“It seems to me to involve grave issues,” wrote District Commissioner Alan Wolsey Cardinall to the acting commissioner of the Northeastern Province of colonial Ghana's Northern Territories Protectorate on June 14, 1919. Referring to the question of policing northern Ghanaians’ migration across recently established district boundaries, he questioned “how … one is to draw the line and to distinguish when a man is a fugitive from work or when he is an emigrant to better his condition.”Footnote 1 In his letter, Cardinall invoked the principle of “freedom of choice” and the risk of colonial chieftaincy becoming “tyrannical.” Four months later, in a prelude to more than a decade of debate on the subject, the commissioner of the Northwestern Province gave a less ambivalent interpretation: “[M]alcontents,” he wrote, “will try to skip to the side where there is little or no work.”Footnote 2
At the heart of this debate about migration was the question of how colonial officials would reconcile colonial labor demands with plans for agricultural and infrastructural development. By 1919, the Gold Coast administration had already begun to envision the region's labor as a resource for development plans in the export-producing South. Northern laborers were sent south to construct large-scale road, railway, and port infrastructure.Footnote 3 Within the North, officials embarked on a much more modest set of “improvements,” demanding forced labor to build both “administrative” and “economic” roads and to construct rest houses for visiting officials. Then, with the formalization of Governor Francis Gordon Guggisberg's ten-year development plan in the 1920s, recruitment of long-distance labor continued alongside the first efforts to cast northern infrastructure and agriculture as elements of development.Footnote 4 These plans also relied on forced labor recruited by chiefs and administrators. By the late 1920s and 1930s, when the British government worked to shape and then comply with international forced labor bans, officials used a new rhetoric of “community interest” to legitimize and institutionalize local demands on northern labor.
Using records from several neighboring districts in colonial northern Ghana, this article builds on Carola Lentz's insight that the infrastructure of colonial administration not only constituted “the material framework of the new territorial order” but “functioned as an instrument in defining and asserting it.”Footnote 5 It asks how demands for labor to build this infrastructure helped define the practice, experience, and limits of colonial authority. Colonial officials did not have the power to fully define the terms on which labor demands were contested.Footnote 6 Colonial developmentalism became embroiled in the local politics of labor as residents devised strategies to cope with the unpredictable and erratic application of colonial force. These strategies, in turn, drew officials into local struggles to attract followers and manage the burdens of an extractive state.
Part I uses monthly diaries of district commissioners (DCs) to show how northerners’ strategies of resistance, combined with the colonial reliance on chiefs for labor recruitment, forced officials to confront the limits of their control over labor.Footnote 7 Scholars have chronicled the widespread use of fines, imprisonment, and other forms of coercion as well as the prevalence of desertion and avoidance in response to recruitment for colonial road building in southern Ghana.Footnote 8 By focusing on local recruitment within the North, this article is able to show how labor demands and strategies of resistance conditioned the first plans for northern agricultural development and efforts to “modernize” rule through automobile transit. These plans neatly sidestepped growing British and international critiques of forced labor, which continued to make exceptions for public works and local projects.Footnote 9 Meanwhile, administrators grafted new demands onto existing structures of force and resistance that continued to draw officials into the local politics of forced labor.
Colonial officials who attempted to impose and legitimize demands for labor found themselves enmeshed in larger processes whereby chiefs and communities struggled to attract residents and secure their labor and loyalty. The clearest illustration of this dynamic came when plans for labor and development became implicated in questions of intraregional mobility. Part II examines cases in which northerners crossed district boundaries to avoid labor demands. Only occasionally did the terms of debate among officials reach the self-importance of Cardinall's concerns about “freedom.” Much more commonly, officials worried about maintaining the infrastructure of rule, pleasing their superiors, and cloaking their practices in the terms of colonial trusteeship. Throughout, northerners forced colonial officials to treat labor not just as something that could be extracted from bodies, but also as a political act involving subjects, chiefs, and officials.
Part III continues with a discussion of the early 1930s, when signers of the International Labour Organization's Forced Labour Convention formalized exceptions for labor that was demanded in the “direct interest of the community” or constituted “minor communal services.”Footnote 10 Far from eliminating demands for unpaid labor, these initiatives increasingly enshrined them in emerging practices of development.Footnote 11 By viewing these changes against the backdrop of pervasive local resistance and growing debates over migration, this article shows that officials were well aware of intense struggles among chiefs, administrators, and constituents over the boundaries of communities and the definition of their interests. In reaffirming the idea of “communal interest,” officials not only relegitimized the use of forced labor, but also washed their hands of the politics of raising it.
Since the 1970s, scholars have explored how demands for labor and a lack of colonial investment in northern Ghana resulted in the “underdevelopment” or “non-development” of the region.Footnote 12 While northern Ghana's marginality in colonial planning was profoundly important to its long-term economic trajectory, it did not make the region any less the object of colonial developmentalism. Jeff Grischow illuminates how successive colonial administrations came to cast the North's hinterland status as a virtue, particularly in contrast to what they saw as the “uncontrolled” or “premature” development of the southern economy and politics. This framework allowed officials to layer successive visions of change on top of “invo[cations] of community … to maintain state power while exploiting the resources of the colonies.”Footnote 13
The observation that development planning was important even in zones of underdevelopment reflects what scholars have long identified as the “marvelous ambiguity of the word development,” to alternately denote a naturalized process, an idealized model of change, a goal of state policy, and a justification for state intervention to achieve it.Footnote 14 This article approaches development as an emerging framework for intervention in the daily lives of colonial subjects, particularly as it related to building the infrastructure that supported colonial administrative and economic planning. Throughout the interwar period, demands for forced labor were the key practical manifestation of colonial incursion in the Northern Territories, and over the period they came to be justified as part of colonial “development” and, ultimately, as in the “interest of the community.” Studies of forced labor in Ghana and elsewhere have often focused on high-profile projects for infrastructure and development and on the use of forced labor for private interest.Footnote 15 In contrast, this article shows that, even in times and places where funding was sparse—and perhaps particularly in these cases—unpaid labor became a central component of emerging colonial developmentalism.
Part I: Administration, Forced Labor, and Development, 1920–1930
British conquest of what became the Gold Coast's Northern Territories Protectorate was a largely speculative and strategic endeavor.Footnote 16 Direct taxation quickly proved impractical, and in 1907 the colonial government eliminated tolls on the caravan trade, the only revenue stream that made significant contributions to the Northern Territories’ budget.Footnote 17 Thus, when civilian staff replaced military officers in 1907, they faced a strict colonial unwillingness to expend resources on the region—a sentiment notably captured by Governor Thornburn's 1912 statement that “the Northern Territories must be content to wait their turn” in favor of government investment in the South.Footnote 18
In lieu of taxation, the Northern Territories administration turned to placing demands directly on the bodies of northern men and women. In addition to demanding northern labor for large-scale infrastructure projects in the South, officials required local labor to satisfy DCs’ constant preoccupation with “trekking” through large districts.Footnote 19 Northerners were ordered to clear and clean roads that ran between district capitals, that served as boundaries between districts, and that gave officials access to towns and villages they wished to visit. Northerners built rest houses for officials, and, before the widespread use of official motorcars in the 1920s, DCs on tour had an almost inexhaustible demand for “carriers” (head porters).Footnote 20 Road labor quotas were officially set at six days per quarter for adult men, though demands varied in practice over time and space and women as well as men were targeted. Demands for carriers and rest house buildings were officially sanctioned, but there was no set quota or limit on these impositions.
The history of enslavement in the region provided official rationale for colonial labor demands, which were cast as an extension of tributary relationships that, officials argued, had been disrupted by slave raiding and imperial conquest by Asante, Babatu, and then the British. This vision led officials to turn a blind eye to domestic slavery in a broad sense and to further classify various forms of labor exploitation by chiefs as forms of taxation.Footnote 21 Imagining that northerners had long been a source of coerced labor, colonial officials argued that the colonial state and its agents were the legitimate beneficiaries. Despite this rhetoric, administrators prioritized meeting the essential goals of colonial administration over understanding existing hierarchies, and officials recognized the authority of a variety of earth priests, chiefs, and local strongmen. Officials were more than willing to unseat chiefs who failed to meet administrative demands and, conversely, to support appointed “sergeant major chiefs” who often appeared to be nothing more than hired guns.Footnote 22
Northern officials had tremendous autonomy and very few resources. Between 1908 and 1930, the total number of DCs and clerks in the Northern Territories was between twenty and twenty-seven, creating a ratio of one official to nearly thirty-five thousand residents and an area of nearly two thousand square miles.Footnote 23 At the discretion of individual DCs, demands for labor were arbitrary, sporadic, and highly localized.Footnote 24 As administrators shuffled among districts, labor regimes could change dramatically and unpredictably. Latitude for official violence was profound, but the practical extent of enforcement was consistently limited. Lawra-Tumu and Navrongo-Zuarungu, two Northern Province districts, present the range of administrative demands and tactics that were established by 1920, highlighting northerners’ well-developed methods of avoidance and resistance. DCs, like chiefs, grappled with the threats to colonial authority presented when northerners refused to “follow” designated chiefs and colonial officials.
From January to November of 1919, A.W. Cardinall was the District Commissioner of the Navrongo-Zuarungu District.Footnote 25 Cardinall's daily records, written while he was working on his first book, The Natives of the Northern Territories of the Gold Coast, reflect his particular mixture of scholarly interest, high-minded ideas about transforming the region, and practical recognition of how administrative demands burdened colonial chiefs.Footnote 26 The most common notes in his records concern the constant complaints from chiefs that locals would not “follow” them. He wrote that this was “a complaint of much elasticity of meaning” and, later, that “‘follow’ is a word I have never understood.”Footnote 27 In his struggle to understand what chiefs meant by “follow,” Cardinall began to recognize that chiefs were using the complaint not as a commentary on the general state of their authority, but instead to explain and excuse their difficulties raising forced labor.
As administrators and chiefs attempted to enforce colonial demands, they faced the limits and contradictions of colonial policy in this administrative hinterland. Cardinall alternated between critiquing Britain for undercutting its own rhetoric of “democracy,” and writing longingly of the ways a “benevolent despot” would be able to realize what he saw as the agricultural and trade potential of the region.Footnote 28 He often concluded such scholarly and ideological musings with more practical recognition that administrators did not need to concern themselves with the workings of authority as much as they simply needed northerners to meet colonial requirements. “From an administrative point of view,” he wrote in July, “it matters little whom they ‘follow’ as long as law and order is observed.”Footnote 29
In November of 1919, just prior to the yearly demand for dry season labor (from December to March), Cardinall was replaced by W.E. Gilbert, a thirty-four-year-old recruit who came directly from his military service in the First World War.Footnote 30 Over the eight months that Gilbert worked in Navrongo-Zuarungu, he saw with fresh eyes the extent of local resistance to forced labor. Gilbert's responses highlight how administrators used the available tools of colonial administration to respond. While on “trek,” Gilbert's usual practice was to stop in a town and note the condition of the roads and rest house under a chief's jurisdiction. He then spoke (through an interpreter) to the chief. Nearly everywhere he stopped, Gilbert fielded complaints that certain sections or towns had refused to “follow” chiefs with respect to forced labor. Consulting his lists of villages, Gilbert determined whether the chief had jurisdiction over the area mentioned and, if so, arrested the headmen he found responsible and imposed collective or individual fines.
Speaking to elderly people in the area in the 1990s, Allman and Parker found that memories of violence, punishment, and its threat pervaded narrations of forced labor in the first decades of colonial administration, when government-recognized chiefs found their work crystalizing around the role of labor recruiter.Footnote 31 Some hints of violence and its threat appear even in the self-reported diaries of officials. Gilbert noted the “Maxim Gun for this Station” when it was returned from Tamale, and when he turned the post over to his successor in July of 1920, he described the process as “handing over the books and ammunition.”Footnote 32
Fines, imprisonment, and the threat of violence were the constant tools of colonial administration, but it is important to remember that the extent of official intervention was often limited. For many chiefs, most of the time, complaints of “not being followed” could explain months in which colonial rule had been more or less ignored and cast vague, collective blame that resulted, at most, in the punishment of a few. Colonial officials often hesitated before challenging local leaders who seemed to have authority. In a series of complaints between 1919 and 1920, successive DCs noted persistent refusals of labor and loyalty by followers of Anam, the chief of Yoragu, just outside the town of Bolgatanga. Despite being fined and imprisoned, officials were loath to unseat a chief who had the power to organize constituents, even though they concluded he was “doubtless at the bottom of all the trouble in this part of the District.”Footnote 33
Monthly diaries suggest that at times officials, like chiefs, were drawn into the complex process of coercing labor. In July of 1920, Navrongo-Zuarungu DC Freeman was required to supply railway labor for the South. At first, recruitment seemed simple. Between August 25 and 26, more than three hundred recruits reported to Freeman's office, and he had “to stop now on account of literally having no more money to subsist any more people.”Footnote 34 Given the growing popularity of labor migration to cocoa farms in the South, an opportunity to be boarded by the government en route may have seemed an attractive option.Footnote 35 A month later, however, Freeman began to report on the widespread return of recruited laborers, listing complaints about the road, the difficulty of getting work and pay, and the nature of the work itself.Footnote 36 Despite Freeman's attempts to “make an example of” particular chiefs and imposing fines for “obstructing recruiting,” he ended up simply paying extra travel money for laborers to return south.Footnote 37
Across the Sisilli River from Navrongo-Zuarungu, in the Lawra-Tumu District, forced labor demands were particularly onerous. Between June and August of 1920, C.B. Shields was posted to the sparsely populated eastern (Tumu) side of the district to make residents build and maintain the hundreds of miles of roads that connected the area to the more densely populated regions to the west and north. A young official named A.C. Duncan-Johnstone was called in to recruit railway laborers from the western (Lawra) side before taking over the whole district in September. Their diaries show that even when officials aimed for mass recruitment, they faced the limits of their control.
Shields spent three months trying to force constituents to quickly build and improve the main roads leading from Tumu to the western towns of Nandom and Wa, to the French border to the north, and to the boundary with Navrongo-Zuarungu district to the east. He noted the urgency and scope of the demand, writing that “the present road scheme” called for “every available man” to be put to work.Footnote 38 His frustrations with constituent resistance are clear. On the first day of a tour, he remarked that inspecting roads made him “furiously angry … the only way to get anything proper is to stand all day and watch every inch of it.” In one of the roads near Tumu, Shields found that a constable had “taken it into his head that there should be a … bulk across the road … it will take 20 men all day tomorrow to unmake it.” In Pina, “nothing has been done on … the road that I wanted, but some that I did not want has been done.” In Dolbizan, where a “[c]onstable could only raise 19 men” of the 100 that Shields had demanded, he found that they had “done a little work, where it was already very good and left all bad places.” In Jeffisi, he found that constituents had spent “20 days messing up the road where it was already good.”Footnote 39
For a brief period, direct supervision and threat of force allowed DCs like Shields to meet administrative demands for road labor. Returning along these roads late in the month, Shields paused to count the hundreds of workers on each section, and “stopped” to “speed things up” where necessary.Footnote 40 By August, when Shields was nearing the end of his term at Tumu, he had concluded, “In the future unless I can be there personally I shall not start any work.”Footnote 41 In October, when Duncan-Johnstone took over as DC of both Tumu and Lawra, he observed the aftermath of Shields’ work. It was an uneven landscape. The residents of Golu, taken with their chief to work on the road to the east of town, had left “a very bad piece of road” on the western road leading to Tumu. At Pullima the road was “as good as anything in the LORHA district,” and in Bellu, a site of frustration for Shields, he remarked, “Not only have [constituents] turned out in large numbers, but they have done very excellent work, and from what I saw of the crowds on the road today, seem to be most cheerful, the work being performed to much singing and dancing.” The same day, he noted that he could “not say the same for the DASIMA division … they have done very little work and very few people have turned out.”Footnote 42
More striking than the variation in the work itself was Duncan-Johnstone's sense that he had little control over the practice or politics that created such stark local differences. In August, after reporting on longstanding recruitment issues in the town of Ulu, Duncan-Johnstone reflected on his limits by ruefully quoting a 1911 poem called “The Passive Resister” by an administrator in Nigeria. The chosen passage read, “Of course it's all right if he means to fight/You can easily burst his bubble./But if he sits tight, then it's hell's delight/He can give a whole world of trouble.”Footnote 43 In subsequent months, Duncan-Johnstone began to report the surprising fact that some chiefs had taken up road-widening and local road construction “by themselves” or on their “own initiative.”Footnote 44 The northern administration was pleased, even if Duncan-Johnstone knew little about how or why chiefs had suddenly found road building to serve their economic, personal, or administrative purposes.
Despite the continuation of limited oversight and even more limited budgets, over the course of the 1920s officials began to more seriously consider the future of northern economic change. Demands for long-distance labor competed with hopes for agricultural transformation in the region, as DCs and chiefs alike feared the practical and ideological implications of young men's journeys to the South. Officials recognized that even local demands for carriers, road laborers, and construction projects threatened to disrupt food production in the region, raising the specter not only of suppressed export production, but also of more basic subsistence crises.Footnote 45 By the 1930s, political action and African accumulation in the South allowed administrators to reconceive small-scale food production as a path to development that would avoid the perceived threat of “uncontrolled” development.Footnote 46
In 1919, Gold Coast Governor Frederick Gordon Guggisberg released a ten-year development plan for the colony, the first of its kind in British colonial Africa.Footnote 47 Guggisberg's plans for growth focused primarily on the South, but the plan also articulated a vision for longer-term growth in northern agriculture, education, social welfare, and infrastructure. Transport was both the literal and figurative instrument for this transformation—as planners envisioned an extension of the railway (and later, the road network) as the “Highway of Progress” for the North.Footnote 48 Throughout the decade, northern officials continued apace in their labor demands. The “highway of progress,” it seemed, would be an unpaved road, requiring annual maintenance by unpaid labor.Footnote 49 Meanwhile, as chiefs gained experience with labor recruitment, officials redefined road building as part of chiefs’ rights to labor for so-called “communal” or “public” purposes.
The rise of motorized transport combined with the newly technocratic interests of the colonial government to create a new set of labor demands on northerners as well as a new set of rationales and staff to implement them. DCs became concerned with the technical requirements of motor transport, and allocated scarce resources to hire road overseers and contractors and to offer nominal payments to forced labor gangs. Demand for “carriers” decreased dramatically by the mid-1920s as DCs began to tour by automobile.Footnote 50 At the same time, DCs’ motorcars changed the frequency and pattern of interactions with northerners, with trips that used to take over a week now being completed in less than a day.Footnote 51 By the late 1920s, DC diaries reported nearly constant visits to the motorcar-accessible towns under their jurisdiction.
In the 1920s, the sudden appearance of private lorry transport had, at best, limited effects on the value northerners saw in roads.Footnote 52 It is certain, however, that it had a negative influence on the experience of unpaid road labor, as lorry traffic took a heavy toll on unpaved roads, increasing the frequency with which constituents were called for repairs.Footnote 53 Making roads accessible to motor vehicles without spending much money also spurred additional schemes.Footnote 54 In June to July of 1926, Northern Provincial Commissioner P.F. Whittall began to experiment with local manufacture of road-making materials. Ordering the laborers at his station in Navrongo to make a kiln out of an old government bungalow, he then had them pack it with clay-and-earth bricks to use on local culverts and bridges. While Whittall seems to have paid the staff at his station, the brick-making scheme relied on forced labor, recruited and enforced by colonial staff.Footnote 55
Administrators adapted old methods to the new tools and demands of developmentalist administration. At the height of the rains, DCs in motorcars monitored the effects of water on roads and bridges, noting the improvements that would need to be made. Then, as the rains began to ebb, the real work began. In October of 1926, Whittall recounted his varied practices for controlling unpaid labor. On October 2, he sent a contractor, Abudulai, to oversee both paid and unpaid labor on the main Tamale-Navrongo road. On October 7, Whittall asked the chief of Sandema to initiate and supervise construction of a “dry-weather motor-track” connecting his jurisdiction to the main road, reporting the chief's assurance that “the work can be done without any trouble at all.” On the twelfth, Whittall went to visit the DC of South Mamprusi as he supervised construction of a low-cost bridge over the White Volta River, checking up on Abudulai's work on the way home. Back in his headquarters, Whittall arranged for a special road to be made so that the governor could visit a government cattle kraal by motorcar, and reported on the success of his brick-making scheme and its promise for future roadwork.Footnote 56
Meanwhile, resistance to labor demands continued. In the Wa district, for example, DC Sumner complained consistently from 1926 to 1928 about “refusals.” In response, he turned to the toolkit of punishments, fines, and oversight forged by his colleagues in the early 1920s—“raising a big row” (August 1926), issuing fines, and assigning extra work as punishment (June 1926, September 1928, October 1928), arresting compound heads (October 1928), and using orderlies and constables to “put the fear of the DC” into recalcitrant workers (October 1928).Footnote 57
By the mid-1920s, colonial reliance on forced labor had become the focus of international as well as local critique. The “Slavery Convention” passed by the League of Nations in 1926 highlighted new European debates about when colonial forced labor constituted “conditions analogous to slavery.” As Suzanne Miers and Frederick Cooper have each argued, British and international advocacy on the subject of forced labor relied on drawing distinctions between forced labor for private enterprise, which was condemned, and for public works, which would simply need to be regulated. Furthermore, and particularly important in light of their reliance on chiefs, the British successfully advocated for the retention of labor demanded by “native custom” and in “personal service” to chiefs who were otherwise not compensated for their administrative work.Footnote 58
As chiefs gained greater experience with labor recruitment, DCs began to report more cases like those seen by Duncan-Johnstone in 1920, in which chiefs started road and bridge projects on their “own initiative.” Official interpretations of “native custom” allowed for administrators to sanction such “initiative” as an extension of chiefs’ rights to demand “labor and personal services” from residents in their jurisdiction. In this interpretation, administrators were charged with regulating the extent and limits of these demands.Footnote 59 The case of Kayani, the chief of Tugu, reveals the slim distance between charges of slavery and the justification of forced labor for public works.
Kayani entered district reports in 1920, when Duncan-Johnstone reported that the chief could make “roads without assistance.” The DC supported Kayani by punishing “malcontents” who refused to follow the chief.Footnote 60 Six years later, Kayani continued to meet new demands for motorable roads. In December of 1926, the DC of Lawra-Tumu reported that Kayani and the chief of Nandom had both built “excellent” roads and bridges “without help or supervision,” praised Kayani's bridge as “show[ing] up any other work on roads in the district,” and remarked that “the Public Works Department could take a lesson from the job of work.”Footnote 61 Concerns about chiefly accumulation in the 1920s, however, made officials inquire more deeply into chiefs’ control over labor. Just eight months after the glowing reports of Kayani's roads, a broader investigation into chiefly affairs in the area raised official concern about his ability to exploit constituent labor for his own farms. DC Eyre-Smith reported that Kayani was “using the people in his villages as his slaves.”Footnote 62
Eyre-Smith's invocation of “slavery” relied on the argument that Kayani's demands for “personal services” were excessive and presented a threat to the economic and political stability of the area. Eyre-Smith speculated that flight from the district and competing demands on constituent time had led to depopulation, agricultural stress, and starvation. He went on to argue that colonial labor demands had provided Kayani with excuses to punish his opponents throughout the 1920s.Footnote 63 In response, Kayani was removed from his chieftaincy in October of 1927. His compound was demolished, his farms were seized, and he was sentenced to six months in prison in Tamale. (He was allowed to return to “live quietly” in Tugu in February of 1928.)Footnote 64
Crucially, the critique of Kayani's practices did not extend to his ability to raise labor “on his own initiative” for roads and other “public” purposes. By April of 1930, the new DC of Lawra-Tumu complained that Kayani's replacement was failing on just this front, as constituents refused to work and “ran ‘for bush’” when the DC summoned them. Arguing that the administration needed “a man with a heavy hand [to be] installed as chief,” he advocated that the administration turn, again, to Kayani.Footnote 65 When the DC was instructed to find a “definite charge” against the current Tugu chief to justify his removal, he turned to the administration's new developmental structures, charging him with halting new locust control efforts as soon as military officials left town.Footnote 66 In June of 1930, Northern Provincial Commissioner Whittall approved Kayani's reinstatement as chief of Tugu.Footnote 67
Part II: Negotiating Mobility as Resistance
When faced with constituents’ refusals of roadwork, DCs turned to the tools of colonial punishment: fines, imprisonment, violence, and unseating chiefs. When chiefs could recruit labor “on their own initiative,” officials defined roads as in the public interest and asked few questions about how labor was raised. This section considers cases in which intraregional migration, in contrast, garnered outsized attention from northern officials. The ability of even small groups of northerners to cross out of a DC's jurisdiction presented new challenges to state coercion. In these cases, DCs became the leaders that constituents refused to “follow.” As a result, government demands drew officials, like chiefs, into the politics of forced labor.
In January of 1919, Michael Dasent, then DC of Tumu, visited Santejan, a chiefly jurisdiction on the district's far eastern border. Stopping in the town of Gwosi, he reported that residents had refused to do road and rest house work and, it was suggested, had burned down the commissioner's stables in protest. When the Santejan chief said that “he could never get labor and that they took no notice when he told them to clean the road,” Dasent moved through his arsenal for enforcing labor demands. After the Santejan chief fined the town ten sheep, Dasent “addressed” them “on the subject of their evil ways,” replaced the Gwosi chief, and stationed a constable in the town. These reports occasioned no comment from Dasent's superiors. The Santejan chief made similar complaints about the people of Kalarsi and Nbenya but then made an allegation that would become a major source of administrative anxiety. He reported that the residents of Nbenya had not only refused to work but were, in addition, “migrating to the Navarro District.”Footnote 68
The following month, Dasent's replacement, E.O. Rake, sent an inquiry to Northwestern Provincial Commissioner H.M. Berkeley about the Santejan case. Reporting that “the people of a village called NAMANEA [Nbenya] … have crossed over into the NAVARO district, their complaint being that there was too much work in this District,” Rake pointed out that while he was “of the opinion that a native is allowed full liberty in choosing his place of residence,” he knew that policies might differ at his new post and asked “what steps” might be necessary “to get these men back.”Footnote 69
Over the next ten months, Rake's inquiry sparked debate over the relationships among colonial labor demands, administrative control, and competing plans for agriculture and infrastructure. While they were couched in terms of competing principles of “freedom” and “authority,” these debates expose competitions among administrators to keep potential laborers within their jurisdiction. In 1919, the boundary between Lawra-Tumu and Navrongo-Zuarungu districts also separated the Northeastern and Northwestern Provinces. Berkeley wrote to his counterpart in the Northeastern Province with the seemingly straightforward request that he “instruct” the DC of Navrongo-Zuarungu to “send these people back.” “They ran away to avoid work when called upon by their Chief,” he argued, and “if this is allowed … [t]he people [would] desert from one side or the other according to the way they are called upon to supply labour.”Footnote 70
Berkeley and Rake's counterparts in the Northeastern Province, however, saw things differently. Provincial Commissioner S.D. Nash and DC A.W. Cardinall both displayed an early interest in northern small-scale agriculture and began to craft an argument against “interfering with free migration,” if “the natives … have changed their residence merely to get better farming land.”Footnote 71 Cardinall moved from the specific to the general, raising the “grave issues” that began this article. He questioned the rights of chiefs and the colonial government to place this kind of restriction on local migration. Nash sent a curt note back to Berkeley stating that he would take no action against the migrants.Footnote 72 When Berkeley turned the case over to Chief Commissioner H.W. Leigh, he contrasted Nash and Cardinall's defense of “freedom” with fundamental concerns for authority and respect, calling the migration “a deliberate attempt to scorn the head chief and so the District Commissioner.”Footnote 73
The letters between Nash and Berkeley reveal the administrative pressures and competitions that undergirded the high-minded debates between “freedom” and “authority.” Tumu, where DC Shields was about to undertake his major push for road construction, was sparsely populated. Recognizing the reality of widespread discontent among northerners asked to labor in the district, Berkeley sought to use district boundaries to set a limit on residents’ options for avoidance. In contrast, Nash and Cardinall, seeing migration as a benefit to agricultural production and labor supply on their side of the boundary, had every reason to defend freedom of movement.Footnote 74 Over the subsequent years, debates between Lawra-Tumu and Navrongo-Zuarungu became more entrenched, even as the northern administration shuffled officials among districts.Footnote 75 Like the chiefs to whom they delegated the task of raising labor, DCs found that competition for “followers” could be fierce.
By 1924, as officials in Lawra-Tumu began to complain of additional out-migration to neighboring districts, the chief commissioner came up with an ad hoc solution that satisfied no one. He required migrants to return to Tumu to fulfill their annual quota of road labor, regardless of where they currently resided.Footnote 76 In July 1925, Austin-Cathrie, DC on the Navrongo side of the boundary, recounted that a headman “at once jumped up and said he did not mind strangers settling on his land but he would expect them to pull their weight on his portion of road work.” A neighboring chief feared that demanding labor from “people [who were] deserting” to another jurisdiction would “cause trouble” between him and the host chief.Footnote 77
Officials likewise fretted over the consequences of a scheme that would make one DC in charge of either the labor or roadwork from another's jurisdiction. After years of supervising labor, they knew the work would only involve them in local politics and, they feared, cause conflict with other DCs. Abandoning the plan, a new chief commissioner summarized its hopelessness, arguing that the DC of Navrongo-Zuarungu had done “all he can do to cope with the roads in his own District.”Footnote 78 Given the well-established resistance that northerners had given to colonial practice, “coping” with their road quotas was as much as DCs could do.
In the midst of trying to stop northerners from migrating between districts to avoid forced labor, officials began to encourage it in the interests of agricultural development. In April of 1923, Chief Commissioner Leigh proposed the first tentative plans to resettle northerners from both Lawra-Tumu and Navrongo-Zuarungu districts to the “more thinly populated South” of the region, in the Gonja and South Mamprusi districts.Footnote 79 He instructed officials to “endeavor to get 5 good families from the environment of NAVRONGO to come down,” promising each “family” ten pounds from the newly established Northern Territories Development Fund and one thousand seed yams for planting. Recognizing the potential deterrent presented by increased labor burdens in less populated areas, he specified that, “these people are not to be called on for communal labor for two years.”Footnote 80 Leigh's idea, which later became known as the “Abangabisi Migration Scheme,” also became embroiled in the politics of labor. Here, development schemes became an opportunity for a small number of constituents to gain recognition of their complaints against a chief and to secure official reprieve from the labor demands of the state.
Leigh's initial plan did not come to fruition. After a flurry of protest from Cardinall about the legality of such a scheme, officials found that they had no interested migrants in the first place. The idea was left dormant until 1928, when it was given new life by early plans for regional agricultural development that saw the southeastern portion of Northern Province as a place of agricultural opportunity. In the late 1920s and 1930s, concerns about “overpopulation” in the far north combined with an emerging interest in supporting small-scale agriculture, rather than a singular focus on export production.Footnote 81 Given the preceding decade of debate over interdistrict migration, officials clearly recognized that intraregional mobility was already established practice.Footnote 82 They nevertheless envisioned that colonial direction was needed to shape and control mobility.
Northern Provincal Commissioner Whittall argued that promoting migration of the kind that was already common would “gain the confidence of the natives over this question of migration.” The “question,” it seems, was less about migration itself than about the risks of working with the state. “Once their fears … are overcome by seeing that it can be done without any harm coming to them,” he argued, “then we might get greater numbers to move farther a-field.”Footnote 83 The Abangabisi Scheme thus anticipated later plans for resettlement tied to tsetse eradication in the 1940s, when attempts to resettle groups of five hundred to two thousand people were commonly tried. Both of these efforts were then dwarfed by plans to resettle populations to the uncultivated lands of the Gonja District, culminating in the attempt to transform the area into a zone of export production during the Gonja Development Scheme in the 1950s.Footnote 84
In December, Cardinall filled in as Northern Provincial Commissioner. He saw the migration scheme as an opportunity to solve a longstanding dispute in his old district. “On my arrival here in December [1928], I found that the age-old quarrel at YORAGU was still alive,” Cardinall explained to his superiors. “[I]t seemed to me that a settlement of this dispute would be attained if the discontented were to move.”Footnote 85 The “discontented” were constituents living outside of Bolgatanga who administrators called “the Abangabisi.” First noted in the early 1920s, this group of people had been involved in a long-running effort to switch their allegiance to Anam, the chief of Yoragu who had so frustrated DCs like Cardinall throughout the decade.Footnote 86 It was the kind of dispute over whom northerners would “follow” that DCs despised, and Cardinall's plan seemed to kill two birds with one stone.
By 1929, ambitions for the scheme were high, with administrators estimating the movement of seventy-five compounds and a total of fifteen hundred people.Footnote 87 Instead, the scheme became a narrow opportunity for a group of northerners to contest the demands of forced labor by putting themselves in jurisdictional limbo. As the Northern Territories administration dispatched a geologist to report on “possible sites for new villages” and worked out the details of the scheme (including the perpetual question of cash payments), officials raised the question of whom the migrants would “follow” once they settled. Rather than requiring the migrants to follow a divisional or village chief—particularly one that would bristle at the prohibition on demanding forced labor—the migrants were authorized to follow the government-recognized Mamprusi king.Footnote 88 Beyond the initial promise of two years’ reprieve from forced labor, officials recognized that when the Mamprusi king offered such arrangements to discontented northerners, it became impossible for local chiefs to enforce colonial labor demands.Footnote 89
Even as Chief Commissioner Leigh reported proudly in June of 1929 that some “eleven hundred people of all ages” had already moved in accordance with the scheme, the Northern Provincial Commissioner and DCs of Zuarungu and South Mamprusi struggled to find out how many migrants had “really” migrated and how many were simply crossing borders to avoid farm clearing or their “24 days” of road labor.Footnote 90 By January of 1930, pressed yet again to ascertain the “real intent” of potential migrants, the DC of South Mamprusi lamented that “these people [the Abangabisi migrants] have chopped and changed about so much that I would never be certain.”Footnote 91 In the end, a mere three compounds were ever determined to have “really” relocated, and the DC of South Mamprusi requested nine pounds to compensate them.Footnote 92 As an early colonial development scheme, the Abangabisi scheme was a colossal failure. For the migrants, however, the scheme had given official recognition to their complaints against a chief and had, most importantly, allowed them to secure an official exemption from forced labor demands. Having spent a decade attempting to prevent northerners from avoiding labor by voting with their feet, northern officials found themselves sponsoring just such a strategy.
Part III: Debates over “Community,” Labor, and Force in the 1930s
In 1930, the British administration in the Gold Coast issued a “memorandum on Forced Labour” that sought to begin to bring the colony's laws in alignment with the International Labour Organization's Forced Labour Convention, which was to come into force in 1932.Footnote 93 In the Northern Territories Protectorate, as Chief Commissioner Jackson noted, any legislation on forced labor would affect policy much more than in the South.Footnote 94 Between 1930 and 1935, as officials debated the extent to which systems of unpaid labor could remain at the center of colonial practice, they worked with fuzzy international definitions of “community” and attempted to parse the differences among coercion, voluntarism, and local taxation in-kind.
The Convention outlined two classes of rationale by which colonial officials could continue to demand unpaid labor. Article 2 of the Convention specified outright exceptions to the ban: (1) “compulsory military service”; (2) “normal civic obligations”; (3) “a conviction in a court of law”; (4) “cases of emergency”; and (5) “minor communal services performed by members of a community and in the direct interest of the community.”Footnote 95 Article 10 elaborated additional cases where forced labor “exacted as a tax … for the execution of public works by chiefs who exercise administrative functions” would be acceptable, at least for the short term. In these cases, several additional provisions required that such work be, among other things, of “direct interest for the community,” that it not impose “too heavy a burden” on this community and that workers would not have to travel long distances or be asked to labor in ways that would interfere with social and religious life or agricultural production.Footnote 96 In the case of labor demanded under Article 10, the convention required that signatories make an annual report about the extent and character of these demands.
When the Convention was first applied in 1932, the colonial government classified unpaid roadwork under Article 10. In November, the chief commissioner of the Northern Territories asked northern officials to report the “extent to which recourse has been had to forced or compulsory labor in your area, purposes for which employed, sickness and death rates, hours of work, methods of payment, and rates of wages,” specifying that “compulsory or forced labor means labor employed on roads, rest houses and station buildings which has not offered itself at current labor rates of pay.” Replies to this query reveal that officials had come to adopt extremely different definitions of what qualified as “forced labor” in this context. In the Gambaga District, for example, reports for June to September of 1932 included a response by J.K. Syme of Bawku that there had been “no labor” called upon, an estimate from Gibbs in South Mamprusi/Zuarungu that 170 laborers had worked for one week, and the rather remarkable report from Olivier in Navrongo that thirty-three thousand people had worked for six days each, at a total cost of eleven pounds and ten shillings.Footnote 97
One year into the application of the convention, British officials began to reinterpret the law in response to their inability or unwillingness to classify, quantify, and challenge the widespread colonial dependence on forced labor. In a lengthy memo on “Maintenance of Roads Under the Forced Labour Convention,” Acting Colonial Secretary G.C. du Boulay explained the change, first pointing out that estimates had now been prepared that foresaw the “eventual cost to this Government” of substituting paid road labor under Article 10 to be “some £140,000 a year.” Citing “subsequent correspondence received from the Secretary of State,” du Boulay explained that “it appears that the view taken by this Government … was rather too rigid” and that the Colonial Office had issued a ruling that “in the majority of the dependencies in Africa the maintenance of minor local roads and tracks is regarded as a minor communal service” that would be exempt from forced labor bans under Article 2. Under this ruling, the colonial government outlined a three-tiered system of road classification. After designating a handful of “Class A—arterial and main roads” that would be maintained by the public works department, the administration then asked each regional commissioner to classify most other roads as “Class C—local roads” that “exist solely for the benefit of the local community” and that colonial officials could therefore “requir[e] villages to maintain.”Footnote 98
Between 1933 and 1935, northern DCs debated which, if any, of the roads in the Northern Province could be classified as class C “local roads and tracks” under a definition that hinged on the idea of “the benefit of the local community.” The staunchest opponent of this classification was W.J.A. Jones, one of the principal architects of indirect rule in the North. Having served as the secretary for native affairs in Accra between 1929 and 1933, Jones was posted as the chief commissioner of the Northern Territories in 1933.Footnote 99 In a series of letters to his superiors, Jones highlighted the patent absurdity of classifying northern roads as having purely “local” benefits. He pointed out that “2,070 miles of motorable roads in the Protectorate … were originally constructed by the people at little or no cost to Government,” and that under the new scheme more than eighteen hundred of these miles would have to be maintained locally.Footnote 100
Jones pointed out that, in sparsely populated areas like Tumu, roadwork had often required northerners to labor far from home. He drew on the experiences of officials to highlight the inefficiency of requiring forced labor, arguing that “compulsory labor is a form of taxation and, as all District Commissioners have observed, an inequitable form in that it is evaded by a large number. It is also wasteful of labor as of those who attend only a small percentage do more than an hour's work in a day.”Footnote 101 Instead, he proposed that the central government dedicate the relatively large sum of seventy-five hundred pounds a year to pay road laborers.Footnote 102 After more than a decade of experience with attempts to enforce local road labor demands, Jones's memos resonated with district officials. In Navrongo-Zuarungu, for example, officials found that only fifteen miles of the road from Navrongo to Lawra could “easily be maintained as a Class C road.”Footnote 103
Jones's attack on the classification of roadwork, however, relied on an affirmation that other forms of labor, especially for more “local” infrastructure and services like dams and wells, fell firmly in the category of “minor communal services.” In these cases, Jones recommended that officials defer to “native customs” in the requirements for labor, arguing, “legislation would convert what is regarded as a social duty to be performed voluntarily into a legal obligation. And there can be no question as to which is the better system.” Nevertheless, he recommended that limits be set at the same level as the previous road work quotas, “restricting to twenty-four the number of days on which a person may be required to work during a year.” He went on to note that even this would be only a fiction, because “for several years to come, it will be impossible to enforce such a Regulation, as there will be no clerks to keep records.”Footnote 104
It was a loophole through which the northern administration could slide all manner of development demands in the coming years, and officials did so with free legal rein. Just a year later, in May of 1936, the colonial secretary's office attempted to put northern roads back into this category, pointing out that there was “nothing in [the Labour Ordinance No 33 of 1935] to prevent a chief and his people maintaining and clearing any road by mutual consent in accordance with native custom, provided that only voluntary labor is employed.”Footnote 105 By the late 1930s, almost all northern roads existed in this administrative gray area. Lacking funding to pay laborers at market rates, the northern administration left roads to the efforts and whims of individual DCs and chiefs.
Conclusion
This article has shown that demands for unpaid labor, primarily for colonial road building, defined the experience of interwar colonialism in northern Ghana. Marginal to colonial budgets and development priorities, the northern administration sought to build and maintain the infrastructure of rule by imposing labor demands through local leaders. Many northerners resisted colonial and chiefly demands by simply not doing what was asked. When they did so, they found that the threat of colonial violence was constant, but its imposition was sporadic and often limited in scope. Chiefs could not always compel colonial subjects to act, a reality reflected in their frequent assertions to colonial officials that people refused to “follow” them. Colonial DCs came to understand this dynamic even more intimately when they began to suspect that northerners were using migration between districts as another strategy to escape labor demands. In these cases, DCs became the leaders that northerners refused to “follow.” Officials found themselves enmeshed in competitions to secure local labor, a politics they neither understood nor could control.
When the colonial administration began its first plans for infrastructural and agricultural development in the North, they grafted new demands and ideologies onto these existing structures of force and resistance. Motor transportation changed the demand for unpaid labor as well as the mechanisms of enforcement. DCs began to prize chiefs who could construct roads on “their own initiative”—a phenomenon that not only saved money but also allowed them to avoid growing international critiques of forced labor by repackaging it as public service. By the 1930s, as officials negotiated the implementation of the ILO's Forced Labour Convention, they managed to place unpaid labor for roads and other works into the murky space that the Convention created for unpaid labor to serve “community” interests. In the 1930s, as plans and funding for development expanded in the Northern Territories, district officials instead used the presence of “communal labor” as a key criterion for granting local development funds.Footnote 106
It is difficult to ascertain whether this maneuver increased or decreased the forced labor demands faced by northerners. When administrators officially devolved responsibility to chiefs under indirect rule, both the extent and mechanisms of coercion were largely displaced from the archival record. As DCs became less enmeshed in the local politics of labor, they likely created both new possibilities for and new limits on northern resistance. (As the example of the Abangabisi Migration Scheme suggests, while official attention was usually the quickest route to state violence, it could sometimes open up limited opportunities to avoid labor demands.) What is clear, however, is that the Forced Labour Convention created a legal mechanism for officials to reclassify local resistance. In the interwar years, concern that northerners refused to “follow” chiefs and DCs drew official attention to the local politics of labor. In the late 1930s, official allowance for labor demands in “the communal interest” made resistance to road labor an argument against disbursing colonial development funds. Practices that defined the limits of colonial control became, in the development era, a problem for the “community.”
