Slavery, Coercion, and Economic Development in Sub-Saharan Africa

Abstract

Recent debates on the economic history of the United States and other regions have revisited the question of the extent to which slavery and other forms of labor coercion contributed to the development of economic and political institutions. This article aims to bring Africa into this global debate, examining the contributions of slavery and coercion to periods of economic growth during the nineteenth and twentieth centuries. It argues that the coercion of labor in a variety of forms was a key part of African political economy, and thus when presented with opportunities for growth, elites turned first to the expansion of coerced labor. However, while labor coercion could help facilitate short-run growth, it also made the transition to sustained growth more difficult.

How did power and coercion contribute to the shaping of the modern economy? This has been the driving question behind histories of empire and slavery for many decades and remains the subject of fierce debates in both academic and public history circles. Recent contributions have centered on the processes of industrialization in first Britain and then the United States, assessing the extent to which the violence of empire and the coercion of millions of enslaved laborers made this possible.¹ Some argue that enslaved labor, and the goods they produced, were foundational to the rise of industrialization.² Others counter that while individual elites certainly gained from the exploitation of enslaved workers, the scale of production by enslaved labor is not sufficient to explain the large structural changes in industrialized economies and that the institution itself undermined prospects for economic development and diversification in the regions where it was most widely used, like the US South.³

The recent focus on the United States and Britain obscures the fact that these same questions could be, and have been, asked of other parts of the world. Labor coercion was everywhere during the period of industrialization and took a variety of forms in different parts of the world. Enslaved or coerced workers were engaged across a variety of professions and industries and had varied degrees of autonomy and living standards. The ways in which these forms shaped patterns of production and the distribution of economic rents also varied. In a study of Brazil, Nuno Palma et al. show that regions more heavily dependent on enslaved labor struggled later, writing that “there is no evidence that slavery benefited the societies that relied largely on it. Not only is slavery abhorrent from a modern normative perspective, but it also mostly had negative development consequences: while slave-owners and a few narrow sectors profited from it, overall society lost out.”⁴

After the abolition of slavery, coercion remained a key feature of many labor markets, ranging from outright forced labor to indirect means like legal restrictions on wages and mobility. How these structures shaped development has also been debated. In South Africa, for example, what is known as the “capitalism and apartheid” literature focuses on the role of racial discrimination in shaping South Africa's economy.⁵ On one side are scholars arguing that policies that kept African wages low were essential in keeping mines and manufacturing industries profitable. Critics argue that the market distortions of apartheid policies hindered long-run development by reducing the size of the domestic market through keeping majority incomes low and disincentivizing investments in human capital. In a critical review of this literature, Nicoli Nattrass adopts a middle ground: while low wages for African workers were important in maintaining the profitability of mines during the late nineteenth and early twentieth centuries, the shortage of skilled workers made later productivity gains in manufacturing more difficult.⁶

This paper surveys these same debates for sub-Saharan Africa. It focuses not on the Atlantic slave trade, which has dominated studies of labor coercion in Africa, but rather on the nineteenth and twentieth centuries.⁷ The continued use of slavery and forced labor after the abolition of the Atlantic trade—often referred to as the “second slavery”—has been the subject of increasing scrutiny in recent years by scholars critiquing the idea that slavery was an antiquated institution incompatible with the expansion of capitalism. Rather, many regions saw expansion rather than retrenchment of the use of coerced labor during the nineteenth century.⁸ In sub-Saharan Africa, global changes in demand and transport costs saw rapid increases in the production of minerals, cash crops, and forest products. As this paper shows, this process of growth relied heavily on the coercion of labor, either through the expansion of the slave trades within Africa or through other, more indirect, methods of labor coercion.

This paper follows Frederick Cooper in treating slavery and coercion “not narrowly an issue of labor and markets but of political economy,” part of wider efforts by elites to control access to wealth and power.⁹ This paper argues that the persistence of coercion in the mobilization of labor in Africa reflected both factor prices and the structure of political and economic institutions. Periods of rapid economic growth during the nineteenth and twentieth centuries often involved the expansion of labor coercion. As in South Africa, however, the turn to coercive policies also made it more difficult to achieve sustained economic growth.

Varieties of Coercion

Any discussion of slavery and coerced labor in Africa must first deal with the dizzying array of institutions and structures in which power was used to mobilize labor instead of wages. Historically, the emergence of coercion in Africa was favored by comparatively low population densities that made labor scarce relative to land.¹⁰ This not only shaped labor markets but also the broader structure of political institutions. As Sean Stilwell puts it, “it is now axiomatic that in Africa, possession of—or control of—people meant power. Power over people provided access to labor, to the reproductive capacity of female slaves, and to political agents or functionaries who then bolstered the power of elites and other big men.”¹¹ However, the precise structures by which this control over people was exercised varied both between and within African societies. Further, the changing geopolitics of the colonial period led to the introduction of new forms of coercion, both direct and indirect, layered on top of institutions which persisted from the pre-colonial period. Many of these were then inherited by post-independence states, resulting in a complex and multi-layered history of labor coercion well into the second half of the twentieth century. This section reviews these different types in order to illustrate the range of different methods of labor coercion.

Owing to the relative scarcity of labor, slavery was prevalent in African economies long before the beginning of the Atlantic trade. Enslaved labor was key to the trade economies of the medieval Sudanic empires, both in the production of traded commodities and as a commodity itself. Enslaved workers were employed in the salt mines in Taghaza, in the gold mines of medieval Mali, and as both porters and commodities in caravans traveling Trans-Saharan trade routes.¹² The same was true for the medieval Swahili coast, where enslaved workers were employed in food production in the hinterland of coastal trading centers, in the production of exports, as domestic workers, and in skilled occupations in urban settings. Slaves were also traded both within the region and farther afield.¹³

These regions were, and remained, areas of intensive slavery.¹⁴ In these areas, slaves were often a substantial proportion of the population and tended to work in slave-dominated units focused on the production of commodities for the market. Also included in this category was the Cape Colony from the eighteenth century. The Dutch East India Company began importing slaves from Asia and East Africa in the seventeenth century to meet labor demands from European settler farms.¹⁵ Using emancipation records from the nineteenth century, Kate Ekama and coauthors show substantial variation in the activities of both male and female slaves.¹⁶ While most male slaves were laborers, significant numbers were also employed in animal husbandry or clerical work. Similarly, among female slaves, most were domestic workers but others worked in artisanal production or animal husbandry. A small number of female slaves were also employed as laborers. Similarly, in areas of intense enslavement in the Indian Ocean region, female slaves worked in occupations ranging from concubines to prostitutes to domestic servants to water carriers, as well as in mining and textile production. Male slaves were employed in agriculture, transport, fishing, and commerce, as soldiers, and in the bureaucratic service.¹⁷

There were also many other regions of what is known as low-density slavery, in which slaves were a smaller proportion of the population and worked within households. This model of low-density slavery was outlined initially by Igor Kopytoff and Suzanne Miers, who conceptualized slavery in Africa as a matter of degrees of marginality within local communities and stressed the ability of slaves to be integrated within those communities.¹⁸ Slavery of the low-density type tended to emerge in regions with less centralized political institutions, and often lower levels of overall population density. Slaves lived within households and had extensive “face-to-face” interactions with their owners, and frequently became absorbed within their households.¹⁹

Slavery was not the only means of labor coercion, and in most African economies the labor force was composed of both slaves and coerced laborers mobilized through other means. In the context of land abundance, land rights did not often serve to secure debt. This purpose was instead served by pawning, or “a system in which individuals are held in debt bondage as collateral for loans.”²⁰ Pawning was widely used in a variety of contexts, though its importance in the provision of credit might have meant it was more common in regions heavily involved in external trades.

Communal labor was another means of mobilizing labor through power relationships within a village or other community. In the cycle of rural life, cooperative labor arrangements were used for tasks too large to be tackled by the family, such as house-building, the clearing of fields for planting, or the harvest, and often operated on a system of rotating obligations between neighbors or other kin. There remains some debate about the extent to which communal labor was prevalent prior to the colonial period given the limited chronological scope of the oral history evidence on which much knowledge of systems of communal labor relies. Gareth Austin notes in his study of Asante that “it seems likely that the rotating exchange of labour between neighbours or fellow-villagers was a recognized institution in nineteenth-century,” but also that “recourse to it was relatively infrequent because of the availability of slaves and pawns within households.”²¹ On the other side of the continent, Opolot Okia argues for Kenya that colonial officials had an interest in stressing the “traditional” legitimacy of communal labor to justify its use.²²

Colonial states built on and often expanded existing systems of forced labor. They also introduced their own systems for the direct coercion of labor, including labor taxes, military drafts, and forced cultivation.²³ Recent literature on the political economy of colonial rule has suggested that labor was an important source of revenue for colonial states. In a study of French colonies, Marlous van Waijenburg combines records of forced labor requisitions with wage estimates to calculate the total value of forced labor recruitment to French colonial states. She estimates that revenue from such sources could amount to even more than total cash revenue, at least in the early years of colonial rule, though its importance as a revenue source declined over time.²⁴ The decline was uneven, however, and poorer colonial states tended to rely more heavily on forced labor until later in the colonial period.

There were a number of other, less direct, methods to bring Africans into the labor force, either by restricting their options for production or by imposing taxes that had to be paid in cash. Both were imposed widely across colonial Africa, but were most commonly used in colonies with extensive European settlement or foreign enterprises. Hut and poll taxes, for example, were more important sources of revenue in colonies with European and other settler communities.²⁵ While the primary aim of these taxes was to raise revenue, they also served a second intended purpose, which was to compel Africans into the wage labor market in order to earn the cash with which to pay the tax. How effective this was in a context of widespread tax evasion and limited state capacity remains an open question. Similar aims were served by restrictions on African economic opportunities. Perhaps the most important of these included land seizures that created artificial scarcities in the amount of land available. This had the effect of creating competition over land and forcing those unable to access land into the labor market. Restrictions on the production of cash crops by Africans also limited their ability to earn a cash income without participating in the market for wage labor.

This section has reviewed the various forms of labor coercion used in African economies during the nineteenth and twentieth centuries. Their variety reflected the long history of labor mobilization using power and social relationships rather than wages in sub-Saharan Africa, where comparatively low population densities made labor rather than land the scarce factor of production. Systems of property rights and political expansion had thus focused on people and not territory. These systems provided the foundation for processes of economic change in the region during the nineteenth and twentieth centuries, as discussed in the next section.

Patterns of African Long-Run Growth

The first two decades of the twenty-first century have seen a rapid expansion in research on African economic history.²⁶ This work has focused particularly on assembling quantitative data with which African countries can be compared to each other and to countries outside Africa. Challenging ideas that data in Africa are inadequate to support rigorous analysis, this work has generated new insights across a wide range of topics, from trade to public finance and to inequality.²⁷ These data have helped illuminate not only the structure and functioning of historical economies but also the ways in which historical events have shaped current economic outcomes.²⁸

Starting from the late nineteenth century, it is possible to calculate aggregate measures of African economic output, which further enable comparative work. Figure 1 shows annual estimates of gross domestic product (GDP) per capita for a selection of African countries. Previous estimates of historical GDP per capita for African countries were for single countries or benchmark years only, establishing that many countries were already above subsistence (sometimes substantially so) by the 1870s.²⁹ Annual estimates allow for a closer look at patterns of economic growth since the 1880s, rather than just levels. These data show several episodes of growth from the late nineteenth century onward. One period was between the late nineteenth century and the outbreak of World War I, reflecting rapid expansion in the export and government sectors. The outbreak of war and the turmoil of the interwar period also affected African economies, but growth resumed during the 1940s as Africa became a key supplier of strategic goods for the war effort. This phase of growth continued through the postwar golden age.

Figure 1. GDP per capita in a selection of countries, 1885–2008. (Source: Stephen Broadberry and Leigh Gardner, “Economic Growth in Sub-Saharan Africa, 1885-2008: Evidence from Eight Countries,” Explorations in Economic History 83 [2022]: 101424.)

The gains from this growth were by no means distributed equally, and one of the contributions of the recent renaissance in African economic history is to show rising levels of inequality in a number of African countries beginning in the colonial period, often linked to the expansion of export production. In Botswana, for example, the commercialization of cattle exports from the 1930s prompted an early increase in inequality that was compounded by the postindependence shift to diamond exports.³⁰ Similarly, Ghana's level of inequality was already comparatively high in the early colonial period due to the comparatively high incomes of cocoa farmers, which increased more from the 1930s as the incomes of civil servants and skilled workers began to rise.³¹

Not all countries experienced similar rises in inequality linked to export production. Despite being one of the leading export producers of East Africa, Uganda had only a limited increase in inequality during the postindependence period.³² A recent survey of inequality trends in colonial Africa argues that the extent to which export production contributed to rising inequality depended on the capital-intensity of export production.³³ In Uganda, much of the cotton exported was produced by smallholders alongside subsistence crops with the intention of earning a small amount of cash income to hedge against the risk of famine. This contrasts with the large capital investment required for the export of beef or tree crops like cocoa.

Alongside the “renaissance” of African economic history, there have emerged new insights about long-run economic growth and the workings of pre-industrial economies around the globe, which, brought together with the literature on Africa, have suggested that Africa was perhaps not as much of an outlier as earlier literature suggested. Estimates of GDP per capita for Europe and Asia dating back to the medieval period show that, much like African economies, pre-industrial economies were not subsistence economies with little or no growth. Rather, many had levels of per capita income much higher than subsistence level, and experienced periods of growth which could sometimes last decades.

However, the gains from these periods of growth were often erased by subsequent periods of shrinking.³⁴ It was only when the severity and duration of the shrinking in those periods were reduced that the early industrializers were able to achieve sustained improvements in per capita income.³⁵ Even before that, however, these growth episodes contributed to rising inequality in preindustrial economies.³⁶ Anecdotal evidence suggests that the same characterization could be made of sub-Saharan Africa for the period prior to the late nineteenth century.³⁷ Annual estimates of GDP per capita illustrate that most countries were subject to long periods of shrinking as well as growth.

In sub-Saharan Africa, the severity and duration of its economic shrinking is the key factor in explaining its position of relative poverty today. As the regional comparisons in Figure 2 show, GDP per capita in the region was at the same level or even slightly higher than that of leading Asian economies like China and India up to the final decades of the twentieth century. By then, however, the prolonged period of shrinking beginning in the middle of the 1970s contrasted with persistent growth and structural change in Asia.

Figure 2. GDP per capita in Africa, Asia, and Latin America. (Source: Broadberry and Gardner, “Economic Growth.”)

The one partial exception is South Africa, shown in Panel A of Figure 1. While South Africa did suffer significant periods of shrinking—notably during the South African War (1899–1902) and in the final years of the apartheid regime during the 1980s—the rise of manufacturing in the interwar period reduced the vulnerability of the economy to commodity price shocks.

The next three sections each focus on one of these periods of growth: the nineteenth century, the early colonial period, and the postwar “golden age.” They examine the various roles that labor coercion played in the process of economic expansion, and conversely how economic and political changes contributed to shifts in the forms of labor coercion used. While it is not possible to quantify these effects, anecdotal evidence shows that coercion was widely used in the production of important export commodities. The paper concludes by speculating on the ways in which the use of that coercion may have made the shift to sustained growth more difficult; namely, rising inequality, underinvestment in human capital, and institutional instability.

Nineteenth-Century Globalization

The nineteenth century saw two conflicting trends with regard to the use of slavery and forced labor. On the one hand, there were increasing international restrictions on the export of slaves. These were imposed gradually and with different levels of intensity in different regions. On the other hand, however, growing demand for African commodity exports—ranging from minerals to agricultural products like palm oil and cloves and forest products like timber and ivory—increased demand for enslaved labor within Africa. Warfare and political upheaval also expanded the number of enslaved. Slavery and forced labor thus remained persistent and central to Africa's economic and political institutions.

Pressure to reduce or eliminate the export of slaves from Africa began in the first years of the new century, not long after the Atlantic slave trade reached its peak in the late eighteenth. Denmark abolished the slave trade in 1803, followed by Britain in 1807, the United States in 1808, and other European countries shortly thereafter.³⁸ The immediate impact of abolition was limited, as Gareth Austin puts it with regard to West Africa: “the 1807 British abolition act was the effective beginning of the end of the Atlantic trade.”³⁹ However, there remained significant markets for African slaves across the Atlantic in Brazil and Cuba, among other destinations.⁴⁰ Slaves continued to be exported from places like the Gold Coast and Bight of Benin until the middle of the nineteenth century, though efforts by the British Navy to enforce the ban prompted a shift southward to West-Central Africa.⁴¹

Efforts to suppress the trade in other regions began later, and were similarly ambiguous in their immediate effects. The decline in Ottoman privateering in the Mediterranean following peace treaties between North African and European states reduced the flow of Christian captives coming from Europe to North Africa, leaving sub-Saharan Africa as the only source of slaves for North African markets.⁴² The trans-Saharan trade saw significant expansions during the nineteenth century, particularly during the Egyptian cotton boom of the 1860s, when the use of slave labor expanded to meet demand for cotton generated by the shortfall in American exports during the US civil war.⁴³ In East Africa, the impact of the Moresby Treaty of 1822 between the British government and the Sultan of Muscat and Oman—which restricted the geographic range of slave exports from the East African coast—was counteracted by the growing demand for enslaved labor on island plantations first established in the eighteenth century.⁴⁴ In southern Africa, British policies of abolition sparked a proliferation of other varieties of labor coercion (some of which were discussed in section 2) that eventually provided a model for other colonial economies.⁴⁵

The multifaceted impact of abolition, combined with other major shifts in the global economy, meant that the gradual decline of slave exports—and the growth of commodity exports—did not imply a decrease in enslavement within Africa itself, as proponents of “legitimate” commerce had predicted.⁴⁶ In fact, the opposite occurred, and the number of enslaved people increased over the course of the century. This expansion was driven by two primary factors. First was the rise in demand for African commodities driven by European and American industrialization, along with innovations in transport technology that made it cheaper to export lower-value-to-weight goods.⁴⁷ The nineteenth century thus saw rapid growth in the export of minerals, forest products, and agricultural produce from Africa to other regions of the world.⁴⁸ It is important not to exaggerate the impact of this increase for most of the African population. John Tosh notes with regard to West Africa that the high cost of internal transport combined with ecological constraints would have limited the commercial production of exports to a narrow band of territory in the forest zone near the coast.⁴⁹

However, as this section will show, coastal production could have ripple effects far into the interior. One feature of the period was the expansion of plantation production in both West and East Africa. Frederick Cooper, in his study of plantations in Zanzibar, defines plantations as “large-scale, specialised units that developed in order to serve the needs of a vast and widespread market for particular commodities,” and which developed in areas where the soil and climate were particularly suited to producing them.⁵⁰ The rapid growth in the production of cloves—a crop indigenous to Southeast Asia, which before the nineteenth century had never been produced in East Africa—was primarily by plantations on Zanzibar and the nearby island of Pemba. These could range greatly in size, from small plantations with one or two slaves to large plantations with several thousand slaves.⁵¹ The expansion of clove production on the islands also increased demand for food surpluses to feed plantation workers, and plantations were also established on the coastal mainland to produce food crops like grain and coconut. In both cases, the vast majority of the slaves came from the East African interior, largely from the hinterland of Kilwa, in what is today Tanzania and the Lake Malawi region.⁵²

Similarly specialized units emerged in the production of palm oil in West Africa, as both Mohamed Salau and Paul Lovejoy have shown.⁵³ But not all export production by enslaved workers took place on plantations. Rather, as before the nineteenth century, the structures of slavery and production by enslaved workers were diverse. In Asante, slaves contributed to the expansion of gold mining and kola nut production from the 1820s to the 1860s. Kola trees tended to be located at some distance from Kumasi, and so the production of kola nuts by slaves through sharecropping arrangements facilitated the expansion of the Asante interior (discussed below). Gold mining was concentrated in the dry season, and therefore would have been combined with other activities.⁵⁴ Slave labor was also crucial in the expanding production of groundnuts, which were used in Europe to make margarine and soap. However, the form this production took varied, from sharecropping to plantation agriculture.⁵⁵

A second factor in the increasing slave population was warfare and state-building, in which slaves became both captives in conflict and instruments of territorial conquest.⁵⁶ Most notably, the Islamic jihads of the early nineteenth century created a number of newly centralized states that relied on newly captured and existing slave populations for much of their conquest of the frontier.⁵⁷ The numbers of slaves held by these new states could be large. Lovejoy estimates that the Sokoto Caliphate, the largest of the jihad states, had a slave population comparable to that of Brazil or the United States in 1860.⁵⁸ Outside the new caliphates, other expanding states used similar practices. As noted above, the Asante state used slave labor to clear forest and expand agricultural production.⁵⁹

The nineteenth century was a period of significant economic and political turmoil across Africa. Emerging opportunities for export production placed new demands on labor. While, many of the commodities that fueled this growth could be produced using a variety of methods of production, the centrality of slavery and other forms of coercion in the African political economy combined with the constraints of Africa's factor endowment meant that the acquisition of slaves frequently went hand in hand with increasing production of both exports and other goods. As Catherine Coquery-Vidrovitch puts it:

In both West and East Africa, from the King of Dahomey to the Sultan of Zanzibar, African leaders quickly seized new opportunities and enthusiastically opened their realms to new markets, but they did not give up their old sources of revenue. On the contrary, growth in exports was gained through slaves, who were used in armies, as porters in caravans, as labourers on plantations, and as workers in the textile industry.⁶⁰

Colonial Expansion

The increase in the volume and value of exports continued to accelerate during the early colonial period, particularly prior to World War I. This expansion was a key source of the economic growth documented in section 3. As in the nineteenth century, coercion played a key role. How much of colonial export production relied on coercion? Just like in the previous period, this is difficult to say with any certainty. Not all export success stories were stories of coercion, and in other cases colonial efforts to use coercion to increase production had decidedly mixed results. Cotton is one example. Michiel de Haas's account of cotton production in Uganda shows that what became one of the leading centers of cotton exports in the region relied primarily on smallholders producing a small amount of an export crop to supplement family incomes. Families did not rely exclusively on the production of cotton for the market, but planted small amounts of cotton for the market alongside food crops.⁶¹ Across the continent in northern Cote d'Ivoire, however, French authorities gave cotton production quotas to Senoufo chiefs, who mobilized traditional systems of communal labor to meet those quotas. “In contrast to the old system,” notes Catherine Boone, “the product of this collective effort was not a public good: chiefs and lineage heads sold the cotton to the French, used the proceeds to pay head taxes, and kept the rest for themselves.” This redistribution to the chiefs and their households had a negative impact on the workers themselves, straining “the capacity of Senoufo villages and households to provide for subsistence needs.”⁶²

The story of rubber is similar. The most notorious use of forced production was probably the “red rubber” regime in the Congo Free State (CFS) under King Leopold II of Belgium. When the prices for natural rubber spiked in the early twentieth century, Leopold saw it as an opportunity to reverse the failing financial position of the CFS, which he had claimed as his own personal property in 1885. Under the red rubber regime, Africans were required to produce minimum quotas of wild rubber, often at gunpoint. Those who did not produce the set quota suffered brutal atrocities at the hands of CFS officials. While this strategy did produce a short-term rubber (and revenue) boom, it proved unsustainable in the long run. The worst-affected areas saw extensive depopulation as Africans fled the reach of the CFS, and rubber trees were destroyed by overtapping. The international outrage prompted by public revelations about the atrocities committed led to the Belgian government's takeover of the administration of the Congo.⁶³ Still, this brief period left lasting scars on the development of the Congo.⁶⁴ Other efforts to coerce rubber production had more limited impacts, as James Fenske shows in his study of rubber in Benin, where rubber production only took off in the postindependence period, when the incentives shifted for African producers.⁶⁵

One area in which coerced labor played an important, if indirect, role in promoting export growth was in the construction of infrastructure, particularly railways. Although African railway networks were thin on the ground compared to other regions, access to the railway had a dramatic impact on transport costs and made export production profitable in regions where it had not previously been so. Roads had a similar impact during the interwar period. Over time, the construction of infrastructure had a dramatic impact on the economic geography of African countries.⁶⁶ This infrastructure was also built largely by coerced labor. Labor on the railways was difficult, and mortality rates were high, and thus coercion became the norm rather than the exception. In some cases, these were not Africans but rather workers shipped in from other parts of the world, including Chinese workers to South Africa and the French Congo or Indian workers to East Africa or Natal.⁶⁷ Such experiments in importing workers were short-lived (although in some cases historically impactful), but in most cases the workforce was African in origin. Workers were recruited either by the colonial state itself or by African elites on its behalf.

As in the case of the forced cultivation of crops, however, the recruitment of labor for work on the railways had complex effects on African economies. While the railways themselves facilitated the expansion of export production, which can be measured, the recruitment of forced labor for their construction had negative effects, which are more difficult to document. In a study of the French Congo, Alexander Keese records that efforts to evade forced labor on the Congo–Ocean Railway were sufficiently widespread to cause significant decreases in local population levels, to the point of entire villages being abandoned.

Deserters had immense success in escaping forced labor, but they paid an enormous price as well: while we need many more detailed studies on these effects, it is obvious that the experiences destabilized rural communities and affected their cycles of agricultural production and their level of food security.⁶⁸

Last, colonial states implemented a wide range of more indirect means of labor coercion, which were discussed earlier in this article. Such policies were most often found in colonies in which export production was in foreign hands, either those of settlers or foreign commercial enterprises. Labor recruitment was difficult for the same reasons as outlined in previous sections, and in many cases the colonial state intervened to restrict the outside options available to African producers.⁶⁹ Many of these methods were pioneered in South Africa and then adopted by other colonies later on.

One of the great ironies of the colonial period was that public justifications for colonial conquest often referred explicitly to the abolition of slavery within Africa. The Brussels Act of 1890, which laid the groundwork for territorial claims by European powers, obliged the countries signing it to commit to ending the slave trade, also stating that the best means of doing this “was to establish colonial administrations in the interior of Africa, to protect missionaries and trading companies, and even to initiate Africans into agricultural and industrial labor.” Through these statements, “it put an anti-slavery guise on the colonial occupation and exploitation of Africa.”⁷⁰

Once in possession of territory, however, understaffed and under-resourced European colonial administrations needed the cooperation of African elites both to govern and to promote the continued expansion of export production to build a tax base.⁷¹ Both tended to consolidate and even expand the use of forced labor, rather than abolish it. “Paradoxically,” Richard Roberts points out, “many colonial states justified the Scramble for Africa in the later nineteenth century as part of their effort to end the slave trade and slavery and promote African commerce, only to resort to forced labor to build in the infrastructure that would enable the colonial project.”⁷²

Coercion and the “Golden Age”

From the interwar period onward, colonial governments faced increasing pressure to reform their methods of labor recruitment. This pressure came from international organizations like the League of Nations and the International Labour Organization (ILO), and from Africans themselves, who were increasingly shifting from migration and avoidance to organized political pressure as a method of protesting labor requisitions. As a result, both late colonial and postindependence governments faced a set of dilemmas about definitions and dimensions of coercion. As Elisabeth McMahon puts it with regard to Zanzibar, “at what moment was labor ‘coerced’? If the government told villagers to build a school for their children, was this a form of forced labor? If a government told a community to build roads so they could bring their clove crops to market more efficiently, was this coerced labor?”⁷³ The final decades of the colonial period and the first decades of independence saw the adoption of new policies of labor mobilization that tried to balance the pressures to abolish forced labor while at the same time retaining options for accessing unpaid workers through systems of “community” mobilization. This section reviews the ways coercion continued to be used in the postwar decades, even as rhetoric changed in response to political pressures.

The rapid expansion in commodity exports prior to World War I had drawn a growing number of Africans into production for the market rather than for subsistence. Not coincidentally, this period also saw rising political instability across colonial Africa, as those affected by price shocks in various ways began to organize protests and strikes. The proliferation of such incidents prompted a rethinking of colonial policies in metropolitan capitals, and from the 1930s they began to adopt more expansive development policies that intended to look beyond infrastructure to increase provision of education and healthcare services. However, the level of resources available for these initiatives were limited in the straightened financial circumstances of a depression and war.

At the same time, there emerged growing, if uneven, international pressure against the use of enslaved or coerced labor. The Brussels Act had lapsed during World War I, but in its place were clauses in the 1919 St Germain-en-Laye treaty and in the foundational documents of the League of Nations referring to its own obligation to suppress slavery. However, these documents did little to define these terms or set up mechanisms for enforcement, often disappointing antislavery activists. After the emergence of a public scandal about continued slave raiding in Ethiopia, the League launched an investigation into the use of enslaved labor that ultimately led to the adoption of the Slavery Convention of 1926.

This 1926 convention was the first international agreement directed particularly at slavery and forced labor. However, Miers describes it as a “paper tiger.” It balanced a need, which even British and French officials recognized, to take some steps to satisfy international campaigners along with their desire to limit the impact on existing forms of labor recruitment. It gave no timeline for the suppression of slavery, nor did it specify any mechanisms for enforcement apart from delegating the ILO as responsible for investigating forced and compulsory labor practices.⁷⁴

Subsequently, the ILO established a Committee of Experts on Native Labor, which produced a Convention on Forced Labor in 1930. The 1930 Convention banned the recruitment of forced labor for private enterprise, but contained exceptions for military conscription, prison labor, communal obligations, and states of emergency.⁷⁵ The 1930 Convention was subject to wide interpretation in actual colonial practice. Coerced labor continued to be used, but it was framed in terms of community contributions to development projects. This was justified by colonial officials as being in accordance with traditional African practices.⁷⁶ In British colonies, responsibility for the mobilization of forced labor was devolved to Native Authorities, or African local governments. This was described as “cost-sharing,” in which colonial governments provided materials and technical expertise needed for local development projects, while local Native Authorities were responsible for providing labor. This approach was used in a variety of colonial projects, perhaps most prominently in the construction and maintenance of roads, in which the recruitment of forced labor through chiefs became the subject of fraught negotiations about the extent of colonial and chiefly power.⁷⁷

After the outbreak of World War II, many colonies expanded their recruitment of forced labor through military conscription and forced cultivation.⁷⁸ During the war there was a rapid expansion in demand for African commodities like rubber and copper, particularly after the Japanese conquest of Southeast Asia cut off alternative supplies. The need to ensure supplies of strategic goods led to widespread government interventions in their production and marketing, conveniently blurring the distinction between public- and private-made in the Convention. Across a number of African countries, the war years saw the creation of marketing boards to which Africans were forced to sell cash crops at below market value.⁷⁹ At the same time, disruptions in shipping made imported goods scarce, threatening food security in some areas of intensive export production. In Zanzibar, for example, interruptions in the supply of imported rice provided grounds for the colonial government to organize a program of forced cultivation known as the “Grow More Food” campaign, in which officials could requisition land and set required quotas for the cultivation of particular crops.⁸⁰

The transfer of power to independent African governments saw the end or evolution of many unpopular colonial policies. Hut taxes were eliminated or transferred to local governments, and in many countries the powers of colonial-era chiefs were undermined. However, postindependence governments faced many of the same dilemmas as their late colonial counterparts. They needed to fulfill promises to a newly enfranchised electorate that expected a rapid expansion in the state's development agenda. Resources, at least in the early postindependence period, were not scarce due to the fiscal effects of postwar economic growth and an increased influx of foreign aid.⁸¹ However, funds were still insufficient to compete with the private sector for labor, and many of these projects drew on some of the same tools of labor recruitment used during the late colonial period. Benedetta Rossi documents continuities in the recruitment of forced labor in Niger for development projects in arid regions. As opportunities for migrant workers expanded, it was primarily women who bore the brunt of labor demands for development projects.⁸²

In the postindependence period, recruitment of forced labor was often associated with national development efforts, particularly those focused on agriculture. These included projects like Operation Feed the Nation in Nigeria, Operation Feed Yourself in Ghana, and Operacao Producao in Mozambique. In a detailed study of one such initiative, Operation Production in Liberia, Mark-Thiesen notes that the project “was explained as a voluntary initiative for economic development and economic sovereignty that put the nation first (and, hence, the individual second).”⁸³ One well-known example was Julius Nyerere's campaign in Tanzania intended to create rural villages that were self-sufficient and not reliant on export markets. Nyerere called the campaign ujamaa, a Swahili phrase referring to extended family or community. Despite the name, the state had to resort to force to get reluctant farmers to move to the new villages.⁸⁴ In what was then Zaire, the government under Joseph Mobutu used similar rhetoric of national self-sufficiency to compel the production of maize to feed workers in mines and cities.⁸⁵

There remain debates about the extent to which the transfer of power altered the underlying political economy of African countries. Alexander Keese points out that it is difficult to draw direct lines between colonial and postindependence uses of forced labor, and thus he argues that postindependence practices simply reflected an inheritance from the colonial period.⁸⁶ In the case of the French Congo, for example, it's possible that the ruling elite adapted ideas about collectivization from Soviet contexts during a period of explicitly socialist rule. However, in practice, elites used vagrancy laws developed by French authorities in the final decade of colonialism to mobilize that labor, suggesting that colonial institutions at the very least paved the way for the continued use of forced labor, even if colonial practices are not the sole explanation for postindependence policies.

Forced Labor and African Economic Performance

This paper has shown that forced labor has formed a part of periods of economic growth in Africa throughout the nineteenth and twentieth centuries. In many places, labor coercion was integral to African political economy; and during periods of economic change, elites repurposed and redesigned institutions of labor mobilization to meet this new demand. During the nineteenth century, gradual moves to abolish the overseas slave trade were accompanied by increased internal demand for slaves to expand the production of minerals and cash crops. The increasing commercialization of cash-crop-producing regions led, in turn, to the increased use of slave labor to produce food and other goods for local consumption. While colonial rule was justified in part as an effort to abolish the practice of slavery, under-resourced colonial governments not only preserved existing practices but also developed new means—both direct and indirect—to mobilize labor through coercion. Many of the institutions developed under the colonial period persisted into postindependence period, as new phases of economic change put pressure on existing factor markets.

Coerced labor thus features prominently in the history of African economic growth during the nineteenth and twentieth centuries. But growth is not the only thing that shaped Africa's long-run economic performance. As in other regions, understanding why periods of shrinking are longer and more severe in some places is key to understanding long-run growth.⁸⁷ Theories of shrinking have yet to be developed as fully as theories of growth.⁸⁸ However, there are several factors that might either exacerbate or alleviate periods of shrinking. One is the degree of economic diversification; or put another way, structural transformation away from agriculture. African economies have generally experienced little structural change, and through the period under consideration here remained dependent on a narrow range of exports.⁸⁹ Changes in demand for these exports can generate more severe periods of shrinking than in economies with a more diverse range of industries. However, most theories of shrinking look beyond economic structure to consider the interaction of economic shocks and broader social issues. Unequal economies, for example, may be more prone to growth reversals.⁹⁰ Similarly, economies with limited human capital may be less able to adopt new technology that might improve productivity. States with less capacity also have less ability to intervene to mitigate the impact of shocks.⁹¹

There is more work yet to be done to link African institutions to the duration and severity of periods of shrinking. However, the overview of systems of labor mobilization presented in this paper suggests that while labor coercion may have facilitated growth in the short run, it might also have contributed to the causes of shrinking. Easy access to slaves and other forms of coerced labor may have helped limit the incentives of African elites to invest in more capital-intensive production of exports, so that high labor costs in Africa did not have the same impact as in the United States and the United Kingdom. Studies of the US South and Brazil, both economies heavily dependent on slave labor, and coercion in South Africa, have made a similar argument about the absence of investment in structural change. Equally, the widespread use of forced labor might also have contributed to increasing inequality and made elites less likely to support investments in human capital.⁹²

Many other factors—ranging from colonial policies to global shifts and the lack of state capacity in both the colonial and postindependence periods—also influenced the absence of structural change and low rates of human capital. A detailed examination of these factors is beyond the scope of this paper. What this paper aims to show, however, is that the wider role of forced labor in Africa's political economy shaped its long-run performance in diverse ways even after the end of the Atlantic slave trade. Given the constraints imposed by Africa's factor endowments, it is difficult to develop a plausible counterfactual by which the growth of the nineteenth and early twentieth centuries, in particular, could have occurred without the use of forced labor. However, its use may have made the hoped-for economic transitions of the postindependence period more difficult.