In the study of African politics, state formation and social inequality have often been treated as separate issues. In diagnosing a failure or weakness of African states, many authors blame selfish politics and the greed of the ruling elites [e.g. Collier Reference Collier2007; Bates Reference Bates2008; Englebert Reference Englebert2009], while inequality and poverty are treated as a given rather than as a political challenge. Approaches that move beyond the institutional surface to study the actual practices of the political authority of states “at work”—while revealing that African states are lively and powerful organisations within society—similarly treat social stratification as a side story or as a background aspect [Lund Reference Lund2006; Péclard and Hagmann 2011; Bierschenk and Olivier de Sardan 2014; Migdal Reference Migdal2001].
In this contribution, we put the political challenge of social inequality back at the centre of analytical attention for the study of state formation. Widespread poverty and what is perceived to be a lack of “development” has been a central theme of African politics ever since the capitalist transformation of the continent began and modern states in the nineteenth and twentieth centuries emerged. The stability of political authority seems to hinge on ruling elites’ ability to successfully manage, overcome or legitimatise inequality and poverty. However, neither is the perception of such “social questions” a given nor is a state’s reaction obvious. We claim, however, that such interpretations and reactions to the social question go a long way in explaining the trajectories of the widely diverging processes of state formation in Africa.
We want to demonstrate that trajectories of state formation [Bayart Reference Bayart1996; Lonsdale Reference Lonsdale1981] become more apparent when we employ a longer historical perspective. In the following, we want to show the connection between social inequality and state formation by comparatively analysing how and why rulers in British Africa detected and reacted to “social questions” between c. 1880 and 1950. The guiding empirical questions are: Which aspects of social change did ruling elites perceive to be a destabilising or morally unacceptable social question? What did state institutions do about these social questions? Which consequences had these policies for the formation of states? By answering these questions, we seek a new understanding of the role of social inequality and social conflict in the history of state formation.
Our method of interpretive historical comparison seeks to understand differences and similarities as well as convergences and divergences [Kaelble Reference Kaelble2012]. A comparison of three countries in the early stages of their modern state formation, Egypt, South Africa and Uganda, serves as the basis of the investigation. This “Cape to Cairo” sample brings together political entities under British rule or influence. We will show that despite remarkable differences due to different global historical timings, all three cases have, in fact, some things in common: Forms and vicissitudes of world market integration mattered for the rise of urban “dangerous classes,” and the reactions of British colonial rulers were not free of contradictions. Furthermore, we see in all three cases that state formation through the social question was a highly internationalised process in Africa as it was elsewhere.
The selection of cases is also intended to provide a range of varieties and similarities regarding economic and political histories. While the colonial encounter was tumultuous in all cases, in Egypt, the British protectorate (1882-1922) was enforced amidst the social dislocation and discontent that a previous state-led modernisation project had created. The semi-independent Union of South Africa (1910-1961), a cluster of previous settler colonies, was formed after a century of regional warfare, and in the wake of a gold and diamond-driven economic boom. Meanwhile, in the protectorate Uganda (1894-1962), an overwhelmingly agricultural entity, the British connected radically different pre-colonial societal models ranging from centralised kingdoms with highly unequal internal political and economic structures to segmented acephalous language communities, forming a new political entity.
This paper argues centrally that the expansion of capitalism and state-building practices in specific countries and colonies, often further propelled by major global political developments, resulted in unplanned and undesired social dynamics. These latter developments gave rise to what we call “social questions”: the impression of ruling elites that the state needed to manage these undesired dynamics lest social and moral orders be undermined. Stately management efforts, while often not achieving their immediate aim of mitigating social change, usually achieved the growth of state bureaucracies as well as the increasing stateness of societies. This is what we call state formation.
With this article, we aim to contribute to a rather new strand of welfare state scholarship that seeks to delineate the specificities of African social policies [e.g. Mkandawire Reference Mkandawire2009; Kpessa and Béland Reference Kpessa and Béland2013]. Different from attempts to conceptualise types of welfare statism, as Bevan [Reference Bevan, Gough and Wood2004] and Cerami [Reference Cerami2013] made, there has recently been a shift toward focusing in more detail on the effects of colonialism. In this vein, however, scholars have limited themselves to comparisons of welfare systems in former French and British colonies [Kangas Reference Kangas2012; Schmitt Reference Schmitt2015]. In this article, we show that important political differences existed within empires. This article is a contribution to filling that gap. As we will argue, our findings speak to the research on state formation and the historical institutionalist assessments of colonialism.
We will proceed by providing a short summary of the historical location of the “social question.” In the following three sketches, we will use mostly secondary literature and only a few primary sources to reconstruct the main linkages of social differentiation, policy reactions and state formation. In the conclusion, we provide a summary of our argument.
The social question and state formation: Europe and Africa compared
The concept of “the social question” originally figured in French and German political discourse in the nineteenth century [Tönnies Reference Tönnies1907]. Different meanings are attached to this expression, ranging from moral indignation about poverty and the poor to the fear of “dangerous classes” [Fregier Reference Fregier1839] and the “spectre of communism” [Marx/Engels 1848]. While in the period stretching into the early twentieth century it had a “structuring presence” for political discourse all over Europe [Case Reference Case2015: 3], it was not always clear what the expression actually encompassed [Howerth Reference Howerth1906: 254]. In retrospect, it connotes “a concern about a society’s ability to maintain its own cohesion” [Castel Reference Castel2003: 3]. This worry has had many historical objects. The European pre-industrial embodiment of the social question was the “vagabond” [Castel Reference Castel2003: 5], but countless other instances of social phenomena or stereotyped groups were, over time, seen as “reasons” to worry. Ruling strata perceived all of them as requiring some form of state-led intervention, aiming to restore a sound societal order. Social relief and regulation, as well as the physical suppression of dissent, remained the typical twin answers of ruling strata to the social question [Lüdtke Reference Lüdtke1982].
In the German Reich, since 1871, “the social question” was an expression that combined the conservative fear of the workers’ movement with charitable care about the evident expressions of poverty. The reaction of Bismarck’s government was a mixture of repression and care: the “socialist laws” that prohibited any activity of the social democrat party were combined with the introduction of compulsory social insurances [Wehler Reference Wehler1995: 902-915], nowadays considered to be the birth of the German welfare state. Since then, social security has been one of the fastest-growing arenas of the bureaucratic state [Scherpner Reference Scherpner1974: 171-177]. The formation of European states, both regarding their capacities and scope, has thus been closely related to various instances when “the social question” was posed.
Social policy creates new forms of domination. Once states intervene in the provision of healthcare, housing, pensions, education, labour regulation and financial assistance, a new and deepened relation of domination (Herrschaftsverhältnis) is developed. The encroachment of the state into these fields brings forth its own particular dynamics of legitimisation and social power formation (and deformation). Social policy directly connects the interests, values and expectations of bounded populations to the organisations ruling over and through them, and thereby constitutes a particular political community. Although social policy today is a normal political field in oecd countries, this was not the case in the nineteenth century, when modern-state infrastructures were yet to be established as the dominant providers of resources, benefits and regulations. Thus, the emergence of social questions may be expected to result in the deepened efforts of states to establish direct modes of regulating social affairs and therefore state formation.
In this respect, no reason exists to believe in fundamental differences between Europe and Africa over the past 120 years. In Africa, many traits of capitalist modernisation could be observed since the early colonial period, i.e. the emergence of wage labour, the crisis of older social institutions, or mass impoverishment and the growth of urban slums. Ruling groups, especially in the context of colonisation and empires, may be expected to have interpreted such social inequality and unrest in largely similar terms as their colleagues in the metropoles, and to have attempted to smooth tensions so as to avoid or quell rebellions in similar ways.
However, differences exist not only between European and African contexts but also inside each continent. Such variances may also lead to divergent interpretations of and strategies for dealing with social dislocations and accordingly different trajectories in terms of state formation. “Polizey” [Heidenheimer Reference Heidenheimer1986], i.e. the state bureaucratic capacity to uphold and repair social and political order through welfare and repression, may take very different forms.
In Europe, the original source of state formation emerged from the outside. As Tilly [Reference Tilly, Evans, Rueschemeyer and Skocpol1985] argues, for early modern Europe, stately protection developed originally as a “racket,” in which populations were compelled to provide material and human resources to assist “their” states in defending themselves against external military threats. In the long run, the ensuing statisation of society [Verstaatlichung der Gesellschaft] through contentious politics also results in a societalisation of the state [Vergesellschaftung des Staats]. Our approach, we argue, shows an additional argument, namely that internal social differentiation can spur state formation as well.
On the African continent, interstate military threats seemingly have not triggered state formation processes for a long time. This was, first, because large parts were only loosely populated so that political centralisation did not take place [Herbst Reference Herbst2014], with some areas, such as the Great Lakes region, the Ethiopian highlands and the Nile valley, being exceptions [Iliffe Reference Iliffe1995: 70]. Colonialism then did away with existential military threats. All the more so, we argue that the question of domestic social and political conflict seems pivotal in the analysis of state formation in Africa. However, besides the country-specific studies on which this paper is largely built, and the literature dealing with the post-independence period, comparative historical analyses on social inequality and social conflict in Africa before 1960 are scarce. Thirty years after publication, a central reference is still John Iliffe’s “The African Poor” [1987], which describes the structural changes of poverty south of the Sahara from the thirteenth century onwards, especially in the colonial era. “Traditionally,” rural sections of society, such as the ill and disabled, slaves and widows, were the ones facing absolute poverty. During colonialism, the crisis of rural economies and urbanisation, however, led to the emergence of the underemployed strata of “able-bodied” men and women who found it difficult to make a decent living and to reliably provide for their dependents. The decline of famine, historically concerning all strata of societies, and the spread of structural malnutrition for some societal segments, is the strongest indicator of these developments. Iliffe does not explicitly analyse colonial social policy. Yet his descriptions of the authorities’ (wavering) tackling of slavery, the (often too late and too little) provision of famine relief, and, most significantly, the segregation of towns and cities between poor and rich (and Black and White) speak to the assumption of a colonial awareness of an African social question.
For the late colonial period, Frederick Cooper [Reference Cooper1996] provides a perspective on the reactions of colonial administrations to the spread of worker strikes, the return of a large number of discharged soldiers after the World Wars, and the ambitions of indigenous evoluées elites in British and French Africa. Administrations slowly gave up on their previous intentions of slowing down capitalist modernisation and keeping Africans confined in rural and tribal social settings. They developed “modern” labour relations by allowing for unionisation, and they sought to accommodate African aspirations by introducing development agendas and minuscule kernels of welfare provision. The inadequacy of these slow policy corrections in the face of massive social change, further hampered by administrative in-fighting, and nationalist elites’ appropriation of the modernisation debate, however, undermined these efforts.
The social question in British Egypt (1882-1922)
The British-Egyptian imperial-colonial project lasted from 1882 to 1954. Having occupied Egypt in 1882, the British instituted a ruling system that, at least formally, left Ottoman sovereignty over Egypt intact. Only with the outbreak of World War I did the British formally incorporate Egypt as a protectorate that lasted until 1922. Even though the British continued to dominate Egyptian affairs until the negotiated retreat of their troops in 1954, this section focuses mainly on the period until 1922 when at least formal independence was granted to Egypt whose everyday affairs were governed by national elites.
Compared to other contexts in Africa, this was a rather short-lived imperial-colonial project. It nevertheless had major implications, not only for Egyptian twentieth century developments but also for the wider Middle East. Footnote 1 For Egypt, in particular, the colonial experience included the restoration and empowerment of the landholding elites, the consolidation of monarchical rule, and an eclectic set of liberal reformist policies. These were being implemented in quite rebellious contexts, as one rebellion (the ‘Urabi uprising of 1881-1882) created the context and opportunity for Britain’s entry, and another one (1919) triggered the end of British colonial order. This demonstrates that the British could not sufficiently legitimise their hold over Egyptian elites and the wider populace. They also did not have anything to offer—in materialist terms, something proper, as the consul-general Lord Cromer envisioned—to curb the growing anti-imperialist spirit that became an important foundation of modern Egyptian nationalism [Khuri-Makdisi Reference Khuri-Makdisi2010]. This, in fact, was the social question of the time, of which the British were becoming increasingly aware.
State formation before the British
British rule began at a highly conflictive conjecture. It brought to an end earlier developmental projects that were intended to strengthen Egypt’s factual autonomy from the Ottoman Empire, on the one hand, and the British and French empires, on the other. Muhammad Ali’s reign until 1848 promoted new forms of state ordering, basing his rule on a new conscript army, the first attempts of primitive industrialisation and import substitution, the expansion of cotton plantations and the development of new cotton seeds and corvée labor mobilisation for public works, especially for the expansion of new perennial irrigation schemes and railway lines.
This statist-authoritarian mobilisation came with serious dislocations of traditional village communities, as the gradual commodification and privatisation of rural property [Baer Reference Baer1969] triggered the first waves of land ownership concentration and urbanisation. Land flight was increasingly criminalised, even by the use of the armed forces, on the one hand, and continual peasant rebellions, on the other [Mitchell Reference Mitchell1988: 42-43]. The latter occurred in the 1820s, Footnote 2 in particular, and were often framed in religious-messianic tones. They were directed against high rents and taxation levels, the increasing shift to tax collection in cash, and conscription into the new army [Baer Reference Baer1969]. New regulations, such as the Law of the Peasantry (1829-1830), demonstrate the necessity of maintaining public order in the countryside, with repressing dissent being the sole mechanism through which the state apparatus aimed to control the mobility of its populace. This was, so to speak, the pre-British Egyptian social question: the threat produced by these early modernisation attempts, which called for new public regulations to “manage” [Ener Reference Ener2003] those newly dislocated groups that had fled their villages and military conscription. They found in the urban centres an increasingly differentiated system of faith-based private welfare and charity [Singer Reference Singer2008], which provided soup kitchens, poor shelters, hospitals and schools, often tied to mosque complexes. Starting in the 1840s, the state as such entered this field of relief provisioning by taking over established institutions to clear, or at least try to clear, the streets of beggars and to organise urban public life under the increasing control of the gendarmerie (dabtiyya), which had to confront issues of crime, health, prostitution and deviance [Fahmy Reference Fahmy1999]. However, this developmentalist model led to a dramatic increase in public debt, opposition and, finally, a military uprising against the Khedive under Colonel ‘Urabi [Cole Reference Cole1993; Toledano 1990].
Cotton and the rural sphere
British troops ended the ‘Urabi uprising of 1882, restored the authority of the returned Khedive Tawfiq (r. 1879-1892) and very early on formulated the purpose of the occupation and the timely expectation of the British evacuation of Egypt. As outlined by the ruling Liberal Party, British control was to be only temporary; it was not to alter the status quo of Egypt as part of the Ottoman Empire. Its main goal was to restore Egyptian solvency vis-à-vis the multilateral Caisse de la Dette and to introduce administrative reforms to maintain security and calm in Egypt. For Lord Dufferin, who issued the first report on the potential of reform in 1882, the prime mechanism for achieving these goals was to uplift Egypt’s agrarian potential, increase hydraulic investments, and rationalise cotton plantations, in particular. Neither he nor Sir Evelyn Baring, the first British Consul-General of Egypt (Lord Cromer since 1892), remained silent on the notion of urban and rural poverty, but they would not formulate any policy intended to tackle social changes in the countryside. The only measures that were instituted included the suspension of the corporal punishment of peasants (courbaj) and, although only gradually before 1889, of corvée labor mobilisation for public works. Also, to lift both the general welfare and the spirits of peasants, the British used the minimal financial leverage they had, until 1895, to lower the overall tax rates on land by setting maximum limits for the amount to be collected as land tax [Tignor Reference Tignor1966: 109]. In addition, many peasants benefitted from tax reductions that were introduced to release them from their high levels of indebtedness.
The continuous rise in agricultural output and exports over the next three decades, which was made possible through public investments in irrigation, increases in technological expertise and the concentration of land ownership, resulted in the gradual dissolution of traditional forms of solidarity, which the British still deemed quite essential for the guarantee of stability and calm. Footnote 3 One measure was intended to limit the dynamic of peasant expropriation and large landholdings: the Five Feddan Law of 1911 was “designed to protect and win the favor of small landholders in Egypt, by removing one of their greateast sources of discontent” [Tignor Reference Tignor1966: 239]. According to this scheme, peasants owning five feddan of land or less were to be protected from being evicted after they had already become indebted to local money lenders. However, given the enormous degree of authority of the local agrarian elites, this law was virtually unenforceable, and the dynamic of accumulation continued in the countryside.
Urban activism and nationalism
While no new peasant rebellions broke out until 1919, most anti-British activism had, since the 1890s, moved into the cities as the new, main sites of contention. The colonial government was faced with a combination of new nationalist groups and the first waves of strikes. While the latter began among anarchistic Greek and Italian workers (cigarette rollers, port workers), the ideas and practices of labour resistance quickly diffused among Egyptian workers, whose protests and strikes (esp. by tram operators in Cairo and Alexandria) in the 1900s were repressed [Lockman Reference Lockman1988]. However, these collective protest actions laid bare a serious legitimacy problem of the British, as they allowed the first group of nationalist actors (Nationalist Party 1907) to link socioeconomic with national and identity issues. It was this culmination of social and political forces that induced the British to issue the first Workers’ Protection Law by limiting the types of job activities open to women and children in 1909. Even though this law was, as the Five Feddan Law, unenforceable, it set the stage for the next wave of labour protection laws in the 1930s, when Egypt was already formally independent (1922).
The British rulers of Egypt did perceive the existing social tensions as potentially threatening their rule, but they felt quite secure that re-formed and continued agrarian growth and the deepening of linkages to European capital would safeguard calm in the countryside and in the cities. However, social transformations stemming from this same growth dynamic triggered the first forms of activism within a segmented urban labour class, and the formation of new public spheres and the increasingly politicised expression of ethnic identity and nationalist claims. Only after heavy state repression, which itself strengthened nationalist discourses, failed to rein in labour activism would the first social policy regulations be formulated; however, they were virtually non-enforceable. World War I induced grievances, the crackdown on political parties and activists via the martial law framework the widely perceived British non-responsiveness to Egyptian nationalism and culminated in the 1919 revolution, to which the empire finally reacted by granting formal independence.
Dealing with the threat of the urban poor was considered vital for domestic politics. However, the consul-generals ruling Egypt were able to act leniently in the face of the widespread poverty that was becoming a potentially explosive issue. This was possible because, unlike the political stirrings in the Ottoman Empire, which triggered gradual political reforms until 1876, the institution of citizenship was not wrested from the Khedives since the rule of Muhammad Ali. Neither rich nor middle or poor classes had the right to vote as had been the case with the transformations of English voting in the Reform Acts of 1832 and 1867.
Finally, no mechanism existed between militarisation and social policy formation. The Sudan missions in the 1880s and 1890s, in which the British very much relied upon Egyptian soldiery, did not come with any major expansion of social provisioning to them [Tignor Reference Tignor1966]. In contrast to India, in particular, the British did not draw on the Egyptian army for its military expeditions against the Ottomans in World War I. Being considered more loyal to their Ottoman-Muslim brethren and thereby regarded unreliable for warfare, Egyptians were drafted only to form labour battalions to set up military infrastructures at the British-Ottoman fronts [Rogan Reference Rogan2015; Goldberg Reference Goldberg1992].
Thus, the Egyptian case stands for the obvious neglect of local political claims and the British resorting mainly to repressive measures in answering the social question. This type of imperial politics left heavy burdens for the Egyptian “liberal order” between 1922 and 1952. While agrarian elites and the king could, to a certain extent, institute Egyptian self-rule, their legitimacy was challenged by their visible lack of capacity and will to deliver public policies for lower strata. This ended when the “Free Officers” intervened militarily and suspended both parliamentarism and the monarchy. They triggered the first Arab Socialist Revolution in 1952 and finally delivered “modern” social policies as the core of their legitimising mission.
The social question in the Union of South Africa (1910-1948)
Before 1994, the South African state interpreted the social question as a race question. Over the course of the twentieth century, the elite’s angst at losing political and economic hegemony led to two connected strategies. On the one hand, violent repression was the White elites’ chief means of coming to terms with the problem of ruling as a racially defined minority. On the other hand, the fragile legitimacy of a racially differentiated economic and political order was bolstered with the introduction of (in the African context) exceptionally large welfare provisions. These welfare policies were designed to reproduce racial inequality but also to build and maintain the loyalty of poor and working classes, and to maintain the economic sustainability of rural areas. For the young South African state, welfare policies were a major means of controlling and regulating the diverse population, and of overcoming the power of provincial administrations.
Two important historical periods of dealing with the social question and the related dynamics of state formation before apartheid (introduced by the National Party from 1948 onwards) can be discerned, roughly speaking before and after World War II. The historical background of the welfare reforms in the 1920s included the violent process of territorial unification domination in the nineteenth century and the emergence of an urban frontier society after the discovery of gold and diamonds. Initially, mostly the White and Coloured populations Footnote 4 benefited from social policies while, for Africans, repressive regulations were reserved. During the 1930s, increasingly well-organised social and political protests by the non-White population increased pressures on elites. Yet only when South Africa’s active participation in World War II created a labour shortage and eased economic depression was a liberal-minded bureaucracy able to expand the benefits of the welfare state, albeit with strong restrictions, also to Africans and Indians.
Guns and gold: South Africa’s frontier society in the nineteenth century
The emergence of a specifically South African social question is closely connected to the discovery of very large deposits of diamonds in the 1860s and gold during the 1880s. The urge to participate in these riches led to the foundation of large urban conglomerates with Johannesburg at the centre. An unruly White and African mining proletariat, beer brewers, cooks, cab drivers, domestics, washers, prostitutes, gangsters, tradesmen and impoverished day labourers populated these towns and cities—leading to a classic colonial frontier society in the midst of a gold rush [Van Onselen 2001].
While urbanisation and proletarianisation have been decisive social dynamics, wars also reverberated in the discourse on social policy, as did the economic downturns connected to them. Armed conflict and warfare dominated South Africa’s political history in the nineteenth century, from Britain’s seizure of the Cape Colony from the Dutch in 1795, to the ultimate victory of the empire over the independent Boer republics in 1902. Parts of this conflict history also included the many wars against African ethnic groups. Settlers’ expropriation of land partly caused and regularly followed these conflicts. The introduction of hut and poll taxes, which demanded the first forms of bureaucratic control, further diminished rural means of support. This led to a long-lasting rural economic crisis in the African reserves, which also contributed to rapid urbanisation at the turn of the century. After the British had achieved violent unification, they allowed the foundation of the Union of South Africa in 1910 as an autonomous dominion within the British Commonwealth. The new state was organised democratically but excluded a majority, i.e. most non-Whites, from the vote [Terreblanche Reference Terreblanche2002: 179-232].
Provide and divide: The introduction of racially exclusive welfare
Both in monetary terms and in the number of recipients, the centrepiece of the South African welfare state in the twentieth century was the old-age social pension system. This section will focus on the emergence and trajectory of this South African form of welfare provision, whose development provides insights into many aspects of the social question in the country. The first group considered to be in need of help was the “poor White” strata of society, which had fallen victim to the crisis in rural farming after the Boer Wars. In the mining areas, economic depression following World War I also hit the livelihoods of non-educated White families. About 10 % of the White population was considered poor [Sagner Reference Sagner2000: 525-526; Devereux Reference Devereux2007: 541]. The discourse on social risks and the necessity of the pension system, which had already begun about ten years prior, was dominated by two moral arguments. First, that many poor old-agers “deserved” pensions because their poverty was not the result of individual failure but rather depression and warfare [Sagner Reference Sagner2000: 528]. Second, White politicians warned against the mixing of Black and White poor populations in urban slums. “Civilised races” needed to be socially as well as geographically separated from the “swartgevaar” emanating from the Black population [Seekings Reference Seekings2007: 378-382; Iliffe Reference Iliffe1987: 116-121].
Pressure from below also put the social question on the table. White mine workers, in 1922, organised the “Rand rebellion” to prevent a breach of the colour bar as well as a decrease in wages. The government declared martial law and militarily quelled the uprising. Although mine owners and the state may have feared strikes by underprivileged Black workers and day labourers (which occasionally took place), collective action by their White “colleagues” for the maintenance of racial privileges remained politically more decisive [Frank Reference Frank1974: 140]. Riding public discontent concerning the brutal quelling of the Rand Rebellion, the National Party and the Labour Party won the following general election. The Afrikaner-dominated National Party considered rural, impoverished Afrikaans speakers (and those in danger of social descent) to be a major constituency. The Labour Party, on the other hand, catered to the White urban working class. In 1928, the coalition government introduced the old-age pension system, from which mainly their constituencies benefited. The introduction of tax-financed pensions revolutionised the South African welfare system [Seekings Reference Seekings2007: 393]. The very limited social welfare services that existed prior to this had been the domain of churches and provincial governments, which had received funds from the central government. The pension system, in contrast, was administrated and budgeted by the central government, which thereby drastically increased its importance and institutional strength against church resistance [Iliffe Reference Iliffe1987: 121; Seekings Reference Seekings2008]. Pensions were not a right but rather were accorded to “deserving” individuals based on an extensive means test. The young centralised state established its bureaucratic presence and power among the poor Whites [Sagner Reference Sagner2000: 525-526].
One argument in the discourse on the pension system was its role in the reactivation of social boundaries between poor urban Whites and non-Whites. A number of laws in the same period, among them the Natives Land Act (1913) and the Natives Urban Areas Act of (1923), established bureaucratic control over the African population, fostered its social and geographic segregation, and protected White farmers and workers from their economic competition. In this period, Africans and Indians were accordingly totally excluded from the pension system, while needy Coloured elderly, a considerable voting bloc in the Cape Province, received reduced pension levels. For the African population, only occasional emergency aid was made available. White elites argued against welfare services in a typical colonialist vain. Social assistance would supposedly undermine the “communal spirit” and social structures in the rural areas. Urban workers, the argument went, should and could return to their rural places of origin once they needed assistance. From the capital perspective, social assistance could potentially increase labour costs and even lead to a shortage of workers. In any case, it was argued, White taxpayers would need to finance social services for the non-White population [Seekings Reference Seekings2007; Terreblanche Reference Terreblanche2002: 239-275; Duncan Reference Duncan1995].
Inclusion of non-Whites during and after World War II
However, at least liberal White elites were aware of the increasing potential of social and industrial conflicts with the Black urban population, and accordingly, they pressed for a more equitable social system. In 1944, and against the resistance of the radical-right political camp that would win elections four years thereafter, the pension system was extended beyond the White and Coloured populations [Sagner Reference Sagner2000: 530-533; Devereux Reference Devereux2007: 541-542].
How did this governmental change of mind come about? Footnote 5 Sagner [2000: 33-35] argues that a liberal-minded ministerial bureaucracy made use of an opportune moment to implement long-desired reforms. While the segregationist policies of previous decades may have been modelled on the United States example [Terreblanche Reference Terreblanche2002: 240], also contemporary welfare reforms and policies in Great Britain and the US aggravated the reformist spirits [Duncan Reference Duncan1993: 107; Sagner Reference Sagner2000: 537; Devereux Reference Devereux2007: 542-543]. Economically, the end of global depression and the war-related boom in manufacturing allowed for the expansion of state expenditure. In the context of World War II, White labour was greatly reduced in availability, which increased fears of social unrest. The mining industry, which in 1928 had argued against pensions for Africans, now welcomed their introduction. Part of the calculation may have been the de facto subsidisation of rural areas, which allowed for a further reduction in African migrant labour wages [Sagner Reference Sagner2000: 537-538].
Yet domestic pressures also put the “native” social question on the governmental agenda. Public protest against racial and social discrimination was rampant in South Africa in the 1930s and 1940s, as in other parts of southern Africa [Cooper Reference Cooper1996]. The African National Congress (ANC) and Black trade unions grew in strength and claimed, besides civil rights, social rights for all South Africans [Duncan Reference Duncan1993: 106; Van Niekerk 2003: 363-364]. At the same time, a social crisis in rural areas and the dissolution of traditional social bonds became ever more obvious. In the cities, meanwhile, especially juvenile delinquency and violent youth gangs became intensely debated topics [Iliffe Reference Iliffe1987: 137-138]. The White conservative-liberal government argued that inactivity would increase social and political costs in the long run, and a kind of trusteeship responsibility existed for the welfare of the African population. However, the introduction of social assistance did not increase state legitimacy, as can be seen from the continuation of protests against racial discrimination [Sagner Reference Sagner2000: 535], although that failure has little explanatory power regarding initial rationales.
Nonetheless, Africans and Indians received drastically decreased rates allocated according to racial belonging [Sagner Reference Sagner2000: 524; 533-534; Seekings Reference Seekings2004: 300]. “Moreover, the native population was subdivided into three categories: city residents, town residents, and rural residents” [Devereux Reference Devereux2007: 543]. Black city dwellers received the highest rates, and rural residents found themselves at the lowest end of the pension scale. This complex system was thus a grudging acceptance of the fact that a part of the Black population, after retiring from urban jobs, would not and could not “return” to rural areas and their often imaginary systems of traditional welfare. At the same time, the system was still geared towards “de-urbanising” as many Black people as possible. In 1965, the system of different rates depending on residence was abolished, as policymakers at that time (contrary to their predecessors) considered it to provide incentives for urban migration [Devereux Reference Devereux2007: 543-544].
The electoral victory of the National Party and the official introduction of Apartheid in 1948 may be understood as the greatest rupture in South African history. Regarding the welfare system, and especially the pension scheme, it surprisingly did not lead to a turnaround. While the National Party had strongly opposed the expansion of the pension system to the African population, it did not abolish Black pensions during its decades-long reign [Sagner Reference Sagner2000: 536]. Quite the opposite, the number of recipients as well as the level of payments almost continuously increased. Initially, the National Party tried to decrease the number of Black recipients through a stricter application of eligibility criteria. At the same time, pension levels for the African population rose by 59 %, while gaps between pension-rate levels for different “race” groups became even greater. Whites received at least four times higher pensions—and during some periods, even twelve times higher pensions—than the most disadvantaged African pensioners [Devereux: 2007: 545-546].
Already before the advent of apartheid, the social question in South Africa was interpreted as deeply connected to the country’s racial politics. Welfare policies introduced in the 1920s completed a series of laws from the same period that aimed to politically, socially and geographically separate the population along racial lines, and install a hierarchy between Whites, Coloureds, Asians and Blacks. At the same time, social provisions cushioned the most severe consequences of these violently imposed measures. Discussions about social questions, among political and economic elites, were triggered by a range of factors: social mobilisations of workers and party constituencies, but also the observation of poverty in rural areas and the mixing of Black and White populations in rapidly growing cities. Welfare provisions were thus designed to reproduce social-racial hierarchies, maintain social distance between poor Whites and non-White sections of the population, but also to ease social misery in rural areas. While in the 1920s, the latter aspect targeted mainly White farmers, after World War II the pension system became an important form of income in impoverished Black reservations (“homelands”). Perhaps as an unintended consequence, the South African welfare system also greatly strengthened the central state vis-à-vis the provinces and the Christian churches, and provided for the bureaucratic incorporation of large sections of the poor population.
The social question in colonial Uganda (1894-1950)
Soon after the beginnings of the British Uganda Protectorate in 1894, the de facto colonial government noted that social mobilisation was increasing. Ironically, until Uganda gained independence in 1962, colonial policies were both causes of and supposed solutions to the change that was observed. In the protectorate, a highly diverse society was enshrined: on the one hand, a number of centralised kingdoms in southern and central Uganda with semi-feudal divisions of labour, and on the other hand, particularly in Northern Uganda, a number of rather segmented language communities, which were much less and only later integrated into the emerging colonial capitalism. These differences also marked the three main phases of the development of the social question in colonial Uganda: a first phase of colonial economic transformation, a second phase marked by the developmental state of late colonialism and, finally, the overtaking of the colonial state by a petty bourgeoisie at the time when the protectorate gained independence.
Colonial mise en valeur and social differentiation
From the early years of the Ugandan protectorate, two main policies had an effect on social dynamics. The first was the land tenure policy, and the second was the introduction of direct taxation in the form of a hut tax, already stipulated in the Buganda Agreement of 1900. It subordinated Buganda, the largest and centrally located kingdom, to British “protection.” Here, the “mailo” system of land tenure did not simply institutionalise de facto feudalism. By introducing land titles on square miles (“mailo”) for single clans, customary land tenure was abolished. This measure created a Bugandan landed gentry that enjoyed the advantages of stable colonial rule. It became the most important local carrier group of the colonial system [Mamdani Reference Mamdani1976: 42; Thompson Reference Thompson2003: 44]. This system, however, excluded many others from the possession of land and led to the first riots [Bazaara Reference Bazaara1994: 29]. In its pursuit of an increased productivity for the protectorate, the colonial administration became aware of this imbalance in the 1920s and tried to encourage freehold land tenure, which turned land into a commodity [Doyle Reference Doyle2006: 166-170].
Taxation policies were applied to fund an increasingly costly colonial state. However, they functioned as well as a capitalist mise en valeur [Sarraut Reference Sarraut1923], a valorisation of land and labour that had not been integrated into a monetarised world market. Taxation was intended to serve as leverage in forcing the population into active roles in the colonial capitalist economy. The “native farmer,” an active cash-crop producer, not yet understood as an entrepreneurial self, was the imagined ideal subject of this first phase of colonial rule in Uganda. With forced labour being the alternative to paying taxes in cash, the colonised who could not enter cash-crop production soon began to take up all forms of wage labour that existed during the first years of the protectorate; this also occurred in the North [Middleton Reference Middleton and Turner1971]. According to Fuller [1977: 78], wage labour increased rapidly at the beginning of the twentieth century for two reasons: first, the poll tax ordinances from 1905 and 1909 enforced monetarisation and, second, the acculturation of European consumption practices fostered the demand for cash. Unemployed labour soon emerged. Already in 1905, unregulated settlements of about 20,000 unemployed in Entebbe were reported [Fuller Reference Fuller1977: 84].
A “social question” in the sense outlined in the introduction has thus existed in the protectorate at least since the 1920s. From the colonial government’s perspective, a major concern was to balance the need to produce a sufficient labour force with the perceived danger of triggering undesired social dynamics. The latter were perceived as a destabilisation of the precarious social contract, mostly with indigenous cash-crop producers. However, disturbing social phenomena soon became a major concern of colonial rule. In 1926, as Mamdani [1976: 122] notes, a clear understanding of the tensions built into the colonial project existed. One therefore would date the birth of the “social question” in Uganda to this decade, also because it was in this period that the colonial state reacted with active measures to streamline the social dynamics of the protectorate.
The answer to undesired dynamics, which the colonial administration noticed, was “civilization”: With similar if not identical undertones to the church officials’ discourse in nineteenth century Europe, missionaries in the protectorate soon complained about the effects of monetarisation, such as the loss of “respect towards the Europeans” [Fuller Reference Fuller1977: 95]. This would be addressed by the formation of a new type of colonial subject similar to the French évolué model: the educated man who would aptly fit into the colonial administration. Mission schools delivered primary education to this end. The other measure for improving the quality of the colonial labour force was the building up of secondary education, embodied also in the transformation of Makerere College into a full university of British East Africa in 1934.
Huge differences in labour organisation emerged. Internal labour migration, in particular to the urban centre Kampala, was an outcome of an increasingly market-oriented agriculture, in the early years mainly of cotton, and later of coffee and tea [Middleton Reference Middleton and Turner1971]. While all of these products are labour-intensive, for a long time they did not lead to an organised labour force in the agricultural sector. Traders and cash-crop producers organised themselves, but the unions that emerged in Uganda from the late 1930s onwards arose out of the urban economy and the milieus of colonial state employees.
Late colonial rule and the idea of development
The last thirty years of the Uganda Protectorate witnessed the transformation of parts of the landed gentry into a more differentiated labour force, working mostly for the colonial state [Mamdani Reference Mamdani1976: 127]. The “benevolent paternalism” of the colonial administration aimed at the “preservation” of a peasant economy, an ideal that came to contradict very diversified forms of land tenure as well as market access and licenses for running businesses [Ehrlich Reference Ehrlich1963: 275].
Historians agree on the point that, at least before 1940, British colonial administration was not actively seeking to transform African societies into capitalist economies. The idea was rather to preserve an imagined “traditional society” against the dissolution and decay that were thought to come with market mechanisms and private capital [Butler Reference Butler1999: 34]. In a famous article, Cyril Ehrlich [Reference Ehrlich1963] denounced this paternalist attitude as contradictory, as it led to policies that hindered Africans from doing what the colonial authorities also expected them to do: become entrepreneurial subjects.
With the beginning of the World War II and the question of reconstruction at its end, an intense debate arose about how colonies should be governed, addressing also observable social dynamics [Lee Reference Lee1967]. The effects of war planning, the rise of social democrat thinking throughout Europe, and growing pressure from the US on the colonial endeavour led to the adoption of the idea of “development” in the late phase of colonial rule in British Africa and, as a part of it, in Uganda [Butler Reference Butler1999: 31]. In the debate on how best to pave the way forward, market-liberal positions stood against strong state interventionism, as it was embodied in the Fabian Colonial Bureau, opened in 1940. This department, which the Fabian Society created in 1940, seems to have played a pivotal role in the emergence of “social policy” in the British Empire. Although its publications and reports referred to the situation in the colonies, its spirit and probably also its remedies were born out of European reformism [Goldsworthy Reference Goldsworthy1971: 113-123].
The ensuing debate in the circles of British colonial institutions revolved around the right balance between state engagement and private initiative, and it involved the shortage of capital and the peaceful transition of “developing” societies into stable and sufficiently institutionalised polities. The debate had the entire empire as a historical background, and it had a much greater effect on policies in Uganda than did the local constellation and dynamic. This early debate about “development” featured a number of characteristics that would accompany “development policy” as a way of addressing “the social question” today. First, debates and policies were based on reviews and reports that travelling specialists with international experience produced. Second, the birth of “development” was from its beginning bolstered with academic expertise: American foundations funded big research schemes to assess the social and economic situations in British colonies [Lee Reference Lee1967: 21; 81]. Third, the discursive field consisted of an endless row of arenas in all parts of the empire, with no centre and a number of different public institutions producing policy papers and suggestions.
In the Ugandan arena, social conflict took on organised forms rather slowly. With the formation of the “Buganda African Motor Drivers Union” in 1939, a first union emerged and, from the end of colonial rule, strikes and unrest—in particular, against the taxation regime—continued with an irregular frequency. Social differentiation also led to conflicts within the “Bugandan community,” quite in contradiction to the popular belief that “tribes” formed units of social integration [Mamdani Reference Mamdani1976].
The colonial state reacted to general strikes in 1947 and 1949 with a “Police Ordinance” in 1947 and a “Trade Unions Ordinance” in 1952. General unions were declared illegal [Mamdani Reference Mamdani1976: 191]. Nevertheless, civil servants organised, as did railway workers, and the 1950s also saw a spread of political parties. An important background to this wave of political organisation was the return of more than 60,000 war veterans who used their experiences and gratuities to develop independent ideas. As in other parts of the British Empire, the “social question” would soon turn into the question at the end of colonial rule with “Africanisation” as the preparatory phase.
“Tribalism” and policing the social question
Social problems persisted and became an object of policing. For the police leadership in the colonial state, they were seen through the lenses of “tribalism.” Institutionally separate policing developed only in late colonialism; in the 1930s, the police force comprised about 20 officers [Harwich Reference Harwich1961: 3]. Before that, the King’s African Rifles, the British colonial army in East Africa, was responsible for political policing, such as quelling rebellions. Still, in 1959, when the Bugandan political party organised a trade boycott of mostly Asian traders, the inspector general of the police at the time saw it as “lawlessness with xenophobic undertones” [Macoun Reference Macoun1996: 39]. Rebellions against an increase in graduated tax were seen as typical for “a country already beset by tribal and fractional strains” [Macoun Reference Macoun1996: 44]. It was determined that colonial policing should be a clear answer to the “increase in civil disorder, violent strikes and succession of minor revolts followed by major revolt” [Macoun Reference Macoun1996: xviii].
Politically, social dynamics instead found their expression in the formation of ethnic consciousness around kingdoms, in which newly emerging royalists also proclaimed “public morals” against the “decadence” of Ugandan youth [Peterson Reference Peterson2012: 257]. In this highly dynamic social and political landscape, the late colonial rulers’ understanding of their own roles was that they should balance the variety of interest groups in the protectorate and prepare for a smooth transition into a stable future. In the case of Uganda, the roles of kingdoms and their relations to the emerging republic were seen as a much greater challenge than organised labour [Thompson Reference Thompson2003: 28].
The structuration that took place during the colonial period had long-lasting impacts: the first decade of independence did not alter the social structure of Uganda, as it was inherited from colonial rule. Although a small layer of European experts and the landed gentry constituted the upper echelon, political power remained mostly in the hands of a political-bureaucratic class that tried to enlarge its power base through patronage in an expanding public sector, creating a tradition of private enrichment through the public sector [Mamdani Reference Mamdani1976: 273]. Quite typically, the most remarkable social policy measure was the expansion of pension schemes to the entire public sector.
In comparison to the two other cases considered above, the social question in colonial Uganda appeared rather late. It was, however, characterised by certain familiar features, namely the imbrication of social dislocations and perceived dangers of anomy on the one hand, and of moral and reformist agendas that were already part of the British Empire’s political arsenal on the other. In colonial Uganda, social differentiation was primarily translated into two tendencies. First, the emergence of ethnic consciousness as the main register in which to perceive social differences. Second, the growing role of the state apparatus as an agency of redistribution in what was still an agrarian society outside a few urban settings. However, modest forms of social policy and policing that included violent repression were the reactions developed in the circuits of colonial rule to these new lines of conflict in the Ugandan colonial society.
Social questions and state formation in comparative perspective
The major variance between our cases is the intensity of investment in social services and relief. While in the case of Egypt, British rulers perceived enormous social dynamics, they did not develop schemes that went beyond milder taxation for the lower classes. The answer to the social question was relatively postponed. As a result, the social contradictions continued to unfold during the twentieth century. To a certain extent, religious organisation worked as an ersatz for state-led social policies. In South Africa, changes in land tenure and a mining boom triggered social differentiation. The state formation effect of the social question was strongest among the cases considered here, despite and because of increasingly institutionalised racism. In Uganda, social differentiation set in later than in the other cases and to a less dramatic extent. The reaction of the colonial state consisted first in attempts to conserve “tribal” forms of social integration and later “development” and formal education, aimed at the production of useful colonial subjects.
Our findings thus seem to show rather different trajectories: there was no universal scheme that British domination simply applied to any context under its rule. This can be attributed to differences in “world historical timing” [Migdal Reference Migdal1988: 271]. Globally, but also within the empire as a political space, different policy schemes were favoured at different points in time. It is, for example, fairly evident from our three cases that subsequently different ideals of what would be the ideal colonial subject existed. Sometimes these models existed simultaneously in open contradiction. Their relative weight seems to be related to political majorities in the “motherland” and to its respective international situation, as the period around World War II shows most clearly. Differences were also related to local dynamics, paradoxically often triggered by the colonial rulers themselves. Colonial governments were pushed in directions they could not anticipate. With regard to state formation, British colonialism is therefore much less homogeneous than historical institutionalists have recently suggested [Lange et al. 2006]. Nor can differences be satisfyingly explained by the distinction between indirect versus direct rule [Lange Reference Lange2004].
These differences should, however, not conceal basic similarities. First, the global expansion of capitalism and modern states, greatly advanced by colonialism, resulted in very similar developments in our dissimilar sample of cases. In all three cases, a major cause of social questions was the expropriation of land from communities to landowners. What Marx called “primitive accumulation” was usually helped or even centrally advanced by colonial state agencies for various reasons: the expansion and protection of White settler farming to expand imperial control in South Africa; a historic alliance with the Bugandan aristocracy to establish imperial control; or the failure to rein in the ongoing indigenous processes of mise en valeur in Egypt. A second element was the introduction or expansion of taxation, which placed further pressure on rural livelihoods, enforced wage labour and agricultural market production, and came with the first forms of bureaucratic control in the everyday sphere.
Second, the answers to these social questions were similar to some extent and regularly resulted in the bureaucratic expansion of the state. Pass laws, segregation and land laws were attempts to regulate and re-orient undesired social developments without undermining the economic development and the financial basis of the state. Open resistance was quelled with physical force. Social policies, the emergence of welfare systems, were initially reserved for select groups. Only after World War II were these systems timidly made more inclusive. At the time, in the English-speaking global sphere, a number of hegemonic interpretations of social questions and adequate reactions might have favoured the expansion of welfare systems. To various degrees, Ugandan and South African governmental bureaucrats sought to adapt these recipes to local circumstances, also keeping in mind the perceived requirements of ruling an empire.
Our findings may sound as if Europe and Africa underwent very similar developments regarding social questions and political answers. Indeed, one may even argue that they had similar goals of moral betterment and political stability, practices with a typical paternalism of governance experts towards the poor. The only outstanding “African element” in this story may be the orientalist view of English-speaking elites towards local populations. The social question within the empire was clearly translated into cultural differences and institutionalised racism. However, despite this fundamental difference, the social question and the elite’s “answers” to it hint at a hidden aspect of the global history of state formation: although wars and taxation as vectors of state incursion mattered in Africa just as in Europe, another element of a connected and international history [Bhambra Reference Bhambra2010] has been the unforeseen social dynamics forming the state from below.
Acknowledgements
The authors would like to thank participants of the colloquium on Historical International Political Sociology at the University of Bremen, in particular Anna Wolkenhauer and Jude Kagoro, as well as participants of the workshop “Social Policies in Africa, 1880-2010: Transnational and domestic factors,” and the anonymous reviewer, for their valuable hints to improve this article.
