- 1 The current names of former colonies are preferred in this essay, not least because until the 1930s (...)
1This article asks how the legacies of European rule, both generally and in particular categories of colony, have affected post-colonial economic development in Sub-Saharan Africa. The year 1960 is conventionally used as the “stylised date” of independence, for the good reason that it saw the end of colonial rule in most of the French colonies south of the Sahara as well as in the most populous British and Belgian ones (Nigeria and Congo respectively).1 Half a century is a reasonable period over which to review the economic impact of legacies because it allows us to consider the issue in the context of different phases of post-colonial policy and performance.
2The causal significance of legacies varies, in that they affect subsequent freedom of manoeuvre to different extents and in different directions. At its strongest, legacy takes the form of “path determination”, implying that colonial choices determined post-colonial ones, or at least conditioned them, such that departure from the colonial pattern was, and perhaps remains, difficult and costly. Besides asking about the strength of the influence of the past on the future, we need to consider the nature of that influence. Did colonial rule put African countries on a higher or lower path of economic change? It will be argued here that the “path(s)” on which African economies were (to a greater or lesser extent) set by the time of independence are most usefully seen not as necessarily initiated in the colonial period, but often rather as continuations and adjustments from paths of change established before the European partition of the continent.
3The following discussion has three preliminary sections. Thus, chapter 2 first attempts a summary of the economic record since independence in order to define the pattern for which colonial legacies may have been partly responsible. Chapter 3 outlines contending views of those legacies. Chapter 4 tries to define the economic and political structures and trends within Africa on the eve of the European partition of the continent. It identifies an emerging African comparative advantage in land-extensive forms of production, which West Africans in particular were already exploiting and, by their investments and initiatives, deepening.
4In this framework, chapter 5 then introduces the colonial regimes, highlighting their fiscal constraints and comparing different national styles of colonial rule, focusing on the largest empires, those of Britain and France. It is a theme of this essay, however, that another kind of variation between colonies was more important, i.e. that defined by the extent and form of European appropriation and use of land: “settler”, “plantation” and “peasant” colonies. Chapter 6 considers how far colonial rule (and the actions of European companies that it facilitated) reinforced the emergence of a comparative advantage in land-extensive primary exports and looks at the consequences of this for the welfare of the population. Chapter 7 explores colonial contributions, and their limits, for the very long-term shift of African factor endowments from labour scarcity towards labour abundance and a relatively high level of human capital formation, such as helped Tokugawa Japan, and more recently other parts of Asia, to achieve “labour-intensive industrialisation” (Sugihara 2007). Chapter 8 assesses the impact of different kinds of European regime on African entrepreneurship and on institutions facilitating, hindering or channelling African participation in markets. Chapter 9 completes the substantive discussion by commenting on the long-term effects of the colonial intrusion on the capacity of the State in Africa for facilitating and promoting economic development.
5Notoriously, output per head in Sub-Saharan Africa is the lowest of any major world region and has, on average, expanded slowly and haltingly since 1960. But there have been important changes, and variations over space, in policy and performance. In policy, structural adjustment in the 1980s marked a watershed: a fundamental shift from administrative to market means of resource allocation. The change, however, was less dramatic in most of the former French colonies, where (except in Guinea) the maintenance of a convertible currency had enabled governments to avoid some of the supplementary price and quantity controls which had increasingly been imposed in the mostly former British colonies outside the franc zone. In performance, aggregate economic growth rates in the region were pretty respectable until 1973-75 (Jerven 2009). Ironically, in the decade or so following the adoption of structural adjustment they were stagnant or negative, before the Chinese-led boom in world commodity prices eased the region into 12 years of gross domestic product (GDP) growth at an average of 5% a year before the crises of 2007 (rising fuel and food prices, then the beginning of the international financial crisis) and 2008 brought about a “great recession” in 2009 (IMF 2009).
- 1 For a general account see Rimmer (1992, esp. 5, 228).
6There were notable exceptions to the general growth trends, both before and after the turning-point in the early to mid-1970s. Côte d’Ivoire and Ghana made a particularly interesting contrast: similarly-sized neighbours with relatively similar factor endowments and geographical features, but with different colonial heritages. Côte d’Ivoire underwent what might loosely be described as a magnified version of the standard growth trajectory. It averaged an annual GDP growth of 9.5% from 1960 to 1978 (Berthélemy and Söderling 2001, 324-5) but then had several years of stagnation followed by civil war. Meanwhile, Ghana did almost the opposite. Ghanaian GDP per capita was barely higher in 1983, when it began structural adjustment, than at independence in 1957.1 However, as one of the two most successful cases of structural adjustment in Africa (the other being Uganda), Ghana averaged nearly 5% annual growth during the quarter-century after 1983. Thus, roughly, while Côte d’Ivoire was rising Ghana was falling, and vice versa. Only one Sub-Saharan economy, Botswana, sustained growth over three, indeed four, decades since its independence, which was in 1966. Botswana averaged 9.3% annual growth (Berthélemy and Söderling 2001, 324-5).
- 2 For Africanist commentaries see Austin (2008b) and Hopkins (2009).
7A feature of the theoretical and ideological debate about the history of economic development in Africa is that it is possible to reach rather similar conclusions from very different scholarly and political starting-points. Regarding the colonial impact, the case for the prosecution, which a generation ago was urged most strongly by dependency theorists and radical nationalists (Amin 1972; Rodney 1972), is now championed by “rational choice” growth economists. Daron Acemoglu, Simon Johnson and James A. Robinson (2001; 2002) have argued that Africa’s relative poverty at the end of the 20th century was primarily the result of the form taken by European colonialism on the continent: Europeans settling for extraction rather than settling themselves in overwhelming numbers and thereby introducing the kinds of institution (private property rights and systems of government that would support them) that, according to Acemoglu, Johnson and Robinson, was responsible for economic development in Europe and the colonies of European settlement in North America and Australasia.2
8Colonial extraction in Africa could be seen most decisively in the appropriation of land for European settlers or plantations, a strategy used not only to provide European investors and settlers with cheap and secure control of land, but also to oblige Africans to sell their labour to European farmers, planters or mine-owners (Palmer and Parsons 1977). Even in the “peasant” colonies, i.e. where the land remained overwhelmingly in African ownership, we will see that major parts of the services sector were effectively monopolised by Europeans. Then there was coercive recruitment of labour by colonial administrations, whether to work for the State or for European private enterprise (Fall 1993; Northrup 1988). Of potentially great long-term importance was the unwillingness of colonial governments to accept, still less promote, the emergence of markets in land rights on land occupied by Africans, whether in “settler” or “peasant” colonies (Phillips 1989). From the perspectives of both dependency theory and “rational choice” institutionalism, the original sin of colonialism in Africa was that it did not introduce a full-blooded capitalist system, based upon private property and thereby generating the pressures towards competition and accumulation necessary to drive self-sustained economic growth.
9A narrower but important argument was made by the then small group of liberal development economists between the 1950s and 1970s. At a time when development economists (especially but not exclusively those writing in French) tended to favour a leading role for the State in the search for development in mixed economies (Hugon 1993; Killick 1978) P. T. Bauer (1953; 1972) attacked the late colonial State for introducing statutory marketing boards and thereby laying the foundation of what he considered to be deadening State interventionism.
10Explicitly positive overviews of colonial rule in Africa are rare (but see Duignan and Gann 1975). Many studies, though, mention the suppression of intra-African warfare, the abolition of internal slave trading and slavery, the introduction of mechanised transport and investment in infrastructure, and the development of modern manufacturing in the “settler” economies and in the Belgian Congo. Excited by the late 20th century wave of economic “globalisation”, some economic liberals have argued that the British empire pioneered the process through its general opposition to tariff protection (1846-1931) and by other pro-market measures (Ferguson 2003; Lal 2004). With respect to tariffs, this case would apply less strongly to French colonies because of the protectionism of the French empire. It is also much less true of the final 30 years of British rule in Africa, which saw not only tariffs but also the creation of marketing boards. From the perspective of institutional change, a fundamental observation applicable to the region generally was highlighted by John Sender and Sheila Smith (1986). Writing in the “tragic optimist” tradition of Marx’s writings on British rule in India, they emphasised that wage labour was rare at the beginning of colonial rule and increasingly common by the end of it. For them, as for Bill Warren (1980), imperialism was the “pioneer of capitalism”.
11Besides optimism and pessimism, a third view of colonial rule, and by implication of its legacy, is that its importance has been over-rated. There are different routes to this conclusion. Many historians are struck by the brevity of colonial rule south of the Sahara, i.e. about 60 years in most of tropical Africa (Ajayi 1969), and by the weakness of the colonial State (Herbst 2000). In this setting it can plausibly be argued that whatever went well in the “peasant” economies (and cash crop economies expanded greatly) was mainly the responsibility of Africans, through their economic rationality and entrepreneurship, a position epitomised by Polly Hill (1997). More ambivalent are the arguments of Jean-François Bayart (1989; 2000). Building on the familiar observation that rulers in Africa have usually found it hard to raise large revenues from domestic sources, Bayart argues that, during colonial rule and since, African elites became clients of colonial or overseas States. Thereby they forged relations which, though unequal, benefited themselves as well as the foreigners. Whereas dependency theory emphasised the primacy of foreign agency in determining historical outcomes, Bayart insists that African elites played a calculating and key role in establishing the “extraverted” pattern of African political economy.
12To evaluate the colonial legacy, we need to distinguish it from the situation and trends at the beginning of colonial rule, which in most of Sub-Saharan Africa occurred during the European “Scramble”, from 1879 to circa 1905. At that time the region was, as before, characterised generally (not everywhere all the time) by an abundance of cultivable land in relation to the labour available to till it (Hopkins 1973; Austin 2008a). This did not mean “resource abundance” as much of Africa’s mineral endowment was either unknown or inaccessible with pre-industrial technology or was not yet valuable even overseas. For example, many of the major discoveries (notably of oil in Nigeria and diamonds in Botswana) were to occur only during the period of decolonisation. Moreover, the fertility of much of the land was relatively low or at least fragile, making it costly or difficult to pursue intensive cultivation, especially in the absence of animal manure. Sleeping sickness prevented the use of large animals, whether for ploughing or transport, in the forest zones and much of the savannas. The extreme seasonality of the annual distribution of rainfall rendered much of the dry season effectively unavailable for farm work. The consequent low opportunity cost of dry-season labour reduced the incentive to raise labour productivity in craft production. Conversely, the characteristic choices of farming techniques were land-extensive and labour-saving; but the thinness of the soils constrained the returns on labour (Austin 2008a). All this helps to explain why the productivity of African labour was apparently higher outside Africa over several centuries, cf. the underlying economic logic of the external slave trades which in turn, ironically, aggravated the scarcity of labour within Africa itself (Austin 2008b; Manning 1990).
13Within Africa, the structure of incentives encouraged a high degree of self-sufficiency, and by the middle of the 20th century it was widely assumed that pre-colonial economies had necessarily been overwhelmingly subsistence-oriented. The last half-century of research has progressively changed this assessment, especially for West Africa where a strong tendency towards extra-subsistence production was evident in the 16th and 17th centuries. While damaged by the aggravated “Dutch disease” effects of the Atlantic slave trade (Inikori 2007; Austin, forthcoming), this tendency was strongly resumed from the first decade of the 19th century when that trade began to be abolished, with West Africans producing on a wider and larger scale for internal as well as overseas markets. Given the relative scarcity of labour, and in the absence (generally) of significant economies of scale in production, it was rare for the reservation wage (the minimum wage rate sufficient to persuade people to sell their labour rather than work for themselves) to be low enough for a would-be employer to afford to pay it. Hence the labour markets of pre-colonial Africa mainly took the form of slave trading (Austin 2005, chapters 6, 8; Austin 2008a).
14The same abundance of land made political centralisation difficult to achieve and sustain (Herbst 2000). Political fragmentation had facilitated the Atlantic slave trade, in that larger States would have had stronger incentives and capacities for rejecting participation in it (Inikori 2003). This fragmentation later facilitated the European conquest. Ethiopia was the exception that proved the rule, with its fertile central provinces and large agricultural surplus supporting a long-established and modernising State that, alone in Africa, had the economic base to resist the “Scramble” successfully.
- 2 The historiography is too complex to be summarised here. But it can be said that, although the moti (...)
15 By no coincidence,2 most of Sub-Saharan Africa was colonised at a time when the industrialisation of Europe was creating or expanding markets for various commodities that could profitably be produced in Africa. The land-labour ratio, the environmental constraints on intensive agriculture and also the specific qualities of particular kinds of land in various parts of the continent gave Africa at least a potential comparative advantage in land-extensive primary production. By the time of colonisation, especially in West Africa, the indigenous populations were increasingly taking advantage of the combination of these supply-side features and of access to expanding overseas markets. From Senegal to Cameroon thousands of tonnes of groundnuts and palm oil, and from the 1880s rubber, were being produced for sale to European merchants (Law 1995).
16Colonial rule in Africa was intended to be cheap, viz. for taxpayers in Europe. The British doctrine was that each colony should be fiscally self-supporting. Thus, any growth in government expenditure was supposed to be financed from higher revenues, as it was in Ghana in the 1920s when Governor Guggisberg was able to fund the creation of what became the country’s best-known hospital and school, as well as a new harbour and more railways and roads, from customs proceeds that had been fuelled by the colony’s increasing exports of cocoa beans. In practice the French were equally committed to covering costs. In French West Africa too there was a major programme of public works in the 1920s, although, as also in Ghana, within a few years expenditure had to be curtailed when export prices fell and the growth of revenue ended (Hopkins 1973, 190).
17After retrenchment during the 1930s Depression, and especially during the Second World War, colonial administrations found themselves (for a variety of reasons) entering the post-war era with a new public commitment to be seen to promote actively the development of the economies over which they presided. “Developmental” language was partly redeemed by greater spending. In principle this came partly from the metropolitan taxpayer. However, in the French case Patrick Manning (1998, 123-5) has calculated that the government continued to receive more in tax from Africa than it spent in Africa. In British West Africa the new statutory export marketing boards accrued substantial surpluses by keeping a large margin between the price paid to producers and the price that the boards received for the crop on the world market. The surpluses were kept in London, in British government bonds, as forced savings from African farmers (Rimmer 1992, 41-2), which assisted the British metropolitan economy to recover from the post-war dollar shortage.
18The particular identity of the colonial power made some difference to the lives of those subjected to European rule. Contrasts between the two largest empires in Africa are traditionally made with reference to greater British reliance on African chiefs as intermediaries (“indirect rule”) and the French doctrine of assimilating a small minority of Africans into French culture and citizenship. On the whole it is arguable that, in economic terms, the similarities were much greater than the differences, except when the latter arose from the composition of their respective African empires. French rule, like British, relied on African intermediaries, including chiefs, even though France was much more insistent on abolishing African monarchies (as in Dahomey, in contrast to the British treatment of the structures and dynasties of the States of Buganda, Botswana, Lesotho and, after an abortive attempt at abolition, Ashanti). In West Africa the French made much greater use of forced labour, but that was primarily because the French territories were, from the start, relatively lacking in cash-earning and therefore wage-paying potential. That specific policy, Corvée and its use to benefit white planters rather than African farmers, made a difference to the colonial legacy in Ghana and Côte d’Ivoire. It meant that African cocoa farming took off much more quickly and dramatically in the former, so that Ghana was much wealthier at independence, when Côte d’Ivoire was in the process of catching up (and overtaking) after a late start (Hopkins 1973, 218-9), which it proceeded to do by the 1980s.
- 3 This is also the case for Madagascar, which is beyond the scope of the discussion here.
- 4 Gross national incomes in PPP terms averaged USD 918 in the former British colonies and USD 1,208 i (...)
19The contention here that the differences between the legacies of British and French rule in Africa are primarily attributable to variations in the composition of the African empires concerned may need to be qualified in the light of valuable recent research by Thomas Bossuroy and Denis Cogneau (2009). They examined social mobility in five African countries and found that in the former British colonies in their sample, Ghana and Uganda, “the links between origin, migration, education and occupational achievement appear much looser” than in the former French colonies they examined, i.e. Côte d’Ivoire and Guinea3 (Bossuroy and Cogneau 2009, 2). In explanation, they emphasise the importance of greater investment in education in the British colonies than the French colonies in their sample. This is a novel and important line of inquiry. I suspect that the favourable conclusion about the former British colonies also partly reflects the fact that Ghana and Uganda, for reasons that are only partly and indirectly connected to their respective British legacies, were the two major growth successes of structural adjustment in Africa, therefore perhaps offering greater opportunities for educational, physical and occupational mobility from the mid-1980s, which was early enough to be partly reflected in the data and which coincided with economic stagnation and then civil war in Côte d’Ivoire. Finally, the contrast may also partly reflect the legacy of the era of Corvée and “settler” agriculture in Côte d’Ivoire, before the economy took off in the 1950s and 1960s. As of 1990, helped by the legacy of the Ivorian “miracle”, in a sample of 26 former British and French colonies in tropical Africa (so excluding southern Africa) it was the former French colonies that had the higher per capita incomes in Purchasing Power Parity (PPP) terms, by over 30% (Bossuroy and Cogneau 2009, 45, citing World Bank data).4
20The “extraversion” and “monoculture” of African economies is widely lamented and condemned as a victory of colonial interests over African interests. The risks entailed in extreme specialisation, however, need to be set against the long-run income gain to be expected from the exploitation of comparative advantage. But again, although the location of a colonial economy’s comparative advantage could be identified, sooner or later the task of capitalising on it raised the question of what investments might profitably deepen that advantage and, above all, of how the costs and benefits would be distributed. Conflicts of ideology, and especially the balance of power between different interest groups, worked out variously across the range of African colonies. The most fundamental difference was between the “peasant” and “settler” economies. Let us consider the contrasting cases of export agriculture in the former, notably in West Africa, and mining in the latter, most obviously in South Africa.
21We have noted that, by the eve of the European partition of the continent, Africa had already revealed an emerging comparative advantage in export agriculture. In West Africa in particular it was in the joint interests of the population, European merchants and the colonial administrations to further this. In Ghana British planters were initially allowed to enter to grow cocoa beans. But lacking the discriminatory support from the government that their counterparts enjoyed in Kenya and southern Africa, they failed in commercial competition with African producers (Austin 1996a), just as French planters were later to be eclipsed by African ones in Côte d’Ivoire following the abolition of Corvée. Colonial reliance on the efforts of African small capitalists and peasants in the growing and local marketing of export crops paid off in what became Ghana and Nigeria, with more than 20-fold increases in the real value of foreign trade between 1897 and 1960 (Austin 2008a, 612), benefiting British commercial interests as well as (via customs duties) the colonial treasury. The efforts of W. H. Lever, the soap manufacturer, to win government permission, along with the necessary coercive support, to establish huge oil palm plantations in Nigeria continued from 1906 to 1925, but they were always rebuffed in favour of continued African occupation of virtually all agricultural land. Ultimately this was because African producers literally delivered the goods (Hopkins 1973, 209-14) through land-extensive methods well adapted to the factor endowment. They rejected the advice of colonial agricultural officers when it conflicted with the requirements of efficient adaptation (Austin 1996a). The positive contribution of the administrations was to reinforce and permit the exploitation of these economies’ comparative advantage in export agriculture. They did this partly by investment in transport infrastructure, investment to which African entrepreneurs also contributed (Austin 2007). Equally important, although the colonial administration never really established a system of land titling, in Ghana (for example) it upheld the indigenous customary right of farmers to ownership of trees they had planted, irrespective of the outcome of any later litigation about the ownership of the land the trees stood upon. Thus, African producers enjoyed sufficient security of tenure to feel safe in investing in tree crops on a scale sufficient to create, in the case of Ghana, what became for nearly 70 years the world’s largest cocoa economy (Austin 2005, chapters 14, 17).
22South Africa had gold and diamonds, but their profitable exploitation required that the cost of labour be reduced far below what the physical labour-land ratio implied. C. H. Feinstein’s quantitative exercise indicates that without such coercive intervention in the labour market, most of South Africa’s mines would have been unprofitable until the end of the gold standard era in 1932 (Feinstein 2005, 109-12). If South Africa eventually obtained a “free market” comparative advantage in mining, it was only after several decades of using extra-market means to repress black wages, notably through land appropriation and measures to stop Africans from working on European-owned land except as labourers rather than tenants.
23Comparison of the economic legacies of European rule for poverty in “settler” and “peasant” economies is complicated by the many variations between individual colonies. However, some generalisations are possible. It is clear that the distribution of wealth and income was, and has remained, much more unequal in the “settler” economies than in the “peasant” ones. Preliminary findings by Sue Bowden, Blessing Chiripanhura and Paul Mosley (2008) support the proposition that possession of land put a floor under real wages in the “peasant” colonies, enabling labourers migrating into export-crop growing areas to share in the gains from exports that were otherwise divided between European firms, African and Asian middlemen and African farm-owners (see also Austin 2005). Bowden, Chiripanhura and Mosley find real wages beginning to rise from the 1920s and 1930s in the “peasant” colonies of Ghana and Uganda respectively and not falling back afterwards to the 1914 floor. In contrast, it was only in the 1970s that the real wages of black gold-miners in South Africa began a sustained rise above their early 20th century level (Lipton 1986, 410). In the sample taken by Bowden, Chiripanhura and Mosley it was only in the “pure settler” economies, South Africa and Zimbabwe (Southern Rhodesia), not in the “peasant” colonies of Ghana or Uganda nor even in the intermediate case of Kenya, that there were declines in rural African living standards over periods of longer than 15 years within the 20th century. This pattern in real wages, together with the long-term expansion of African export agriculture which underpinned the growth of real wages in the “peasant” colonies, was reflected in the earlier onset of falling infant mortality in Ghana and Uganda compared to Southern Rhodesia and South Africa.
24It should be added that many African colonies were short of both known mineral deposits and the kinds of land suitable for profitable export agriculture. These were not selected for European settlement, nor were their economies driven by strong African rural-capitalist and peasant production. They had to rely on seasonal exports of male wage labourers, and on growing the less lucrative cash crops such as cotton, the timing of whose labour requirements conflicted with those of food crops, thereby creating risks to food security (Tosh 1980). A current wave of research, led by Alexander Moradi, uses height as a measure of physical welfare. The average height of African populations rose during the colonial period in Ghana and even in the “semi-settler” economy of Kenya (Moradi 2008; 2009). When this research is extended to poorer colonies such as southern Sudan, Tanganyika (mainland Tanzania) or those in the West African Sahel, it would be no surprise if welfare improvements there are found to have been smaller than in the better-endowed economies studied so far. It was particularly in (selected areas of) the less favourably-endowed economies that colonial governments sought to raise productivity through very large-scale, capital-intensive and authoritarian projects, notably the massive irrigation scheme of the Office du Niger in Mali and the mechanisation campaign of the East African Groundnut Project in Tanganyika. Both were spectacular failures in their own output and productivity terms, not least because they were inefficient in relation to the prevailing factor ratios and physical environments (Hogendorn and Scott 1981; Roberts 1996, 223-48; Van Beusekom 2002).
25Poor as was the record of “settler” colonialism for the living standards of the indigenous population, it was in colonies where Europeans appropriated land on a large scale, for settlers or for companies, that the earlier and larger beginnings were made in modern manufacturing.
26Where industrialisation has occurred in Asia, it has tended to follow a more labour-intensive route than in Europe and North America, substituting longer working hours for additional machinery where possible (Sugihara 2007) and generally having a higher proportion of labour to capital at any given level of output. A region in which labour as well as capital was scarce in relation to land, such as Sub-Saharan Africa, was not well suited to follow either route in the early 20th century.
- 5 In Katanga, in contrast to South Africa, the black labour force was “stabilised” from the 1920s onw (...)
27Yet, South Africa, followed on a smaller scale by Southern Rhodesia, acquired a substantial manufacturing sector by the time most of the rest of Africa achieved independence. The “artificially” low cost of black labour helped, but only in unskilled jobs because the skilled ones were anyway reserved for whites and the choice of technique was generally capital-intensive. Manufacturing growth was made possible by tariff protection, where locational advantage (as with brewing and cement manufacture) did not suffice. Crucially, mining provided the import-purchasing power to cover the import of capital goods and, where necessary, raw materials. It was also the direct or indirect source of much of the revenue used by governments to invest in manufacturing, whether directly or through the provision of infrastructure. The large European populations were a source of both educated workers and capital, but arguably their most important contribution to industrialisation was the political commitment to support it even at the cost of consumer prices that were often above world market levels (Austen 1987, 181-7; Kilby 1975; Wood and Jordan 2000). Moving up the value chain became an ambition of substantial proportions of white voters where they controlled governments, as in South Africa after 1910 and to a large extent in Southern Rhodesia from 1923, as well as of African voters since independence. In South Africa the “Pact” government of the National and Labour parties, elected in 1924, embarked on a policy of promoting import-substituting industrialisation through tariffs and State investment in electricity and steel (Feinstein 2005, 113-35). Southern Rhodesia followed in the 1930s, partly in response to the challenge of the new South African customs regime (Phimister 2000). Besides these “settler” colonies, there was a third case of precocious growth of modern manufacturing, i.e. the Belgian Congo. This was absolutely not a case of settler independence or autonomy. As in southern Africa, however, mines provided a favourable context for import-substituting industry, providing infrastructure, import-purchasing power and part of the market.5 South Africa remained the flagship of manufacturing in the region, but the scope for further expansion was increasingly restricted by the high price of skilled labour in an economy where only a minority of the population had access to secondary education and by the limited market for mass-produced goods that resulted from the low level of black wages. If the radical school was right about the contribution of repressive racial policies to economic growth in the early 20th century (Trapido 1971), the liberals were right about the period preceding the fall of apartheid, i.e. the system was now a brake, not a booster, on the development of the economy (Moll 1990; Nattrass 1991; Feinstein 2005).
- 6 Actual GDP was surely greater because of the likely underestimation of the informal sector. This im (...)
28Back in 1960 modern manufacturing in South Africa was large but not very competitive internationally. In the rest of Sub-Saharan Africa it was much smaller. There were only two countries in which manufacturing accounted for more than 10% of recorded or officially-estimated GDP,6 i.e. Southern Rhodesia (16%) and the Belgian Congo (14%). Next, on 9.5%, the “semi-settler” economy of Kenya tied with Senegal (Kilby 1975, 472). The latter was a “peasant” colony but, as the administrative and commercial centre of French West Africa, had an exceptionally large resident European population, which increased the supply of people with managerial experience, technical expertise and access to capital (Kilby 1975, 473, 488-90). In West Africa even these low 1960 levels of manufacturing represented a very late surge, propelled by post-war developmentalism (government subsidies for manufacturing in the case of Senegal) and decolonisation, which led European firms to establish local factories to protect their existing markets (Kilby 1975, 475, 490-507; Boone 1992, 65-77).
29Given the relative scarcity of labour and the small markets, together with the comparative advantage in land-extensive primary production, it is not surprising that there was not much more manufacturing by the end of the colonial period. Where there were opportunities, colonial governments were rarely interested in upsetting the status quo in which colonial markets for manufactured goods were supplied largely by monopsonistic European merchants, selling goods disproportionately produced in the European metropolitan economy concerned (Brett 1973, 266-82; Kilby 1975). But given that, despite rising population, the factor endowments of even the larger African economies were not suited to industrialisation in 1960, the more important question is perhaps whether colonial rule, directly or indirectly, laid foundations on which Africa might later develop the conditions for a much larger growth of manufacturing.
- 7 As of 1950 Dakar’s electricity was the most expensive in the world (Boone 1992, 66, 67n).
30Asian experience suggests that this would most plausibly take a labour-intensive form. In the long term the most fundamental change of the colonial period was probably the start of sustained population growth, which in aggregate can be dated from the end of the 1918 influenza pandemic, although local timing varied. How far the demographic breakthrough was the result of colonial actions, such as the suppression of slave raiding, the post-1918 peace within Africa and public health measures that reduced crisis mortality, is difficult to determine (Iliffe 1995, 238-41). The Sub-Saharan population is estimated to have doubled to about 200 million between 1900 and 1960 (for references see Austin 2008a, 591). So the demographic conditions for cheaper labour were beginning, but only beginning, to be established. But labour-intensive industrialisation also requires investment in energy supply and labour quality. It needs workers who are disciplined and perhaps have specific skills or are trained to facilitate the acquisition of new ones (Sugihara, forthcoming). School enrolment rates rose during the colonial era from low or non-existent levels and in many countries doubled or tripled between 1950 and 1960. This was especially helped by African politicians gaining control of domestic budgets during the transition to independence, such as in Nigeria where primary enrolment was raised from 971,000 to 2,913,000 and secondary enrolment was raised from 28,000 to 135,000 (Sender and Smith 1986, 62). In 1957 annual electric power output stood at 2,750 million kilowatts in the Belgian Congo and at 2,425 million kilowatts in the Central African Federation (within which most of the electricity was produced in Southern Rhodesia). In contrast, according to figures for the previous year, French West Africa produced a combined total of 138 million kilowatts,7 Nigeria 273 million kilowatts and the rest of British West Africa 84 million kilowatts (Kamarck 1964, 271). Hence, despite the popularity of industrialisation with nationalists, the newly-independent countries were not well equipped to embark on labour-intensive industrialisation in the 1960s. Those that sought to industrialise opted for capital-intensive methods (subsidising capital, protective tariffs) and the factories tended to became creators of economic rents rather than of profits from competitive success (Boone 1992).
- 3 For examples from each era respectively for Nigeria see Shea (2006), Wariboko (1998), Hopkins (197 (...)
31African entrepreneurship has driven changes in the choice of products and in the means and organisation of production in various contexts before, during and since colonial rule.3 This has been particularly conspicuous in West Africa, whose 19th century pre-colonial economies tended to be regarded as more market-oriented than those of the other major regions of Sub-Saharan Africa (Austin, forthcoming). The colonial impact on African entrepreneurship and on the markets in which they operated again turned to a large extent on whether there were large-scale appropriations of land for the use of Europeans, be they individual settlers or corporations.
32This familiar division between “settler” and “plantation” colonies, on one hand, and “peasant” (and rural capitalist) colonies, on the other, was far from purely exogenous to African economic history. Where African producers were able to enter export markets early and on a wide scale, before European exporters really got going, their success was sufficient to tip the balance of the argument among colonial policy-makers in favour of those who thought it economically as well as politically wisest to leave agricultural production in African hands. As we saw in chapter 6, British West Africa was the major example of this. In South Africa, Southern Rhodesia and Kenya African farmers responded quickly to opportunities to grow additional grain to supply internal markets. But the governments reacted by trying to drive Africans out of the produce market and into the labour market by reserving land for Europeans, while either prohibiting Africans from leasing it back or (as in inter-war Kenya) limiting the time that African “squatters” could work for themselves rather than for their European landlords (Palmer and Parsons 1977; Kanogo 1987). African production for the market proved resilient, however, and the governments eventually accepted this and shifted to imposing controls on agricultural marketing that favoured European producers rather than trying to displace African ones. In Kenya it was only in the mid-1950s, during the Mau Mau revolt, that the government lifted restrictions on African production of high-value cash crops (Mosley 1983). Thus, to the extent that African production for the market in the late 19th century was greater in what became the “peasant” agricultural-export economies than in what became the “settler” economies, that contrast was reinforced by government actions in the latter over the following decades.
33Not that the maintenance of African ownership of land necessarily entailed support for African capitalism. Admittedly, we have seen that the colonial State in Ghana protected the property of agricultural investors, in the sense of preserving the ownership of a farmer over trees or crops that he or she had planted, irrespective of the outcome of legal disputes about the ownership of the land on which they stood. But in “settler” and “peasant” colonies alike colonial governments were hesitant and usually hostile to the emergence of land markets in areas controlled by Africans. This policy eventually changed in Southern Rhodesia and Kenya, with selective promotion of land registration, in response to the de facto emergence of land sales and individual proprietorship (cultivable land having become increasingly scarce in the areas left to Africans) and with African land-owners being seen as a politically conservative force in the context of Mau Mau (Mosley 1983, 27-8; Kanogo 1987). In West Africa, without the settler pressure on African access to land, and given the expansion of cash crops that occurred early in the colonial period and again in the 1950s, neither the political case nor the economic case for compulsory land titling was as yet compelling (Austin 2005).
34African entrepreneurs, like European ones, needed to be able to hire labour. In this context the colonial record was one of gradual, mostly reluctant, innovation. Sooner or (often) later, they legislated against slavery. But in West Africa, the region with evidently the largest slave population at the start of the 20th century, the replacement of the slave market by a wage labour market depended very much on the progress of African cash crop agriculture (Austin 2009). During the inter-war decades the continued use of forced labour by colonial administrations came under sustained pressure from the International Labour Office in Geneva. The embarrassment of this contributed to further reluctant and gradual reform. By the end of the Second World War, as Frederick Cooper (1996) has shown, British and French authorities had accepted that wage labour had become a regular occupation for Africans, rather than a seasonal sideline from farming. Indeed, Cooper went on to show that in London and Paris the long-run fiscal implications of having to give workers in Africa the same rights as workers in Europe contributed to the decisions to withdraw from tropical Africa. For African societies the end of slavery and the rise of wage labour was arguably a condition of continued large-scale participation in international trade. As early as 1907 the chocolate manufacturer Cadbury had moved its cocoa-buying to Ghana following bad publicity about “slave-grown” cocoa in the Portuguese colony of Sao Tome, where it had been buying before (Southall 1975, 39-49). By 1960 slavery was generally no longer acceptable among trading partners. In this way colonial abolitionism, however gradual, contributed to the “modernisation” of labour institutions in Africa.
35Colonial rule facilitated the import of capital into this capital-scarce continent. But only in mining, and to some extent in “settler” and “plantation” agriculture, did this happen on a large scale. The survey by Herbert S. Frankel (1938) of external capital investment in white-ruled Africa remains the only comprehensive study for the colonial period. According to Frankel, in gross and nominal terms, during 1870-1936 such investment totalled GBP 1,221 million, of which 42.8% was in South Africa. This meant GBP 55.8 per head in South Africa, but only GBP 3.3 in the French colonies and GBP 4.8 in British West Africa. Public investment constituted 44.7% of the grand total and almost 46% of the non-South Africa total (Frankel 1938, 158-60, 169-70). Governments, and to some extent mining and plantation companies, invested in the transport infrastructure required for the development of, mainly, the export-import trade. In Nigeria and Ghana Africans also took a leading role in building motor roads and pioneering lorry services (Heap 1990). In institutional terms the colonial period saw the eventual abolition of human pawning, with its replacement by promissory notes and, in those areas of West Africa where it was possible, by loans on the security of cocoa farms. It also saw the introduction of modern banking, but the banks were much more willing to accept Africans’ savings than to offer them loans, partly because of the colonial governments’ non-introduction of compulsory land titling (Cowen and Shenton 1991).
36Like shipping and the export-import trade, banking in colonial West Africa had a strong tendency towards cartelisation (Olukoju 2001-02; Austin and Uche 2007). The initial imposition of colonial rule and boundaries itself disrupted intra-African networks of exchange, and the increasing presence of European merchants in the interior relegated many African traders further down the chain of intermediation between shippers and farmers (Goerg 1980; Nwabughuogu 1982). Organised resistance to European cartels mostly emerged later and was largely confined to particular colonies, given the tradition of cocoa “hold-ups” with which African farmers and brokers confronted successive European merchant cartels in Ghana and the indigenous banking movement in Nigeria (Miles 1978; Hopkins 1966). Until the advent of independence it remained the case in the “peasant” colonies that the markets dominated by Europeans were cartelistic, whereas the markets populated by Africans were characterised by extreme competition (Hopkins 1978, 95). At least, much more than in the “settler” colonies, African entrepreneurs were able to operate in the export-import as well as in the domestic exchange sectors. Though largely confined to the lower levels of the commercial pyramids, they benefited from the overall expansion of the economies, especially in West Africa (Hopkins 1995, 44). The monopolistic arrangements were shaken a bit by decolonisation (Austin and Uche 2007), but the old African sector became the “informal” sector, the old European sector the “formal” one (Austin 1993). At independence, new governments were faced with the familiar problems of this financial dualism, notably lack of cheap credit from the formal sector for informal enterprises.
37Despite asymmetric competition, the more economically successful “peasant” colonies saw the continuation of a tradition of entrepreneurship and mostly (but not always) small-scale accumulation in agriculture, crafts and trade. The result, as John Iliffe (1983, 67) noted, was “a strong contrast between West Africa, with its long-established capitalistic sector and its entrepreneurs from artisanship and trade, and eastern and southern Africa, where entrepreneurs had emerged chiefly through (…) Western education and modern sector employment”. Early post-colonial policy did not always build on this, for example in the case of Ghana, with high taxation of export agriculture and the creation of State monopolies in certain sectors (Austin 1996b).
38It is widely accepted that States have a critical role in economic development, at least in enforcing the rules of economic activity and providing physical public goods. Therefore, we should ask how colonial rule affected the historic constraint on political centralisation in Africa, namely the difficulty of raising revenue. Beyond this we need to consider the size of the State and the nature of authority and legitimacy, i.e. whether colonisation was responsible for fragmenting Africa, as is often said, or whether, as the colonial rulers themselves claimed, they were a modernising force, bringing the State to the “Stateless” and replacing patrimonial authority by bureaucratic authority.
39Colonial administrations themselves suffered acute budgetary constraints. Although European empires introduced to Africa the possibility of raising loan finance (at least in an impersonal, law-governed though undemocratic way), the colonial administrations were restricted in their resort to money markets by the metropolitan insistence that each colony be fiscally self-sufficient and balance its budget. The introduction in each colony of a single currency as legal tender probably reduced net transaction costs (although in some cases the demonetisation of existing currencies hurt Africans holding them). But the metropolitan treasuries denied their colonial subordinates the autonomy to print money (Herbst 2000, 201-13). The French colonies used the French franc. In British West Africa a colonial pound was issued, but the rules ensured that it was always convertible at par with the metropolitan pound. It was only at independence that the new African governments had the option of creating national currencies, an option the former French colonies mostly declined while the former British colonies soon accepted.
40Given that they faced much the same practical constraints as the African States that had preceded them, colonial governments generally continued the reliance of pre-colonial kingdoms upon taxes on trade and people, rather than on land or agriculture. It was the above-mentioned discovery, during the Second World War, that the export marketing board could be a major revenue-raiser, which was the major fiscal innovation of colonial rule. As independence approached, this unintended consequence of a wartime expedient offered African politicians unprecedented opportunities to, for example, transform educational opportunities for their populations. The marketing board as a fiscal instrument was an important colonial legacy, and its possibilities and implications were only beginning to be understood. By the 1980s the limits of the device had become clear, as ordinary traders and producers could evade it by trading on parallel markets (Azarya and Chazan 1987).
41Smuggling brings us to one of the more notorious legacies of the colonial partition of Africa: the imposition of boundaries dividing people of shared culture, the delineation of some States so small as to be of questionable economic viability and the creation of some States so large as to be potentially ungovernable. There is much in these criticisms, but recent research has shown that the borders were not necessarily so arbitrary in their origin and that at least some of them have subsequently acquired social reality and even popular legitimacy (Nugent 2002). Again, while the colonial legacy includes several very small States, most colonies (even the small ones) were larger than the pre-colonial polities on which or in place of which they were imposed; and some of them formed parts of larger regional units (notably French West Africa). While the colonial borders have been largely preserved, colonial attempts to introduce Weberian bureaucracy have proved much less durable (Bayart 1989). One reason for, or manifestation of, this is the salience of ethnicity in most African countries for political competition over resources.
42From the late 1970s onwards a generation of historians and anthropologists tended to argue that ethnicity in Africa, far from being “primordial”, was created, or at least greatly entrenched, by colonial strategies of “divide and rule” (seminal contributions included Iliffe 1979, 318-41, and Ranger 1983, although the thesis was pursued beyond those cautious initial statements). Recent historiography has shown that the emphasis on the capacity of colonial States to invent and manipulate traditions, including those relating to ethnicity and chieftaincy, was partly justified, but it underestimated the capacity of African elites and peoples to influence the outcomes themselves (Spear 2003). By no means all ethnic divisions originated in the colonial period (Vansina 2001), although they were usually deepened and reified by the interaction of colonial and African elites (Prunier 1995). Whatever the precise division of responsibility in this interaction, there is general agreement among scholars that ethnicity has been a more important organising principle of political association and conflict since colonial rule than before it. This matters for economic development because ethnic divisions are often seen, by public opinion and by some economists (notably Easterly and Levine 1997), as primarily responsible for rent-seeking rather than growth-promoting policies in post-colonial Africa. However, that approach has been criticised on various grounds (notably by Arcand, Guillaumont and Jeanneney 2000), and it is arguable that the salience of ethnicity in African political and economic life is as much a response to as a cause of the difficulties of enlarging the economic cake in African conditions and of the continued weakness of State capacity.
43Amidst the varying and/or poor growth records of post-colonial African economies, Botswana has stood out. Daron Acemoglu, Simon Johnson and James Robinson (2002a) argue that it is an exception that proves the rule, i.e. that while Botswana did not have the beneficial institutional legacy characteristic of the “full settler” colonies like Australia, it was exceptionally lightly ruled by Britain and as a result escaped the worst of the extractive propensities that they see as generally characteristic of non-“settler” colonialism. In my view two considerations point to a different conclusion. First, without the discovery of diamonds, it is hard to see how post-colonial Botswana could have grown dramatically faster than colonial Bechuanaland. Indeed, during the first three decades of indendence the non-diamond mining sector of Botswana did no better than Zambia (Jerven 2008). Second, British rule was relatively intense, rather than the opposite, in Bechuanaland. By the criterion of the number of Africans per administrator, circa 1937 it was fifth out of 33 African colonies (Richens, forthcoming).
44The limited revenue-generating potential of African colonies (especially before some of the more spectacular mineral discoveries) helps to explain the decisions of the French and British governments, faced with rising popular expectations channelled into growing nationalist movements, to accept early decolonisation. Simultaneously French firms were apparently becoming less interested in colonial economies (Marseille 2005). If so, it is ironic that the French government remained closely involved with its former colonies after their independence, not least through the franc zone. Again, in the 1950s British firms on the spot expressed concern about their future under independent African governments, but they failed to attract much notice from the decolonising authorities (Tignor 1998; Stockwell 2000). The irony of the latter case is that a few years later the British government’s attitude to the Biafran secession was influenced by the interests of British oil companies (Uche 2008).
45This article has considered the issue of colonial legacies in relation to the longer-term dynamics of economic development in what was in 1900 an overwhelmingly land-abundant region, characterised by simultaneous shortages of labour and capital, by perhaps surprisingly extensive indigenous market exchanges, especially in West Africa, and by varying but often low levels of political centralisation. Colonial governments and European firms invested in both infrastructure and (especially in southern Africa) in institutions designed to develop African economies as primary-product exporters. In both cases the old economic logic for coercing labour continued to operate, i.e. the continued existence of slavery in early colonial tropical Africa and the use of large-scale land grabs to promote migrant labour flows in “settler” economies. But there were changes and variations. While we have noted differences between French and British policy, for example in West Africa, the bigger contrasts were between “peasant” and “settler” colonies.
46In British West Africa in particular there was a genuine coincidence of interest between African farmers, European merchants and colonial governments in enlarging and exploiting West Africa’s comparative advantage in land-extensive agriculture. The resultant income at least enabled many of the slave-owners to become employers instead. In these cases the British government (and the French in Côte d’Ivoire after 1945) correctly recognised where their own self-interest lay when they supported African investment in export agriculture. It was in those “peasant” colonies that were best endowed with lands suitable for producing the more lucrative crops that African populations experienced significant improvements in purchasing power and had the most improvement in physical welfare. In the same countries, however, colonial rulers, partly because of fiscal constraints as well as a probably realistic assessment of the short-term economic prospects, did little directly to prepare the economies to move “up the value chain”. Thus, the first generation of post-colonial rulers presided over economies which were as yet too short of educated (and cheap) labour and sufficient (and sufficiently cheap) electricity to embark successfully on industrialisation. It has taken post-colonial investment in education and other public goods to move West African economies, and tropical Africa generally, closer to the prospect of a substantial growth of labour-intensive manufacturing, if international competition permits it.
47 “Settler” colonies had a worse record for poverty reduction, especially considering the mineral resources of South Africa and Southern Rhodesia, but they had a better one for structural change. The large-scale use of coercion was the basis for the construction of white-ruled economies that, especially in South Africa, eventually became profitable enough for a partly politically-impelled policy of import-substituting industrialisation to achieve some success. Thus, the rents extracted from African labourers were channelled into structural change, although the process became self-defeating as it progressed, contributing to the fall of apartheid.
48As promoters of market institutions, the colonial regimes had a very mixed record; but probably in all Sub-Saharan countries there was far more wage labour, and a lot more land sales, and a lot more people more deeply dependent on markets, by 1960 than there had been in 1890 or 1900. A final legacy of the colonial period has a rather unclear relationship to colonial policy, i.e. the sustained growth of (total) population since 1918 has progressively transformed the factor ratios and, on the whole, increased the long-term economic potential of the continent.