Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Foreign direct investment

On average over 95 per cent of investment in any developing country is financed from domestic savings (IMF, 1993). As discussed above, however, FDI is sought for its anticipated advantages in giving access to technology and export markets (Jenkins, 1987). Much has been concentrated in a small number of countries with large domestic markets, rich natural resources, or advantages as a base for export-oriented production. In 1975, 19 per cent of the world's stock of FDI was in LDCs. By 1982, the proportion of global FDI that was in LDCs had increased to 24 per cent but Africa's share of it had decreased from 21 per cent to 11 per cent (Cantwell, 1991, pp. 187189). Its share of new FDI was even less (Cantwell, 1991). The IMF acknowledged that, even when the smaller, resource-poor African countries offered substantial incentives and imposed few restrictions, they were unsuccessful in attracting investment (IMF, 1985, p. 4). Thus between 1976 and 1986 significant net inflows occurred to only half a dozen countries. Most countries experienced net disinvestment, not least because of the effects of structural adjustment (see below; Cockcroft, 1992, p. 337).

The sectoral composition of worldwide FDI changed over the period, initially away from oil, mining, and agriculture to manufacturing. Nationalizations and the effect of policies to restrict entry of new foreign capital lessened TNC interest even in oil and mining in Africa (IMF, 1985; UNCTC, 1991). Nevertheless by the early 1980s half the total stock of FDI was still in the extractive sectors, compared with about 20 per cent in non-African LDCs, and almost 40 per cent of foreign-owned primary commodity enterprises, concentrated in oil extraction and the mining of copper, iron, bauxite, and uranium, were in Africa. To the extent that FDI occurred in manufacturing (about a quarter of the stock in Africa by the beginning of the 1980s compared with half in Asia and over half in Latin America), it has been in import substitution and resource-processing industries in the larger, more developed economies and so has been of only limited importance to urban economies (Cantwell, 1991).

Between 1983 and 1989, FDI outflows worldwide increased 29 per cent per annum, as the global economy recovered from the early 1980s' recession, although the share of LDCs in this investment fell to about a fifth (UNCTC, 1991). The destination for over three-quarters of these flows continued to be 10 countries, of which only Egypt was in Africa. Although Africa's share of worldwide FDI remained stable, at just over 2 per cent throughout the 1980s, and its share of FDI in LDCs increased slightly in the second half of the decade, the volume of investment was minute by world standards and concentrated on oil-related investment in Egypt and Nigeria (UNCTC, 1991). Low-technology, low-value-added import substitution (IS) industries continued to dominate most countries' industrial structure, yielding lower rates of return to foreign investors than manufacturing elsewhere in the world (Bennell, 1990). Only a tiny share of investment originating in Japan and the United States goes to Africa. Most FDI in the continent originates in the former colonial powers (UNCTC, 1985), although Bennell concludes that, for British industrial capital as a whole, Africa is now of only marginal and decreasing interest (Bennell, 1990).

FDI in manufacturing worldwide has continued to grow over the past 10 years. However, a further sectoral shift, which had started in the 1970s, gathered pace in the 1980s. Services FDI, including business and financial services, trade, tourism, and construction, grew to comprise over half of all annual flows of FDI and 40 per cent of FDI stock worldwide. Intermediate services, such as financial, business, and professional services, and those competing for the discretionary income of consumers, are concentrated in developed countries, whereas FDI in services in LDCs traditionally concentrated in trade and construction, and has more recently diversified into tourism and banking, especially in those countries offering "tax haven" facilities (UNCTC, 1989). Recent growth in FDI in services in Africa has been concentrated once again in a few countries (Egypt, Morocco, Nigeria) and in the traditional areas of trade and construction, rather than the financial services sector, which has been so important globally. However, a very high proportion of the tiny but growing Japanese investment is in trade and financial services (UNCTC, 1989). Although Africa has shared in the increased global flows of investment in services, as with manufacturing its position is marginal.

Development in Africa has been held back not only by limited flows of FDI but also by low levels of domestic savings and investment, which fell from 27 per cent of GDP between 1971 and 1975 and 30 per cent between 1975 and 1981 to 21-22 per cent in the 1980s and early 1990s (IMF, 1993; see also Husain, 1993). Reinforcing the shortage of capital for investment, FDI has had a negative effect on the balance of payments: outflows of dividends, royalties, management fees, etc. exceed inflows of new investment and are financed from commodity exports, while the investment in manufacturing is generally for the domestic market and absorbs rather than earns foreign exchange (Cockcroft, 1992). The limited volume of FDI in manufacturing has exacerbated the difficulties experienced by African countries in generating sufficient wage jobs for growing urban labour forces.

Aid dependence

External and internal shocks, a general deterioration in the terms of trade for many of Africa's most important products, and economic mismanagement have, as we have seen, led to a continued need for foreign exchange, which is not satisfied by earnings from exports, commercial bank borrowing, or FDI. As a result, Africa became the largest recipient region of Official Development Assistance (ODA) by the 1980s, accounting for 43 per cent of all bilateral and multilateral development assistance by 1990 (12 per cent to north Africa and 30 per cent to SSA) (Simon, 1995). Aid receipts as a percentage of GNP were 9.6 per cent for SSA as a whole in 1990 or US$34 per capita. The poorest and most aid-dependent countries were Mozambique, its economy wrecked by civil war, in which 66 per cent of GNP in 1990 came from aid (US$60/head), Tanzania (48 per cent of GNP, US$47/head), and Somalia (46 per cent of GNP, US$55/head). Although other countries were less aid dependent, in several between a fifth and a third of GNP came from aid. The recipients of the largest volume of aid per capita, however, were not the poorest countries: the level in Mozambique was exceeded by four other countries, none among the poorest (Egypt US$108/head, Mauritania US$107/head, Senegal US$100/head, and the Congo US$92/head) and that in Tanzania by a further four, with relatively rich countries such as Morocco and Tunisia not far behind (US$39/head) (World Bank, 1992). Aid receipts, clearly, are related to factors other than need. It should also be noted that, whereas net credits from the IMF were positive in the early 1980s, they were negative later in the decade (Bird, 1993; ODI, 1993).

The reliance of African countries on aid has rendered them vulnerable to policy conditionality, both by the multilateral agencies, which supplied just under one-third of ODA to SSA during the 1980s (although a much smaller proportion to north African countries), and, in their wake, by the bilateral donors, especially the United States. Policy conditions have affected both general economic management and approaches to urban development. The latter will be considered in the next section.

The difficulties caused for African countries by the external shocks of the 1970s and early 1980s (oil price increases, world recession, and increased interest rates) were initially seen as temporary trade crises and tackled by IMF stabilization packages. However, endogenous shocks (drought, civil strife) and economic mismanagement combined with external vulnerability showed that the economic crisis had deeper roots, resulting in the addition of structural and sectoral adjustment programmes to the stabilization programmes. By the end of the 1980s, 30 African countries had adopted structural adjustment policies, many implementing a succession of programmes. Only a few small countries had no IMF/World Bank programmes and several had introduced some kind of "home grown" programme (Jesperson, 1992). As a result, 29 per cent of IMF commitments by value were to SSA countries in April 1989 (ODI, 1993, p. 2).

The Structural Adjustment Programmes (SAPs) adopted by most countries comprised three basic components:

(1) "[A] reform of the system of economic incentives with the objective of changing the structure of incentives in favor of tradeables versus nontradeables, accompanied by measures to liberalize economies so that factors of production can seek their highest returns" (Seralgeldin, 1989, p. 5). Thus the strategies advocated are typically export led, aiming above all to increase export earnings in order to foster economic growth and enable the repayment of outstanding debt. The main macroeconomic policies urged on countries are:

- devaluation, to encourage exports and discourage imports;

- trade liberalization, by removing controls on imports, foreign exchange, and foreign investment;

- price decontrol, to remove distortions that discriminate against export sectors and agriculture and protect inefficient industry. This includes decontrol of the price of labour. The intention is to increase the efficiency of industry by ensuring appropriate prices for the factors of production and by shedding labour where overmanning has occurred, thus enhancing domestic and export competitiveness. In addition, price subsidies are to be removed, inter alia to remove the bias in favour of urban consumers;

- strict control of money supply and credit expansion, by increasing real interest rates. The latter is also intended to fight inflation, promote saving, and allocate investment capital to the highest bidders;

- tax reform to replace taxes on production with taxes on consumption and value added.

(2) "[S]treamlining the public sector... while directing scarce public financial resources toward the provision of basic infrastructure; supporting vital economic services; and developing human resources... and freeing of domestic resources for private sector investment and production" (Seralgeldin, 1989, p. 5). This is to be achieved by:

- reduction of public expenditure. Although the World Bank holds that this can be achieved by improving revenue generation, in practice streamlining of the civil service is required, based on reducing the number of civil servants employed and freezing wages;

- privatization of government enterprises and parastatals, to reduce the size of the public sector and increase efficiency; reorientation of public sector investment towards fostering productive sectors, especially agriculture, and providing basic services in education and health. In a climate of general expenditure cutbacks, the provision of basic services is to be accompanied by the introduction or raising of user charges and the removal of subsidies.

(3) "[A] comprehensive restructuring of external debt with the objective of relieving the resource constraint" (Seralgeldin, 1989, p. 5); restructuring debt (rather than writing it off, although some write-offs may be negotiated) and improving repayment are intended to encourage new flows of aid and commercial investment.

To achieve these aims, macroeconomic policies aimed at removing distortions in markets are needed, in particular to favour trade, as well as sectoral policies to encourage the development of productive sectors and resources, and policies to improve resource mobilization, both private investment (domestic and foreign) and public resources, by a balance between improved revenue generation and public expenditure cut-backs. By the end of the 1980s, the Bank's emphasis had shifted from stabilization and growth to place rather more emphasis on equity, stressing the need for programmes to foster the participation of the poor in the process of economic growth, by improving their access to jobs, income-generating assets, and basic services, and to "provide for action programs to avoid social distress in the short term" (Seralgeldin, 1989, p. 7). The latter, generally termed the "social aspects of adjustment," are compensatory transitional provisions to protect vulnerable groups' access to services and to compensate redundant workers and enable them to make a new start (Ribe et al., 1990). SAPs have continued to evolve in response to both donor and country experience and the negotiations between them as programmes are extended (Green and Faber, 1994).

Despite the difficult circumstances in which, it is acknowledged, most countries have adopted and pursued SAPs (fluctuating and/or declining terms of trade, high real interest rates exacerbating the debt burden, and droughts), Seralgeldin detected a number of positive experiences in the first two-thirds of the 1980s. Between 1980 and 1985, in 20 countries relatively little affected by unusual weather or external shocks, World Bank figures suggested that real GDP growth rates (allowing for inflation and population growth) of 1.5 per cent per annum on average were experienced. However, in 1986/87, GDP growth more than doubled to about 4 per cent per annum in countries that were considered to have had sustained adjustment programmes, whereas it fell by half in non-adjusting countries. In addition, in adjusting countries inflation and the fiscal deficit are said to have been reduced, and real interest rates to have become positive (Seralgeldin, 1989, pp. 3-4).

Not all assessments are so positive, and some reach contrary conclusions. Mosley and Weeks (1993) concluded that, contrary to World Bank assertions, Africa as a whole did not recover after 1985, although there was some improvement for some countries; countries with Bank-supported structural adjustment did no better than those without; and "there may have been no significant difference between the economic performances of 'strong' and 'weak' adjusters that can be attributed to structural adjustment packages" (p. 1588). Difficulties in assessing the outcomes arise from methodological problems, differences in definitions, and disagreement over the relative roles of domestic policy reform and exogenous shocks or the effects of war and civil disorder (Green, 1993). As a result, adjustment measures have not been disaggregated and assessed in a way that is useful to improving policy. In partial recognition of these and other criticisms, the World Bank's most recent assessment (World Bank, 1994) is more measured in its conclusions." It notes that there has been more progress with liberalizing trade than with reforms in the agricultural, financial, and public sectors, and, although it acknowledges external constraints on achieving economic growth, it attributes growth in per capita GDP, exports, and industrial and agricultural production in those countries that have achieved these primarily to far-reaching and consistent domestic policy reforms (see also O'Brien, 1994).

Killick, however, in an assessment of the outcome of IMF programmes, concludes that the effects of Fund programmes and their ability to influence macroeconomic policy (with the exception of exchange rates) are overrated. High programme failure rates and lack of clear evidence that objectives are being achieved throw doubt on the Fund's ability to operate in low-income countries. Although the programmes have, he acknowledges, raised awareness amongst some governments of the importance of financial discipline, the increased inflow of private capital expected to result has not, as noted above, materialized (Killick with Malik, 1992; ODI, 1993). Riddell (1992), Stewart et al. (1992), and Jesperson (1992) are even more critical:

Broadly speaking, stabilization achieved positive but modest results in Sub-Saharan Africa in the 1980s... [but] [w]ith few exceptions, stabilization was accompanied by sharp losses in GDP growth, investment and human capital development... The belt-tightening undertaken by most African countries would perhaps have been acceptable if adjustment had triggered the desired changes in economic structures and eventually led to expansion in food production, manufacturing activities and non-traditional exports. However, from this perspective also, the improvements realised in the 1980s were not satisfactory, despite profound reforms in privatisation, the liberalisation of prices and foreign trade and the mobilisation of external resources. (Jesperson, 1992, p. 14).

An analysis of 24 countries that initiated adjustment programmes in the 1980s showed that:

(i) capital accumulation slowed in 20 of the countries, owing to low rates of public and private (domestic and foreign) investment;

(ii) the share of manufacturing in GDP increased in only 6 of the countries (in 10, industrial output declined and in a further 9 it stagnated or grew very slowly; Stewart et al., 1992);

(iii) export volumes increased in only 11 countries, although even in these countries "the impact on the balance of payments was almost always negligible because of the fall in the export prices of primary commodities" (Jesperson, 1992, p. 14). In the other 13 countries, export volumes stagnated or diminished.

Overall, in SSA in the 1980s, only 12 countries (representing less than a fifth of the region's population) recorded positive rates of growth of per capita GDP. Per capita GDP fell in the remaining 21 countries, including many that had achieved growth in the previous 15 years (see also Stewart, 1991; Helleiner, 1992; Stoneman, 1993). Even including the more buoyant economies of north Africa, IMF figures show that, whereas real GDP growth was positive between 1973 and 1992 (increasing at about 3 per cent per annum in the 1970s and 2 per cent per annum in the 1980s), per capita real GDP was stagnant or negative in seven out of the eight years between 1983 and 1990 (Bird, 1993). There was little change in the overall picture in the first half of the 1990s. These conclusions conceal marked variations in implementation and outcomes, which themselves need careful explanations related to political and institutional as well as economic factors (Harvey, 1993).

SAPs have also had uneven impacts on economic groups within countries. It was hoped, for example, that increased producer prices and reduced food and services subsidies would benefit rural populations, especially small farmers, redressing the perceived anti-agriculture and anti-rural bias in government policy. In practice, these impacts have been more complex than anticipated by the multilateral agencies and in many cases strongly negative, despite half-hearted compensatory programmes and lip-service paid to integrating anti-poverty measures into SAPs (Gibbon, 1992). Many of the impacts have affected urban populations in particular and will be analysed in more detail below. An even more recent aspect of policy conditionality has been the association of political liberalization with economic reform. However, as has been shown in ODI (1994) and Simon (1995), the relationship between type of political system and economic policy is complex and the nature and significance of democratization imperfectly understood in either theoretical or practical terms.

Although policy reform was needed in African countries, their increased dependence on external assistance in the 1980s made them susceptible to uniform policy prescriptions, determined, despite apparent opportunities for negotiation, largely by outside agencies. Although the donors claim that there is now evidence of improved macroeconomic performance as a result, few others are convinced. Instead, the policy formulation hegemony of the colonial powers to which Africa was subjected in the first half of the twentieth century seems to have been substituted by a new hegemony which no less clearly has the interests of transnational capital and northern countries at heart (Bier, 1994). Such a hegemony, it is argued, is illustrated by the lack of influence of alternative agendas for reform, such as that of the UN Economic Commission for Africa (UNECA, 1989; Stewart et al., 1992; Parfitt, 1993; Simon, 1995).

Urban change and urban management

What, in the context of Africa's external dependence and vulnerability, marginalization and aid dependence, has happened to cities? What effects has the economic deterioration precipitated by the first oil price increases and reinforced by subsequent exogenous and endogenous forces had on the rate of urbanization, on the economic structure of cities, and on urban management? In the first part of this section, urban trends will be related to the internationalization of capital and their manifestation in Africa will be explored. Aid dependence, and especially the implementation of SAPs, have had particular implications for urban populations, in terms of both incomes and access to jobs and services. Finally, the donors, especially the UN, the World Bank, and the United States Agency for International Development (USAID), have adopted a series of policies related to lending for urban development which reveal congruences and contradictions with their wider policy stances.

Global cities and urbanization in Africa

One of the manifestations of the internationalization of capital has been the emergence of a hierarchy of cities with particular roles in the capitalist economic system. At the apex of the system are the so called "world cities," sites for the control and management of TNC operations, specialized business services to back these up, and nodes in the world banking and commercial system (Friedmann and Wolff, 1982; Sassen-Koob, 1985; Thrift, 1987; Sassen, 1994). Second to these global cities are regional or continental cities, which perform similar functions within the world capitalist system to global cities, but within a more restricted geographical region (Sit, 1993). LDC cities of most obvious regional significance are those in the newly industrialized countries (NICs), which are increasingly integrated into functional networks of economic linkages with global or core cities, especially in Asia, where TNC headquarters and R&D functions remain in Tokyo but investment occurs in a regional network of cities in East and South-East Asia. Within these regional networks, "growth triangles" have been identified based on complementarities across national boundaries and urban corridors based on mega-cities (Yeung and Lo, 1996). At the third level in the hierarchy are national cities, which are foci for national accumulation but also provide a location for transnational offices and operations, banks, and corporate services, and are thus linked into the world economic system (Sit, 1993; Simon, 1992, 1993).

Not unexpectedly, as Simon shows in earlier work (1992, 1993) and in chapter 3, given Africa's marginality to the world economic system, none of the world cities is located in the continent, although Cairo, Nairobi, and increasingly Johannesburg have regional roles. Africa is not even part of the semi-periphery, and the functional city systems linked across national borders that have begun to emerge in Asia are not evident in Africa except insofar as cities in the interior must use ports in other countries for trade. Most large African cities are centres of national economies, although they are connected to the world economy through the unequal trade, investment, and aid relationships analysed above.

Despite its marginal position in the world economy and the economic difficulties experienced more or less consistently since the mid-1970s, urbanization has continued (table 2.1). By 1990 it was estimated that a third of Africa's population was urban compared with a quarter in 1975. The rate of urban growth was about the same in the 1980s, it was suggested, as in the post-independence decade, running at just under 5 per cent per annum. As in the earlier period, rates of growth in the least urbanized regions (east and west Africa) were above the average for the continent as a whole, while those in the most urbanized regions (south and north Africa) were below (UN, 1993). However, there are major difficulties with these figures: of the 53 countries included, only two-thirds had had a census since 1980 (14 in the early 1980s and 17 since 1985); figures for 19 were based on censuses carried out in the 1970s and 3 in the 1960s. Not only the recency but also the reliability of the figures varies widely: of the five largest countries (20 million+ in 1980), only three have had reasonably recent semi-reliable censuses (Ethiopia, South Africa, and Egypt²), while Zaire appears to have had only one (1984) and Nigeria's last reliable one was in 1963, because there are some doubts even about the most recent (1991) census: these latter two countries account, on 1980 estimates, for 22 per cent of the continent's total population and 23 per cent of its urban population.

The inability of so many African countries to carry out regular and reliable censuses is a symptom of their poverty and underdevelopment, as well as of civil war and political instability. Much of the data on recent urbanization trends is, therefore, unreliable and many of the apparently precise figures are mere estimates, based on extrapolations of earlier trends. This is a theme that is developed in more detail by Simon and taken up again in later chapters, especially chapter 13. In addition, given the series of exogenous and endogenous shocks that have affected the development of most African countries, it is risky to make assumptions based on previous patterns.

By 1970 there were eight cities of 1 million+ in Africa, and by 1990 there were estimated to be 24, in 18 countries (UN, 1993). So far, the number of very large cities is limited. The emergence of Cairo in Egypt, as elaborated upon by Yousry and Aboul Atta in chapter 5, owes much to the country's geography (the limited stock of fertile, watered agricultural land, the significance of the Nile as a source of water and transport route), as well as to its long urban history and the economies of agglomeration which operate in any major city (Rondinelli, 1988). However, efforts to encourage urban and industrial growth elsewhere, especially in Alexandria and also to some extent in the new towns, have borne some fruit and the proportion of Egypt's urban population that lives in Cairo is expected to decline from a peak of 39 per cent in 1980 to 34 per cent by 2000. Lagos has never been as dominant in Nigeria as Cairo has in Egypt. It accounted for only 8.5 per cent of Nigeria's urban population in 1950, but has increased its dominance in succeeding decades and now is estimated to accommodate a fifth of the country's urban population (UN, 1993). From its colonial origins as a port and administrative centre, a history of concentrated public and private sector investment underlain by poor communications with the rest of the country, political and administrative dominance combined with a politics of patronage, and dependence on imported inputs for industry and construction has, as Abiodun shows in chapter 6, reinforced its early dominance. Beavon (chap. 5) describes the early mining and later manufacturing economic base of Johannesburg which, together with its services functions, accounts for its growth and agglomeration into the Pretoria-Witwatersrand-Vereeniging conurbation. Even in countries that do not have very large cities, the proportion of the urban population that lives in the largest city is high and increasing, while secondary cities and small urban centres continue to be underdeveloped (Rondinelli, 1988).

The relationship between economic and urban growth, as Simon shows in chapter 3, is complex and far from direct. Despite Africa's economic difficulties, many large cities have continued to grow - they are communications and transport hubs, and often the seats of government, and rates of natural increase are high despite the spread of AIDS. Their location as the place of residence for politicians, senior civil servants, and diplomats has helped to bias public expenditure on specialist health and other services and infrastructure towards them; this and the need for access to government offices has in turn made them the most likely locational choice for both transnational and domestic investment in manufacturing and for the offices of TNCs with mining or agricultural enterprises elsewhere in the country. Even though transnational investment in services in African countries is limited, what there is occurs in the largest cities, the fulcrums of flows of international capital, travel, and communications (Thrift, 1987). FDI is even more concentrated than domestic investment in the largest cities, because of executives" greater knowledge of the city economic environment and the presence of commercial facilities to intermediate between extraction/production and the international market where necessary (Sit, 1993). In 1988 over half of all urban wage employment in Kenya was in Nairobi (Simon, 1992, pp. 94-97; see chap. 9).

The locational attractions of the city are often magnified by policy with respect, for example, to transport tariffs, energy and service prices, and incentives for industrial development. Official policies to encourage decentralization are often counteracted by the spatial effects of non-spatial policies: Oberai (1993) quotes the example of Nigeria, where over 90 per cent of total net subsidies granted to industries benefited those located in Lagos.

The leading economic functions in the international and national economy have a multiplier effect, generating both formal and informal sector employment in wage and consumer goods and services industries. Despite the decreasing ability of the formal sector to absorb increased numbers in the labour force, as described by Rogerson in chapter 10, the cities continue to exert an attraction for migration. Many rural areas provide few economic opportunities for their growing populations, despite the pro-agriculture policy changes which have formed part of SAPs. Although the deterioration in infrastructure and services that has resulted from public expenditure cut-backs has been countrywide, rural areas started in a disadvantaged position. The chance of a better life in the cities, however much of a gamble, continues to attract migrants (Dogan and Kasarda, 1988). The evolving economic, social, political, and physical characteristics of these cities, marked as they are by so-called "informalization," are the subject of both the city case-studies and the thematic chapters in parts II and III of this volume.


Contents - Previous - Next