Revised Version

.

(Paper Prepared for Economic Commission for Africa, Addis Ababa, Ethiopia)

.

Alemayehu Geda

Institute of Social Studies (the Hague)

and KIPPRA (Nairobi)

.

.

FINANCING DEVELOPMETN IN AFRICA

TABLE OF CONTENTS

.

I. An Overview

This study is concerned with issues of financing development in Africa. The analysis commences by organizing major possible sources of financing development in the continent in to seven sub-sections of the study. First, the paper briefly reviews the recent growth trend and the daunting task of reducing poverty in the years to come. This is done by examining both the growth performance of the continent, the challenges of reducing poverty, as well as how this growth might be financed. The implications for financing (both from internal and external sources), drawn from this analysis, are pursed at length in the rest of the paper.

The brief overview of issues of poverty, growth and financing requirements sets the scene for the rest of the paper. That is, once the enormity of the task of reducing poverty and its huge resources implication is drawn, the next stage of the analysis is executed under two broad themes:

• What is the nature and source of such finance (ODA, FDI and Other Capital flows, Domestic resource mobilization) and,

• The financing implication of addressing finance related problems such as issues of Capital flight, External Debt, and the International Financial Architecture.

1. The rest of the study is organized as follows. Section three examines issue of ODA both from theoretical and empirical perspective. Section four deals with foreign direct investment and other capital flows such as portfolio, banks, equity and bonds. Section five examines the role of domestic financing. Section six deals with the issue of capital flight and the implication of its reversal for financing development. In a similar way, section seven examines the issue of debt and the possible effect of addressing it for financing development. The paper concludes by briefly examining the implications of the new (international) financial architecture for developing countries such as those in Africa. Finally, an attempt to draw short to medium term policy and research implications is made.

.

II Recent Economic Performance And Future Challenges: The Implication For Financing Development

1. In sharp contrast to 1970s and 1980s, the mid-1990s witnessed a growing optimism about Africa’s economic future. Real GDP growth of more than 4 percent, exceeding the rapid population growth, compared to nearly 1 percent growth in the first half of the 1990s, the doubling of the growth of exports, from around 4 percent in the mid 1990s to nearly 8 percent in the second half of the 1990s (See Appendix I), growing emphasis on appropriate policy reforms and the new political discourse of the ‘African renaissance’, associated with political reforms, may partly explains this optimism.

2. Such optimism was not fully embraced by African policy makers, governments and the academics at large, however. This is because the enormity of the challenge facing the continent is apparent from the grim social data they are confronted with. This is in particular true of poverty in the continent. By the end of the 1990s even the macro indicators started to slide down. The recovery in the per capita real GDP growth that started from -2.9 percent in 1992 and picked to 2.6 percent in 1996 has decelerated to 0.4 percent in 1998 (See Table 1.2 in Appendix I). Exports as the share of GDP that rose form 27 percent in 1990 to 30 percent in 1995 started to slide to 27 percent in 1998 (See Table 2.2 in Appendix I). This unhealthy trend is rather vivid when one looks at the social data.

3. According to ECA’s study, which is based on household surveys of 20 countries that constitute 60 percent of the continent’s population and 76 percent of its GDP, the continent is characterized by (a) a high degree of inequality of income (with a Gini coefficient 0.44). The highest inequality is fund in South Africa and Kenya (58 percent) and the lowest in Egypt (32 percent); similarly the bottom 20 percent of the population received only 5 percent of the income (the top 20 percent getting 50 percent); (b) about 44 percent of the population is leaving below the poverty line of $39 per person per month. This incidence being 22 percent and 51 percent in North and Sub-Saharan Africa, respectively (the highest incidence is found in Guinea-Bissau, 70 percent, and the lowest, 15 percent, in Algeria). In gross terms the average income of the poor for the continent as a whole is found to be only 83 US cents per person per day (this figure being $1.5 for North Africa and 67 US cents for Sub-Saharan Africa). In general the average incidence of poverty for the 32 sample countries is 46.7 percent; the poverty gap being 27.3 percent (ECA 1999).

4. There is noticeable regional variation in the state of poverty. North African countries are in a better shape having an average score of 21.4 percent and 5.5 percent for incidence of poverty and poverty gap, respectively. West African countries follow this, with a corresponding value of 48.6 percent and 18.8 percent, respectively. Eastern Africa countries have registered the third rank with incidence of poverty of 50.7 percent and poverty gap of 18.3 percent. Although the sample from Central Africa has only two countries, it has the most severe state of poverty – a head count ratio of 58 percent with a poverty gap of 29 percent. This is followed by Southern Africa with incidence of 57.5 percent and a poverty gap of 27 percent.

5. Looking the issue in a broader perspective, the main conclusion that can be drawn from evaluation of recent economic performance and sustainability in the continent, based on ECA’s African Economic Report (ECA 1999), is that there is a clear difference between performance and sustainability. The study noted only three African countries (constituting about 6 percent of the population) are found to be good enough to sustain growth and development. Twelve countries (having nearly 25 percent of the population) are also found in the ‘good’ category. Thus, it is reasonable to conclude that African countries in general are not in a good shape in terms of overall performance and, in particular, its sustainability. This underscores the need to conduct an in-depth study of the growth process in the continent so as to understand constraints to growth as well as the sustainability of growth attained.

6. In fact, the second half of the 1990s witnessed a proliferation of what can loosely be termed as ‘growth literature’ that focused on Africa. These studies range from those that attribute the slow growth to geographical and related factors (Bloom and Sachs 1998), as well as prevalence of social tension-cum-external vulnerability (Rodrik 1998, Easterly and Levin 1997) to micro-based explanation that are blamed to hinder proper market operation by making economic activities risky and costly (Collier and Gunning 1999). The proliferation of this literature, as aptly noted by Azam and others (1999), although does not provide hard and fast rule to revitalized growth, suggests important lessons for the future. After reviewing this literature, Azam and others (1999) stressed the importance of addressing structural problems such as lack of social capital and deficient political institutions.

7. The discussion above clearly shows the enormity of the challenge facing the continent. The relevant question is what would it take to address these challenges, and most importantly, what is its resource implication. In order to answer these questions this paper has followed two approaches. In the first approach, the level of investment is compared with the level of domestic saving to arrive at the trend of the current level of the resource gap (financing requirement) from a macro perspective. In the second part the challenge of reducing poverty is converted in to its growth rate equivalent and the financing needed thereof. This will be based on the scenario drawn by Amoako and Ali (1998).

8. For Sub-Saharan Africa (SSA) excluding S. Africa, the share of investment in GDP has increased from 15.8 percent in 1990 to 18.8 and 19.5 percent in 1995 and 1998, respectively. This figure is relatively high for the North Africa region that had a ratio of 28.2 percent in 1990, 23.1 percent in 1995 and 22.7 percent in 1998. Although these figures are higher than the ones for SSA, they do show a declining trend. (See Tables 2.3 in Appendix I for details). East, Southern and West African regions broadly followed the SSA trend. Although the trend in SSA and the actual ratio of North Africa are good, investment in both regions is way below the level of domestic resources.

9. Gross domestic saving in SSA (excluding South Africa) has declined from 17.4 percent recorded in 1990 to 13.5 percent in 1995 and picked up again to 17.2 percent in 1996 only to decline to 12.6 percent in 1998. The corresponding figures for North Africa are 21.9, 18.1 and 18.9 percent in 1990, 1995 and 1998, respectively. These figures are far below the 30 plus percent domestic saving rate observed in newly industrializing countries of East Asia. What is worrisome in Africa is not only the low level of the absolute figures but also their oscillating and, more often, declining trend (See Tables 2.4 in Appendix I for details).

10. A simple computation of resource gap following the above two paragraph shows that the external finance requirement in SSA (excluding S. Africa), which had been negligible in the early 1990 rose to 5 and 7 percent of GDP in 1995 and 1998, respectively. The Corresponding figure for North Africa stands at 5 and 4 percent in 1995 and 1998, respectively. Taking the GDP of the regions in 1998, the 1998 ratios indicate an annual external resource gaps of around 20 billions of US$ for Africa. Although these figures point to a very high level of dependence on external finance, the needs of Africa are much higher when the continent’s objective of reducing poverty is taken on board (See Tables 2.3 and 2.4 in Appendix I for details).

11. Apart from such simple computation, there had also been various attempts to estimate the resource requirement of the continent. Almost all such studies are based on the typical Harrod-Dommar set up and usually come up with reasonable estimates. Recently, Easterly (1997) has noted various pitfalls of this approach. However, using some regression results of ‘the Easterly type’ for a sample of African countries, Amoako and Ali (1998) argue that still sensible projections can be made using this approach. In the paragraph below we have reported the result of projection that is made by Amoako and Ali (1998). We have used this estimate for various reasons. First, they have attempted to address some of Easterly’s concerns in the African context; second, they have set up the average and the best performance by taking the actual experience of African countries (such as efficiency of capital use in Uganda). Finally, their projection for the period 2000 – 2005 embraces the period under analysis.

12. Amoako and Ali’s (1998) review of the literature on estimates of expected transfers to Africa shows that over the period 1986-92 an annual inflow of US$ 9 billion over and above the previous level of ODA were expected in the context of UN Program of Action for Africa’s Economic Recovery and Development (UN-PAIRED). This had failed and a new initiative called UN Agenda for Development of Africa (UN-NADAF) estimated a minimum of US$ 30 billion in net ODA in 1992; moreover, this ODA is stipulated to grow by 4 percent in real terms. Although specific estimates are not explicitly made such requirement is also noted in the context of the Copenhagen Declaration. Estimates of resource needs are also made by ADB (1995) and ECA (1993) (See Amoako and Ali 1998).

13. After reviewing such estimates Amoako and Ali (1998) came up with their own estimates. For the purpose of estimation they have set the objective of reducing poverty by half by 2015. Using a general measure of poverty and population growth rate, such an objective can be attained if GDP could grow by annual figure of 8 percent (World Bank and others 2000 put this at 7 percent). This is combined with an average ICOR (7.8) for the sample of countries in their studies. It is further assumed that each country will approach the most efficient one (Uganda with ICOR of 2.5) at the end of the planning horizon. They have also assumed an initial domestic savings rate of 16.1 (observed in the 1990s and noted above) and that this rate is assumed to increase to the feasible level of 23.5 percent by the year 2015.

14. On the basis of the above assumptions they have reported estimates of resource requirement for eight countries in their sample and for SSA (including and excluding S. Africa and Nigeria). This result is given in the Summary table below. The result shows a trend of graduating from aid dependency. This, however, is the result of the authors’ assumption of rising level of savings and efficiency of capital use over the planning horizon. This figure will be much higher if World Bank and others’ (2000) estimated cost of combating HIV/AIDS (estimated to be 1 to 2 percent of GDP) is factored in. This has the strait foreword implication for policy. One way to minimize external finance requirement and meet the objective of reducing poverty is raising domestic saving and efficiency of capital use.

External Resource Requirement (Percentage of GDP, Annual Average)

	1998	1999-2000	2001-2005	2006-2010	2011-2015
SSA	52.4	46.8	32.1	10.0	-7.1
SSA1	48.4	40.0	26.1	8.2	-7.1
SSA2	55.3	47.3	30.4	10.1	-6.1

Source: Amoako and Ali(1998).

Note: SSA1=SSA excluding S. Africa, and SSA2=SSA excluding S. Africa and Nigeria

.

III Official Development Assistance

1. The analysis in section II shows not only the huge external finance requirement of Africa but also the frustration of various initiatives aimed at effecting such transfers to the continent. This section examines the flow of ODA and their future contributions in financing development in Africa. This is done by examining both its current trend as well as its theoretical and empirical determinants.

2. Relative to donor GDP, net disbursement of ODA have dropped almost 30 percent percent in real terms (O’Connell and Soludo cited in World Bank 2000). ODA is also increasingly shifting in composition (towards humanitarian assistances) and facing competition from Eastern Europe. In general, net transfer per capita has fallen sharply form $32 in 1990 to $19 in 1998 mainly because of Africa’s less strategic significances and donor fatigues. However, this net transfer itself is largely being offset by the terms of trade loss. Cumulative terms of trade loss for SSA (excluding S. Africa) between 1970-97 represent a staggering level of 120 percent of GDP (World Bank and others 2000). For Sub-Saharan Africa (excluding South Africa) the recent trend shows that net ODA as the share of the recipients GDP has been 9.4 , 12.1 percent in 1990 and 1995, respectively, and declined to 7 percent in 1997. The corresponding figure for the whole of Africa were 5.5, 4.4 and 3.3 percent in 1990, 1995 and 1997, respectively (See Table 3.1 in Appendix I).

3. Notwithstanding this recent trend, official capital flows represent an important component of financial flows to African countries. Both the theoretical and empirical literature about such flows attempts to answer ‘what determines official capital flows to the South?’ and ‘why such flows?’ The search for an answer to these questions leads one to different, sometimes conflicting, theoretical explanations. Thus, one school of thought maintains that official capital flows are determined by the economic and geo-political interests of donors. Indeed, this suggestion finds support in a number of studies (See, for example, Mikesell, 1968; OECD, 1985; Mosley, 1985; Ruttan, 1992; McGllivary and White, 1993). Another major explanation for aid flows relates to humanitarian or developmental considerations (studies supporting this viewpoint include Streeten, 1976, cited in Gasper 1992; Riddell, 1987; and the ‘aid as a public good’ literature of Mosley, 1985; Dudley and Montmarquette, 1976; and Frey, 1984). A number of studies investigate one or both of these explanations empirically. Indeed, Beenstock (1989), Mosley (1985), and White and McGllivary (1992, 1993) have gone so far as to portray these empirically based studies as representing a distinct approach, devoid of theory. However, we prefer to view these simply as empirical manifestations of the theoretical explanations noted above.

4. In one of the earliest studies, McKinley and Little (1979) develop a ‘recipient need and donor interest model’ which sets out to examine the ‘humanitarian’ versus ‘national (donor) security’ explanations for aid flows. According to this ‘recipient need model’, aid should be allocated in proportion to the economic and welfare needs of the recipient, otherwise, such aid is simply fulfilling the donor’s interests, either as a commitment, or as leverage strategy. After discussing, at length, how these propositions might be operationalized, McKinley and Little conclude that the recipient need argument is likely to be a secondary one. However, while this paper focuses on bilateral issues, syndicated efforts are more widely followed, today, by international and regional organizations in their effort to realize donor interests. In this study, we do not examine this phenomenon. However, we would refer the interested reader to Anyadike-Danes and Anyadike-Danes’ (1992) review of evidence relating to European Community (EC) aid to the African, Caribbean and Pacific (ACP) region. They conclude that aid to ACP countries is strongly associated with the Pre-Lome association.

5. Moreover, most econometric studies of recipient need models are not robust either. Having surveyed such models, White and McGillivray (1993) note that the separate estimation of recipient need/donor interest models suffers from specification error due to the omission of relevant variables, which are usually not orthogonal (i.e., correlated with all included variables). This, they suggest, leads to OLS estimates with bias. They argue that this problem is inherent in the very methodology of this approach. White and McGillivray illustrate how different results may be obtained if one allows for correction of such specification errors (White and McGillivray, 1993: 36-41). To this, one might add the observation that such time series studies might as well suffer from spurious (non-sense) regression.

6. In a similar study of what determines total aid volumes, Beenstock (1980) starts from the assumption that political factors affect the geographical distribution and not the total volume of aid. He points out that, whatever the objective of aid, its volume is constrained by GNP (or GNP per capita), the balance of payment, levels of unemployment and the size of net budget surplus of the donor state. At a statistical level, he found all signs as expected and, with the exception of the budget term, all are significant at a 5 per cent significance level (Beenstock, 1980:142). Using a time trend, Beenstock suggested a tendency for ODA to increase over time. Although Beenstock’s analysis focuses on the supply side of the issue, his analysis does not explain the central reason for aid. Further, it is quite difficult to envisage a situation where politics is used solely in allocation, and not also in the determination of total supply.

7. A summary of the findings of a recent survey of the literature on allocation of aid, undertaken by White and McGillivray (1992, 1993) can throw light on the evidence from the existing body of knowledge. White and McGillivray adopt two broad classification schemes. Firstly, descriptive measures, which are evaluative in their nature, while measuring donor performance (White and McGillivray, 1992:1). And, secondly, explanatory studies which trace their origin to political-economy theories, and base their explanation on ‘political, strategic, commercial and (albeit often begrudgingly) humanitarian motives (White and McGillivray, 1993:2). These surveys raise a number of important issues. They conclude that models should approximate the actual practice of aid determination process. Perhaps more importantly, these surveys also highlight how aid allocation is the outcome of a bureaucratic decision making process, economic, political and other relations between the donor and the recipient (White and McGillivray, 1993:68). We have empirically examined this line of thought in the African context.

8. Although the results of this empirical analysis are not robust and need further research, because of the relatively short time series data used, nevertheless they do provide a useful second best indicator of the determinants of official flows to Africa. In any case, the use of results obtained in the context of this study, no matter how flawed, is preferable to using spurious results arising from existing literature. Besides, a regionally based estimation for Africa is not available. In relation to the actual results of these estimations, these may be summed up as follows. Firstly, there exists a long-run equilibrium relationship between Africa’s relative economic performance and flows of capital to that continent. Secondly, humanitarian and developmental considerations are found to be largely negligible in influencing such flows to Africa. This result would tend to lend support to the apparently obvious notion that capital flows are associated with the level of development, and particularly involvement in trade. Indeed, this remains true even when these flows chiefly comprise aid (See Appendix II, Technical Appendix 3.1 for detail).

9. The major short to medium policy implication of this result is straightforward. Instead of expecting aid from pure humanitarian consideration, identifying the geo-political and strategic interest of donors and acting on them is crucial. Moreover, growth oriented polices might go hand in hand with increased flows. On the second point understanding the bureaucratic/budgetary process of donors while handling aid is crucial for influencing flows to the continent. This conclusion shouldn’t deter countries from vigorously raising external resources through the customary channels such those initiatives outlined at the beginning of this section (See Section IX for detail).

10. Once such flows are secured the next most important task is to enhance aid effectiveness and design ways of graduating from aid-dependency. The main culprit, apart form terms of trade loss, behind aid effectiveness in Africa is the extreme low level of growth which is predominately explained by huge setback in (investment) productivity. According to the World Bank and others (2000) the ICOR in Africa is half that in Asia in 1970-97, this investment productivity decelerated from 25 to 5 percent (and GDP growth from 5 to 1 percent) from early 1970s to now. It is curious to note that the growth recovery since 1994 has relied on productivity gains rather than an increase in investment as such. Thus this failure in growth/productivity combined with the oil price shock and terms of trade loss led to sharp increase in Aid, a cumulative transfer of 178 percent GDP (1970-97). But the increase after 1970-73 (125 percent of GDP) was a little more than the terms of trade loss. Moreover, by 1997 external debt equals to GDP for most countries, see section VI below (World Bank and Others 2000).

11. The sharp increase in aid may also have long run detrimental effects depending on the initial institutional set-up of the recipient countries as well as the emphasis given to that in the design of foreign aid strategy. Azam and others (1999) noted such outcome as symptoms of aid dependence. They argue that when institutions are already weak aid may lead to the collapse of such institutions. The policy implication of these authors’ analysis is quite striking: ‘foreign aid strategies, even for countries with similar per capital incomes, should be differentiated according to there institutional capacity’. In the African context, where institutions are both weak and vary across countries, graduating from aid dependency requires an examination of not only the short run external financing of resource gap but also the long run possible negative impact of such flows.

12. Apart from the productivity issue, aid effectiveness is also a major problem. It is estimated that a typical poor country receives 9 percent of its GDP but the poorest quintile of the population consumes only about 4 percent of the GDP (World Bank and Others 2000). Many factors are given for such failure of aid to address poverty (aid ineffectiveness). This list includes, according to World Bank and others (2000): support provided to ‘trusted allies’ even when they pursue poor policy, donor preference on aid objective and delivery mechanisms that usually face accountability problems as well as debt overhang.

13. The way ahead for aid effectiveness and transiting from aid dependency, according to a recent study (See World Bank and others 2000) requires, inter alia, recipients ownership of policies and programs, transferring accountability to Africans and strengthening the institutions to run aid by themselves, emphasizing capacity building as opposed to technical assistance, emphasizing on sustainability (in particular by working through national budget) and transparency as well as focusing on regional rather than country programs per se are few among many (See Section IX for detail). In general it can safely be concluded that aid could be important in good policy environment but can be even harmful in bad policy environment (See Ali and others, 1999)

.

IV Foreign Direct Investment And Other Private Capital Flows

.

Foreign Direct Investment (FDI)

1. Africa’s share of world FDI is extremely low. It was mere US$ 5.5 billion in 1996, representing only 1.5 percent of the global investment flows. The distribution of this flow is extremely skewed, with the main recipients being Nigeria, Egypt, Morocco, Tunisia, South Africa, Algeria, Angola, Ghana and Cote d’Ivoire who accounted over 67 percent of FDI to Africa in 1996. Between 1991 and 1996 ten countries (Nigeria, Morocco, Tunisia, Angola, South Africa, Ghana, Tanzania, Namibia, Uganda and Zambia) received almost 90 percent of such flows with Nigeria alone absorbing a third of this flows. The main sources countries being France, UK, Germany and US while the favorite sector are oil, gas, metals and other extractive industries (ADB, 1998). In general, by the second half of the 1990s, the average share of FDI in GDP was not only very small but also was declining. Whenever it has a positive trend it is largely related to investment in countries with new resource discovery.

2. Recently, however, there is a surge of FDI in some countries (Kasekende others 1995, Fernandez-Arias and Montiel 1996, Bhinda and others 1999). For all Africa the share of FDI in GDP, which was 0.29 percent (US$ 1.3 billion) in 1990 has increased to 0.56 percent (US$2.7 billion) in 1995 and jumped to 1.2 percent (US$ 6.3) in 1998. The comparable figure for SSA, excluding South Africa, during this period has been 0.41 (US$0.76 billion), 1.61 (US$ 2.7 billion) and 2.4 percent (US$ 4.8 billion), respectively (See Tables 4.1 and 4.2 in Appendix I). Similar trend is observed when individual country data is used (See Table 4.3 in Appendix I).

3. An exploration of the literature on the determinants of FDI leaves much to be desired. Although FDI, as discussed above, does not comprise a major component of external flows to low income countries of Africa, nevertheless, this section will briefly summarize the major canons of the theories of the determinants of foreign direct investment, in the hope that this may help to explain why this type of investment has not been so important in the continent and why there is a surge in the recent past.

4. The early neoclassical approach, summarized in the classic article by MacDougal (1960), hypothesized that capital flows across countries are governed by differential rates of return. The MacDougal model assumes perfect competition, risk free capital movement, mobility in factors of production and no risk of default. The portfolio approach to FDI, presented in reaction to The MacDougal model, emphasizes not only return differential, but also risk (Iversen, 1935 and Tobin, 1958, both cited in Agarwal, 1980). This is strengthened by a theory that emphasizes the positive relationship between FDI and output (sales in host country), along the lines of Jorgenson’s (1963) model (see Agarwal, 1980).

5. A major criticism of these theories relates to the question of perfection in markets. Hymer (1960, published in 1976) and Kindleberger (1969) argue that, if foreign firms were able to compete and succeed in the host country, then they must be in possession of a specific and transferable competitive advantage, both over local firms, and other potential entrants into the local market. This analysis also focuses on the micro foundations of FDI, by moving from a simple capital movement/ portfolio theory to a broader production and industrial organizational theory. Indeed, this school of thought has formed the basis for a whole strand of the literature. According to this line of thinking, some advantages of the competitive foreign firm include cheaper sources of financing, the use of brand names and patent rights, technological, marketing and managerial skills, economies of scale, and, entry and exit barriers (Kindleberger, 1969; Agarwal 1980). A related micro-based theory of FDI has also emerged with the development of the Vernon’s product cycle theory (Vernon, 1966) and its extension in Krugman (1979).

6. A second wave of refinements to the neoclassical capital movement/ portfolio theory of FDI, building upon Hymer’s original contribution, came with the emergence of explanations based on the ideas of ‘international firm’ and ‘industrial organization’. The fact that decision-making about foreign direct investment (FDI) takes place within the context of oligopolistic firm structures, - and that such investment includes a package of other inputs, such as intermediate imports and capital flows, - has led to the development of alternative explanations grounded in the theory of industrial organization (see Agarwal, 1980; Helleiner 1989:1452; Dunning, 1993). This approach attempts to understand flows of FDI by multinational firms desire to minimize transaction costs, a la Coase (1937), to tackle risk and uncertainty, increase control and market power, achieve economies of scale, and ensure advantageous transfer pricing (Hymer, 1976; Buckley and Casson, 1976, Lall 1973).

7. The recent works of Dunning (1993), which he terms the ‘eclectic paradigm’, represents a culmination of this trend towards a refinement of theories of FDI. Without departing much from the Heckscher-Ohlin-Samuelson theory of trade, in explaining spatial distribution of multinational firms, Dunning’s paradigm summarizes this strand of theory under an ‘ownership-specific, location and internalization’ (OLI) framework (see Dunning, 1993). Helleiner (1989) notes that ‘this "eclectic" theory of FDI, drawing on firm-specific attributes, location and internalization advantages - is widely accepted’ (Helleiner, 1989: 1253).

8. In sum, the theory of determinants of FDI covers a range of explanations: the pure capital movement, product cycle, industrial organization, the stagnation thesis as well as other political consideration. In the African context, the pure capital theory does not work since the assumptions simply do not hold. Neither is Krugman’s hypothesis workable, since it is more relevant to countries with a good industrial base and infrastructure such as East Asia. The deterioration in terms of trade, combined with the debt crisis, will greatly undermine the relevance of this theory, in explaining the African context. The most relevant theoretical explanation seems to be found in ‘industrial organization’ and the ‘international firm’ - ‘eclectic’ explanations. More importantly, the concentration of Multinational Corporations in the mining sector of most African countries and, to a good degree, the importance of the colonial history in determining their spatial pattern might be taken as lending support to the importance of the ‘eclectic’ approach.

9. An empirical assessment of the determinants of FDI in Africa using this theoretical insight reveals that, in general, relative market size, mining activity and the historical pattern of FDI together determine the flow of FDI to Africa. (See Technical Appendix 4.1 for econometric results). The importance of some of these factors in explaining FDI flows is also shown in the descriptive analysis of Bhattacharya and others (1997). Bhattacharya and others (1997) grouped the African FDI recipients into three categories: (a) countries which are long term recipients (Botswana, Mauritius, Seychelles, Swaziland and Zambia), (b) countries that recorded large increase in the 1990s (Angola, Cameroon, Gabon, Ghana, Guinea, Lesotho, Madagascar, Mozambique, Namibia, Nigerian and Zimbabwe) and finally (c) countries that have low and/or declining level of FDI but with encouraging turnaround, the best example being Uganda.

10. The main conclusion that can be drawn from the existing body of knowledge (both theoretical and empirical) is that much of the preconditions for sustained flow of FDI to Africa, in the existing global economic framework, relies on the structural transformation of the African economies which will have a positive effect on market size, resource discovery and enabling conditions for high level of growth. Thus, influencing FDI flows using conducive monetary and fiscal polices is important if a surge in flows of FDI in the short run is sought. In the medium to long run, however, bringing structural change in the economy by growth enhancing and growth enabling policies (which could be both market and non-market in nature) as well as joint-venture based exploitation of resources is an important area that need to be looked at (See Section IX for detail).

11. An in-depth study on sources and destination of FDI using country case studies from micro perspective, and a cross-country analysis using time-series macro data need also to be taken as an agenda for future research.

.

.

Other Private Capital Flows

1. Apart from FDI, other private capital flows such as portfolio flows, bank flows and bonds are also important in financing development. Notwithstanding an extreme inconsistency of data obtained from various international data source (See Appendix I, Table 4.3), the available data shows that such flows do not constitute important flows to the continent. This fairly accords with the popular perception about such flows to the continent.

2. What is interesting in examining such private capital flows to Africa is their trend (as opposed to their sheer magnitude). At the end of the 1970s and early 1980s there was a surge of private capital flows (FDI, private equity flows and private loans, the later in turn comprising bank, bond and other flows), SSA accounting for 8.9 percent of total private flows to developing countries. It accounted only for 1.6 percent of such flows in the period 1990-95. This sharp fall is chiefly attributed to the sharp deceleration in private loans starting from mid 1980s (See Bhattacharya and others 1997).

3. This trend and perception is changing in the recent past, however. A recent study by Bhinda and others (1999) came up with the interesting observation that international data sets were not tracking the sharp increase in private flows to Africa. This data and the inconsistency problem are reproduced in Appendix I, Table 4.3. The case study countries for this study include South Africa, Zambia, Tanzania, Uganda and Zimbabwe. South Africa has received higher than all four countries taken together (90 percent of total SSA since 1992) in absolute terms. However, relative to GDP, the other countries have levels (10 - 15 percent) as high as the fastest growing Southeast Asian and Latin America countries, while South Africa received only 4 percent (Bhinda and others 1999).

4. Portfolio equity flows: two important categories of ‘other private flows’ (excluding FDI) are portfolio equity and bank flows. Portfolio equity flows, though insignificant in magnitudes (except in South Africa), are growing in recent past. From 1994 to 1997 more than 12 African-oriented funds have been setup with a total size of more than US$ 1 billion. The operation of these funds is expanding from the initial focus on South Africa to Botswana, Cote d’Ivoire, Ghana, Kenya, Mauritius, Zambia and Zimbabwe (Bhattacharya and others 1997)

5. According to Bhinda and others (1999) study, this surge in portfolio flows to Africa is highly inaccurate if one follows the international data. This is because, first, they fail to reflect data from stock exchange and foreign participation in primary and secondary markets such as in Zambia and Zimbabwe; second, there is huge underestimation of inflows through equity funds. Recently, there are three important equity funds with SSA exposure: (a) ‘Pan-African funds’ with an exposure of US$ 692.9 millions, (b) South African dedicated funds with an exposure of US$ 8.057 billion and (c) emerging market global funds with an exposure of US1.5 to 3.5 billion (4 to 10 percent of world total is in SSA). Thus, the total SSA portfolio investment stock ranges from US$ 10.3 to 12.3 billion since 1995. When we look closer at the Pan African funds around 18 percent of it is invested in North Africa, 70 percent in SSA - SSA excluding South Africa having 35.1 percent (See Bhinda et al 1999 for details).

6. Bank flows: again the different international sources do conflict on the size of such flows to Africa (See Appendix I, Table 4.3). All the sources, however, show that the magnitude is not only very small but also erratic. When country data is used for sample of countries, this erratic feature is still there, however, the actual magnitude is in general higher in the latter data set. For 1996 (for which there is comparable figure for all sample countries) for instance, country data shows that such flows in South Africa are 2.2 billions (as opposed to 1.8 billion in BIS, Bank of International Settlement, data); for Tanzania 27.9 million (as opposed to 13 million in BIS data); for Uganda –21 million (as opposed to 30 million in BIS); for Zambia 235.8 million (as opposed to 28 million using BIS data) and for Zimbabwe –8 million (as opposed to 35 million in BIS). In general all data show that such flows are not really picking except in South Africa and to some degree in Tanzania. This is partly attributed to the wish of lenders to change exposure following economic trends, the rapid increase of foreign exchange holding in Africa following financial sector liberalization, domestic financials sector problems such as debt overhang and domestic payment arrears as well as the perception of high country risk. These factors resulted either in the decline or increasingly very short-term nature of bank flows. Whenever such flows are increasing this is explained by the dominance of foreign banks in those African countries where this trend is observed (See Bhinda and others 1999 for details).

7. Bond flows: in terms of magnitude such flows are also extremely unimportant in SSA. Such flows in 1996 were 1.2 billion using IMF data and 586 billion using World Bank data (see Table 4.3 in Appendix I). This low and erratic nature is partly attributed to low credit rating of most African countries.

8. The important question relevant for policy is to understand the determinants of such flows. In international finance literature there are a number of systemic explanations for bank and portfolio flows. These include static/dynamic capital movement theories (MacDougall, 1960; Bardhan, 1967), the portfolio theory of capital movement (See Williamson, 1983: 182-192; Solink, 1974) and theories of credit rationing (See Sachs, 1984; and Stiglitz and Weiss, 1981;for an empirical study of credit rationing see Feder and Uy, 1985 and Lee, 1993). All of these are essentially microeconomic explanations, tailored to a neoclassical framework. The first explains how capital is allocated across countries and generations in order to equate return differentials, with the aim of achieving optimal borrowing. The second approach adopts a portfolio choice method, owing to the emphasis it places on the evaluation of risk and uncertainty. However, the portfolio theory is essentially governed by the same motives as the first approach. The credit rationing theory builds on both of the above approaches, but also incorporates a constraint based on market imperfection.

9. A common problem in applying these approaches is their implicit belief in the workings of a competitive market, albeit with some reservations in the case of the credit rationing theory. A further weakness is the assumption that theories based on micro behaviour may readily be carried over to the macro/ international level. This theoretical underpinning greatly limits the relevance of these approaches for countries in Africa where risk, uncertainty, instability and market segmentation all represent significant factors, which need to be taken into account. Neither is it reasonable to assume that fully developed financial markets are available in all African countries. Indeed, most African countries are sovereign borrowers (with sovereign risk) and systemic explanation might be derived from such an approach. The important question then is what can we draw from highly limited ability of the theory to explain such flows and their erratic trend that we observe form the data?

10. Having such theoretical explanations it is also insightful to examine the determinants of such flows from empirical studies. There are various factors behind the recent surge of portfolio (the most important component being equity flows following by bonds, which in African context is taking a form of Treasury bills) flows. These determinant factors can be grouped as (a) global or push factors – the trend in OECD to invest in emerging markets and growing institutional (usually pension) investors faced with low interest rate and slow down in economic activity at home. For such investors SSA is found to be attractive because its yields have low correlation with other emerging markets; (b) perception of SSA by investors, which ranges from a perception of the region as ‘the final frontier’ to negative bias is also an important factor. This is largely determined by investors information about Africa; (c) there are also national factors such as political and macroeconomic stability, good governance, economic growth and regional integration, standardized regional structure of banks and developed stock markets with positive performance, as well as the existence of motivated labour force (d) finally, in particular, with non-equity flows (bonds and treasury bills) liberalization of the economies, possibility of holding dollar denominated accounts in banks and hence low risk nature of such flows, good credit rating, high domestic interest rate and development of capital markets are important factors for the recent upsurge (See Bhinda and others, 1999: 69-84). A recent empirical study using a dynamic modeling approach on the determinants of such flows in Asian and Latin American developing countries underscored that both global and country-specific factors have roughly the same level of significance in influencing such flows. However, bond flows are found to react predominantly to global factors while equity flows to country-specific factors (Taylor and Sarno, 1997). This finding is consistent with previous studies of Calvo and others (1993, 1996).

11. The policy implication of this should be apparent. Apart from designing appropriate policies that could positively influence the afro-mentioned factors, African policy makers need to understand the extreme volatility associated with such flows. This requires, capacity building on the management of the financial sector. Moreover, policy makers might be confronted with the policy dilemma of a trade-off between high flows owing to increased regional integration and high volatility. This is in particular true given the huge cost of bailing out countries when they are faced with financial crisis associated with such flows. As Bhinda and others (1999) rightly noted ‘Africa has no ‘big brother’ to facilitate a bailout in the first place’, this requires appropriate exchange rate policy, debt management, proper financial regulation and supervision and transparency (See Bhinda and others 1999 and Section IX below for detail).

.

V Domestic Resource Mobilization

1. The emphasis on domestic resource mobilization to bring development has been the preoccupation of economists drawn from the different strands of the profession. This ranges form Lewis’ classic assertion in early 1950s about the importance of savings (Lewis 1954) to the neoclassical growth models of the 1950s and 1960 and the recent endogenous growth models (See Romer 1986, Lucas 1988).

2. Main stream theories of saving such as the permanent income hypothesis as well as the life-cycle hypothesis that are based on the assumption that consumption is determined by life time resources that are directed at consumption smoothing, rather than current income, are increasingly being questioned in the context of developing countries. This is chiefly because the characteristics of households in developing countries are quite different from those in developed countries. These characteristics include poor and large household, the dominance of agriculture and uncertainty of income flows, demographic structure as well as binding liquidity constraints (See Deaton 1989). It is argued that savings in developing countries are largely explained by precautionary motive as well as by the need to finance investment as own finance dominates such economies (See Bridsall and others 1999, for instance).

3. Another important dimension widely cited in the developing countries context is the relationship between saving and economic growth. Most importantly, the direction of causality between saving and output growth is far from clear (Schmidt Hebel and others, 1996; Elbadawi and Mwega, 1998). In Sub-Saharan Africa, Elbadawi and Mwega (1998) argue that regardless of the direction of causation, focusing on policies that enhance private saving is important for at least two reasons. First, even if saving is the result and not the cause of economic growth, empirical evidence suggests that sustaining a high rate of growth requires a high level of accumulation of capital, which requires high level of saving. Second, due to Sub-Saharan African countries’ limited capacity of mobilizing external resource, raising national saving to maintain a high rate of investment and hence growth is essential.

4. Elbadawi and Mwega’s (1998) empirical study shows that saving rate significantly Granger-causes the investment rate although the relationship between saving rate and economic growth is non-significant. Other studies (Dooleey, Frankel and Mathieson 1987, Summers 1998, cited in Schmidt-Hebble et al 1996) find that there is a strong correlation between investment and saving both in developed and developing countries. However, growth found to Granger-cause both saving and investment (Elbadawi and Mwega, 1998). Similar results are also reported in the study by Carroll and Weil (1994). According to Deaton’s (1990) survey, the literature on household saving in LDCs has almost uniformly found that saving will increase with ‘permanent income’ (See Bhall 1979, 1980 for India; Musgrave for Latin America; Muellbauer 1982 for Sri Lanka; Berancours 1971 for Chile’ and Paxson 1989 for Thailand, all cited in Deaton 1990). The implication of this latter body of knowledge is that in order to raise domestic saving it is not only growth but also its sustainability that policy makers need to focus on.

5. External sector related factors are also important in determining the level of domestic saving. In the context of SSA, Elbadawi and Mwega’s (1998) work shows that the foreign aid ratio significantly Granger-causes a reduction in saving rate. Similar results are also reported in Alemayehu (2000). In SSA the Global Coalition for Africa (1993), claims a negative and significant effect of foreign aid on domestic saving. Commenting on such evidence Schmidt-Hebbel and others (1996) noted that the empirical results do widely vary depending on difference in sample, model specification, estimation method as well as the extent of fungibility of the foreign resources (See also Mwega, 1997: 208). This ‘aid-saving debate’ has been carried for nearly three decades and is still an unsettled issue (see White, 1992, for a comprehensive survey).

6. Another open-economy based variable widely used in the empirical literature is the terms of trade. Terms of trade are expected to have a positive effect on private saving, especially when it is transitory. This has supporting empirical evidence (See Ostry and Reinhart 1992, Bevan et al 1992, and Azam 1995). The widely cited work of Bevan et al (1992) on Kenya noted that 60 per cent of proceeds from the Kenyan coffee boom in the mid-1970s is saved. This is largely related to the fact that such incomes are windfalls that result from fluctuation in commodity prices (Deaton 1990). In Elbadawi and Mwega’s (1998) study a growth in the terms of trade has a positive and significant effect on saving rate. However, in Masson and others (1995) the terms of trade are found to be statistically insignificant. An extreme result of negative terms of trade effect is reported in Mwega’s (1997) study.

7. Macro policy issues relevant to enhance domestic resource mobilization are also examined in some of the African empirical literature on the issue. This can broadly be categorized as fiscal policy (in particular public saving and issue of crowidgn-out/in) and monetary policy. Corbo and Schmidt-Hebbel, 1991 cited in Mwega (1997); Mwega (1997) and Elbadawi and Mwega (1998) have used public saving as explanatory variable in their saving equation. For the sample of LDCs they found negative and statistically significant effect of public saving on private saving. On the other hand, government consumption is found to have a positive and significant effect in Mwega’s (1997) and Elbadawi and Mwega’s (1998) studies.

8. One way of augmenting public saving is through taxes. It is argued that this situation brings about what is called the Ricardian Equivalence. Most empirical studies for industrial countries reject the Ricardian equivalence. Studies for developing countries also dismiss it in its pure form and agree that public saving offsets some private saving (Haque and Monties 1989; Corbo and Schmidt-Hebble 1991; Easterly, Rodgigues and Schmidt-Hebble 1994; Edwards 1995, all cited in Schmidt-Hebble and others (1996) and Masson and others (1995)). These results show that public saving is an effective tool to raise national saving (Schmidt-Hebble others 1996: 99). Thus, the evidence is not conclusive.

9. A related fiscal issue is the impact of public investment on private investment and hence saving, if saving and investment have positive relationship. Serven and Solimano (1993) have examined the impact of public investment on private investment in developing countries and reported a positive and significant correlation in the panel data of developing counties, as well as in separate studies of Latin America and East Asia. Similar stylized fact is found across many developing countries (See Taylor 1991). What the empirical evidence suggests about the impact of public investment is that different types of public investment are likely to have different kinds of effect. Empirically, such examination shows that infrastructure investment generally has a positive impact while investment by public enterprise does compete with private investment (See Easterly and Rebelo (1993) cited in Schmidt-Hebble and others 1996). In general as noted by Azam and others (2000) the evidence is mixed. It is also noted that there are certain categories of public investment such as investment in infrastructure, human capital and law and order that tend to crowd-in private investment. Moreover, there appear to be lower (uppers) threshold below (above) which public investment may not be effective (See Azam and others 2000).

10. A monetary policy that is important in the empirical literature of domestic saving is the (real) interest rate. This has got prominent status following the Mckinnon and Shaw’s work on ‘financial repression’ that informed much of the design of SAPs in Africa. Thus many in Sub-Saharan Africa entered in financial liberalization since the 1980s. Despite such reforms however, real deposit rates have not significantly increased in many African countries (see Aryeetey and Udry 1999). Aryeetey and Udry noted that the real deposit rates have risen far slower than lending rates in many countries. However, citing the case of Ghana and Nigeria, they added, when there is some stability in macro-economic conditions and deposit rates rise, depositors react positively. However, this evidence is not conclusive. For instance, Giovannini (1985), Schmidt-Hebbel and others (1992) found no significant impact of real interest rate on saving, while Ogaki and others (1995) found positive effects that are small and very sensitive to income levels. A related financial variable used in the empirical studies is the degree of financial depth (usually measured by M2 to GDP ratio). In line with the mixed result in many studies of LDCs, the empirical evidence in Sub-Saharan is not conclusive either. For instance in Elabadwi and Mwega’s (1998) study this variable is found to be non-significant for a sample of LDCs. Mwega (1997) also reported similar result for Sub-Saharan Africa. However, in Elabadwi and Mwega’s (1998) recent study, it turned out to have significant positive effect when a fixed-effect model is used.

11. Both fiscal and monetary policies are related to macroeconomic stability. Macroeconomic stability is another area that is usually emphasized to strengthen domestic resource mobilization. The empirical evidence of the effect of macro-economic instability on saving rate in developing countries is mixed and inconclusive. In SSA there is evidence that macroeconomic stability leads to a rise in deposit rates and depositors react positively to this rise as noted earlier (Nissanke and Aryeetey 1998, cited in Aryeetey and Udry, 1999). But Mwega’s (1997) result shows that inflation (as one of the indictors of macro instability) has no significant effect on saving.

12. Macroeconomic stability is also related to institutional measures such as financial liberalization, appropriate credit policy and ensuring low level of income inequality as the experience of East Asian countries shows (See Birdsall and others 1999 for details). Ikhide (1996, cited in Aryeetey and Udry 1999) argues that institutional and structural constraints to saving are the major reasons for weak saving mobilization in Africa. This is compounded, he argues, by low presence of formal institutions. Extension of commercial bank branches to rural areas in five African countries covered in his study turned out to have the strongest effect on savings. Nissanke and Aryeetey (1998 cited in Aryeetey and Udry 1999) have also suggested that the fragmented nature of financial markets in Africa tend to increase the transaction cost of moving from one segment to the other and hence could act as a disincentive for saving mobilization in Africa.

13. One predominant institutional feature of saving in Africa is the importance of informal saving. Deaton (1989) suggested that saving in such set up is intended to smooth consumption. Aryeetey and Udry (1999) though agree with this notion emphasize that most studies of fund utilization by such association shows that the funds are usually spent on consumer durables and for providing working capital (Miracle et al, 1980; Aryeetey and Gckel, 1991; Chpeta and Mkandawire, 1991, all cited cited in Aryeetey and Udry 1999). This informal financial sector is important because the available evidence indicate that the value of formal sector financial assets is less than half of the financial assets held by households in Africa, although financial asset in general is relatively a small component of the portfolio asset held by households (See Aryeetey and Udry, 1999).

14. Another structural/institutional feature noted to be important in the African context is the transport cost. Aryeetey and Gockel (1991), cited in Aryeetey and Udry (1999) have noted that an average travel time of over an hour is required to reach a bank in rural Northern Ghana and the cost to such travel is about the equivalent of the prevailing minimum wage. This suggests that the incentive to save could easily be offset by the transport cost as long as the cost exceeds the return on saving. Webster and Fdler (1995, cited in Aryeetey and Udry 1999) attributed the low scale of a number of micro finance arrangements in West Africa in part to the low population density in many of the rural areas - indicating the importance of location to access credit.

15. Finally demographic characteristic of the household are also taken as important factors that could influence domestic resource mobilization. For a sample of developing countries, Elbadawi and Mwega (1998) used two demographic variables: the young-age dependency ratio and urbanization. In the pooled model both variables have negative and statistically significant effects. This result becomes insignificant, however, when the fixed effect model is used. Comparing their estimation for LDCs with that of SSA, they noted, for SSA the dependency ratio emerged as the most important and robust contributor that differentiate the performance of high performing Asian economies from SSA. The dependency ratio has negative contribution especially in middle income SSA (Mwega, 1997: 214; Elbadawi and Mwega, 1998: 19). Masson and others (1995) also found that demographic factors have significant negative effect for all but middle income LDCs. Others (See Harrigan, 1995, cited in Mwega 1997) noted that empirical evidence is conflicting and has not resolved the issue. Mwega (1997) reported that adverse effect of high dependency ratio on private saving appears to have little support for the sample of LDCs in his study. Deaton (1989) has shown that for developing countries actual age-composition profiles are not consistent with the predictions of life-cycle theories, thereby undermining the empirical importance of the mechanism. The weakness of the life-cycle model in developing countries is also noted by Collins (1991) based on a study that used a sample of ten developing countries (See Aryeetey and Udry, 1999).

16. The discussion in this section, in particular the exploration of myriad of saving determining factors, clearly demonstrates the challenge of domestic resource mobilization in Africa. The identification of different factors that affect each of the saving determinants, it is hoped, would provide some handle on designing relevant policies geared to domestic resource mobilization (See section IX for detail). More important, however, is the inconclusive nature of most empirical studies both in developing countries in general and in Africa in particular. This points to the need to carry out an in-depth research in these areas.

.

.

VI Issues Of Capital Flight

1. One important area that increasingly attracting the attention of researchers and policy makers is the potential of capital flight reversal in financing development. In the context of Africa this potential is enormous. Notwithstanding the measurement problem associated with such studies, a recent study using a rather large data set based on 22 countries from Sub-Saharan Africa concluded that the continent has the highest incidence of capital flight, exceeding even the Middle East. 39 per cent of private portfolios were held outside the continent. Were Africa able to attract back this component of private wealth, the private capital stock would have increased by around 64 percent (Collier and others 1999). Similarly, Ajayi’s 1997 estimate of capital flight from severely indebted low-income countries of Sub-Saharan Africa, which stood at 22 billion, constitutes nearly half of the external resource requirement estimated by Amoako and Ali for 1999/2000.

2. Recent African literature on the topic touches a number of issues to understand determinants of capital flight. According to Collier and others (1999) capital flight arises from portfolio diversification incentives, return differential incentives and relative risk incentives. The risk associated with domestic capital may range from expropriation to wide range of implicit taxes such as inflation or exchange rate depreciation (Collier and others 1999). The empirical analysis of this study revealed that the high level of capital flight from Africa, despite the continent’s capital scarce characteristics, is explained by overvalued exchange rate, the fact that it is rated as the riskiest continent by international investors, and the level of indebtedness of the continent (Collier and others 1999). The finding about debt is, however, contradicts to the finding of Ajayi (1997) who argued that there is no relationship between debt and capital flight.

3. An earlier study by Hermes and Lensink (1992) on capital flight from Africa used debt-creating flows, exchange rate overvaluation and exchange rate adjusted interest rate differential as explanatory variables. Except the last item that is found to be statistically insignificant, they have found positive and statistically significant coefficients. This basically accords with Collier and other’s (1999) recent finding. Our experiment with similar equation using Hermes and Lensink type of specification using 18 African countries could not result on significant coefficient for the coefficient of the debt creating flows – thus supporting Ajayi’s result. Some of these inconclusive results indicate the importance of further study on these issues.

4. Notwithstanding the inconclusive findings about capital flight, it is possible to identify preliminary pointers. The major finding of Ajayi’s examination of African capital flight points to the importance of ‘trade-faking’ (over and under invoicing of imports and exports), problems of political instability (including the abuse of power), unfavorable macroeconomic environment, and lack of economic growth as important areas that African countries need to work on so as to ensure the reversal of flight capital and maintaining what is reversed (See Ajayi 1997 for details). The Collier and others (1999) findings also basically accord with this view (See section IX for detail).

.

VII Africa’s External Debt

.

1. The total external debt of Africa has raised nearly twenty-five folds from a relatively low level of US $14 billion, in 1971, to more than $300 billion now. The major component being outstanding long-term debt. Over period, IMF credits were increasingly used, with ‘Structural Adjustment’ and ‘Enhanced Structural Adjustment’ facilities comprising an ever-important component of flows to Africa (See Table 7.1 in Appendix I).

2. Changes in the structure of African debt can be described in terms of creditor patterns. From Table 7.1 (See Appendix I), it can be read that bilateral debt comprises the largest share. This is followed by multilateral debt, with private inflows declining overtime. Generally, a larger share of official debt is now disbursed on concessional terms. It can also be noted that the debt problem is being aggravated by capitalization of interest and principal arrears, which constitute nearly a quarter of the external debt burden.

3. Although the share of African debt as a proportion of the total debt of developing countries is low, the relative debt burden born by African nations remains high. The debt to GNP and debt service ratios rose from 21 per cent and around 9 per cent, respectively, in 1971, to a high of 68 and 23 per cent during the late 1980s. In 1998, the last year for which we have data, these ratios stood at 64 per cent and 19 per cent, respectively (See Table 7.2 in Appendix I).

4. Africa’s debt burden may also be assessed by examining net transfers. Thus, if we exclude from Table 7.2 in Appendix I, grants and net foreign direct investment inflows, it can be seen that net transfers since 1990 have, in fact were negative (i.e., flowed from Africa to the developed nations). Further more, the level of such transfers (outflow) has increased, from US$ 3.6 billion in 1985 to nearly US$ 12.5 billion in 1998. Finally, in the 1990s it is worth pointing out that nearly 35 per cent of grants to Africa, in fact, went to ‘technical experts’ that usually came from donor countries (See Table 7.2 in Appendix I).

5. In summary, the last three decades have witnessed an unprecedented increase in the level of African debt. This debt is generally long-term in character; there is a growing importance of debt owed to bilateral and multilateral creditors, a trend away from concessionality to non-concessionality and an increase in the importance of interest and principal arrears (usually capitalized through the Paris and London clubs) as a component of long-term debt. Indicators of the debt burden also reveal that African debt is extremely heavy compared to the capacity of the African economies, and, in particular their export sectors. Moreover, most African countries have been subjected to net financial outflows (of debt related flows) since the mid 1980s. The performance of these economies, coupled with a mounting debt burden, surely indicates that African countries are incapable of simultaneously servicing their debt and attaining a reasonable level of economic growth, let alone addressing issues of poverty alleviation outlined in this paper.

6. The actual size of indebtedness does not usually represent an economic problem in itself, since this debt may usually be mitigated by rescheduling and similar short-term arrangements. However, the size of accumulated debt, relative to capacity, and subsequent impacts on the economy, do represent a serious problem for African countries. In this respect, three inter-related implications of the debt problem deserve mention. First, servicing of the external debt erodes foreign exchange reserves, which might otherwise be available for purchase of imports. This has led to the ‘import compression problem’ in the past, in which shortage of foreign exchange adversely affected levels of public and private sector investment (See for instance Ndulu, 1986, 1991 and Ratso 1994). Second, the accumulation of a debt stock results in a ‘debt overhang’ problem, which tends to undermine the confidence of private investors, both foreign and domestic. A decline in levels of private investment as a share of GDP, from the late 1970s onwards, may partly be attributed to this factor (See for instance Elbadawi and others 1997). Finally, servicing of debt is placing an enormous fiscal pressure on many African nations. Such pressure has had an adverse effect on public investment and provision of social services; this finding is reflected in the decline in the share of public investment in GDP from late 1970s onwards as well as high level of fiscal deficit. Naturally, a reduction in levels of public investment will tend to have adverse consequences for physical and social infrastructure. This effect is significant given the empirical finding that public sector investments, in particular in low income countries of Africa does crowd-in private investment (See Alemayehu 2000).

7. To sum up, the debt issue is a crucial element of the overall economic crisis facing Africa. Except the recent gesture shown to Uganda (a 20 percent debt stock reduction), the much-discussed issue of the HIPC initiatives is not realized yet. This initiative comprises two phases. Phase one requires a three-year record of compliance with an IMF program. This leads to the decision point for acceptance into the HIPC initiative. Acceptance requires having indicators that show the country is beyond debt sustainability thresholds. These are defined as a 20 to 25 per cent debt service ratio and present value of debt to export ratio of between 200 and 250 per cent. If countries find themselves above these thresholds they are eligible for the Naples terms, under which they are entitled to a two-third reduction in debt stock. This reduction is applied to all types of debt, including multilateral. Phase two of the new initiative comprises the implementation of another three-year IMF program and, if accepted within the HIPC category, an 80 per cent debt stock reduction. Since it is widely believed that few African countries are likely to reach this phase in the foreseeable feature, one may not expect to get much from this second phase (See Oxfam, 1997 for a critical review of this issue).

8. The Bank and the Fund have also announced a major expansion of the HIPC initiative in 1999. The new expanded HIPC initiative is believed to provide deeper, broader and faster debt relief by (a) qualifying countries when their present value of debt to export reaches 150 percent (as opposed to 200-250 per cent noted above), (b) commencing debt relief from decision point, and (c) basing the length of the interim period on achievements of key development objectives rather than on pre-specified period. Another new dimension emphasized is the idea of linking the debt reduction to poverty alleviation programs (See World Bank and others 2000). Although this new extended initiative seems promising, it is very difficult to count on its realization and hence its financing development prospects as can be read form the recent resolution of the (July 2000) G-7 meeting in Japan.

.

VIII The New Financial Architecture And Its Implication

1. The recent turbulence and crisis in the financial sectors of emerging economies and to some degree in developed countries has prompted a renewed emphasis on examining the operation of the international financial system. There seems to be a consensus that the problem is global and systemic implying the need to go beyond national boundaries to address this issue. In particular, the Asian crisis underscored the need to carry structural policy reforms both at the national and international levels to restore financial stability and to mitigate and if possible avert future crisis (See Bossone and Promisle, 1999). The major issues that need to be addressed are categorized under the following themes: (a) enhancing standards and transparency, (b) financial regulation and supervision, (c) management of capital account and exchange rate regimes, (d) surveillance of national policies, (e) provision of international liquidity and (g) orderly debt workouts (See IMF 1998, Bossone and Promisle 1999, Akyuz 2000).

2. At practical level, following the Asian crisis, Finance Ministers and Central Bank Governors form 22 countries, which are significant players in the international financial system, gathered to work on a new international financial architecture spearheaded by the Bretton Woods institutions. These Ministers and Governors identified three key areas where immediate action is needed: (a) enhancing transparency and accountability, (b) strengthening national financial systems and (c) managing international financial crisis (IMF 1998).

3. Transparency and Accountability: this relates to the importance of improving the coverage, frequency and timeliness with which data on reserves, external debt and financial sector soundness are published. In particular information on the international exposure of investment bank, hedge funds and other institutional investors is required. Adherence to transparency is believed to enhance the performance and accountability of international financial institutions too (IMF, 1998). In this proposal the international community will set the standards for and improving the timeliness and quality of information on key macroeconomic variables, including transparency of governments fiscal and monetary policy (Akyuz 2000).

4. Strengthening financial systems: this relates to principles and polices that foster the development of a stable and efficient financial system in areas of corporate governance, risk (including liquidity risk) management and safety net arrangements. The domestic financial sector is also believed to be strengthened by cooperation of national and international institutions engaged in supervision and regulation of financial institutions (IMF, 1998). This may be carried along the standards based on Basle Capital Accord that need to be applied by national authorities. This accord has various requirements such as provision for risk and, specific standards of supervision and regulatory framework (Akyuz 2000).

5. Managing International financial crisis: the working group set up to examine this issue stressed the need to encourage better management of risk by private and public sectors. In particular, it wants to see governments’ limiting the scope and clarifying the design of guarantees that they offer. Moreover, controlling short term, liquid capital flows through market based measure such as taxes and reserve requirements, exchange rate management (such as targeting the real exchange rate), IMF surveillance over the policies of creditors as well as debtors, provision of liquidity (to pre-empt large currency swings) by an international institutions such as IMF, and an orderly debt workout (such as temporary standstill on debt servicing and provision of seniority status to new debt and similar arrangements) are tools that can be used in managing financial crisis in the context of the new financial architecture (See Akyuz 2000: 5-14). In particular the international financial architecture need to develop in the direction where the role of international financial institutions such as IMF is to reduce demand for liquidity ex ante (preventing collapse) and increase supply of liquidity ex post (organizing bailout) by designing appropriate strategy to tackle possible moral hazard problems. This is in particular important in developing countries where the welfare consequence of financial instability is extremely harmful (See FitzGerald 1997a 199b).

6. Implications of this new financial architecture for developing countries in general and Africa in particular are many. This list may include,

• Many of the proposal are focused on marginal reforms and incremental changes such as standards, transparency etc instead of systemic instability issues raised in the wake of the Asian crisis – the focus is toward self-defense and market based solution. Such solution is not sustainable (See Akyuz 2000)

• Explicit rule based operation is generally resisted by industrialized countries. This gives industrial countries an excessive discretionary power since they have leverage on international financial institutions. This in turn has implication for reforming the rules of the international financial infrastructure such as the IMF quota system. Moreover, Public disclosure of financial and macro information may enhance instability, in particular if there are no asymmetric action/transparency by creditor countries and institutions such as IMF (See Akyuz 2000, Jha and Saggar 2000).

• The credibility of institutions, which carry the supervision and regulation of national economies (such as credit rating agencies), may create a problem. Even when it is believed, it might have detrimental impact on flows to Africa or the interest rate spread of flows destined to Africa since private investors notoriously lack information about individual African countries. Experience in Asia shows that self-surveillance is much more important (See section IX for detail).

• African countries need also to carry an in-depth study of the implication of internationally based rules and regulation to a region that is marginalized and perhaps participating in a segmented capital market. For instance, most of the debt owed to Africa is publicly guaranteed and in general based on sovereign risk. Thus, the rule developed for market based capital flows could hardly be relevant to such segmented market. (See section IX for detail).

• A recent study by World Bank noted: (a) care must be taken to design appropriate policies at the initial stage of surge in capital flows as this largely determines success in dealing overheating and potential vulnerability, (b) development of well functioning capital market reduces potential instability, and (c) developing countries need to develop shock absorbers such as international reserves that varies with variation in capital account (as opposed to simple import cover), fiscal flexibility (in indebtedness), building up of cushions in the banking system and the like (See World Bank, 1997)

• Finally, African countries need to develop tools (such as global models) that could be used to analyze the impact of Northern polices and external shocks on their region. Such tools do also allow examining policy responses to such shocks (See section IX).

IX POLICY IMPLICATIONS

1. This section attempts to bring together the policy implications of issues discussed in this paper. The major conclusions that emerge from the resource gap analysis is that the continent needs an estimated external resource that amounts to 46.8 percent of its GDP (40 % if S. Africa is excluded), annual average of 32 percent (26 % if S. Africa is excluded), and 10 percent (8% excluding S. Africa) in the period 1999/2000, 2001-2005 and 2006-2010, respectively, to attain 8 percent growth in GDP. The latter is believed to reduce poverty by half by year 2015. This figure, although apparently seems large, is very small if the implication of HIV/AIDS, estimated to bring about 1 to 1.5 percent reduction in GDP is factored in.

2. Mobilizing such high level of external resource is a formidable task. In order to realize this, policy makers need to design policies relevant for each category of (possible) source of external finance. The rest of this section will outline such policy guidelines organized by each financing category.

Official Development Assistance (ODA)

3. The fall in percapita ODA and its declining trend needs to be a real concern since it is the main source of financing the resource gap of African countries. Thus, policy makers need to design ways of reversing this trend by working both collectively and at specific country level. The observed declining trend also strengthen the case for efficient use of the existing level of ODA more than ever.

4. The theoretical and empirical analysis about determinants of ODA flows shows that donor interest, rather than recipient needs, is much more important. The geo-political interest of donors as well as the actual/bureaucratic process of aid delivery are specifically singled out as important. Policy makers, thus, need to find ways of influencing these variables to bring about meaningful change in such flows.

5. At specific level, the pre-Lome association of African countries with Europe and their economic performance are much more important in influencing aid flows than humanitarian considerations. Thus, policy makers need to creatively design policy instruments that can positively influence these variables. More specifically, policy makers need to identify and examine (a) the geo-politics of their country/regions so as to creatively use them, (b) understand the budgetary/bureaucratic process of aid delivery from the supply side; and (b) pursue growth-oriented strategies from the demand side.

6. In the face of the declining trend of ODA, aid effectiveness is also another policy direction. This requires specific polices and strategies aimed at (a) raising capital efficiency, (b) transiting from aid-dependency in the medium to long run and (c) a strategy to tackle the terms of trade deterioration which is increasingly offsetting the aid inflow. This may take the form of export diversification and developing collective bargaining position at regional and continental level.

7. Aid dependency is also found to have long-term detrimental effect on existing institutional capacity of African countries that are already weak. Policy making, thus, need to focus on insulating existing institutions from such effect by designing appropriate strategy. The latter may include capacity building both at country and regional level. Such capacity building needs to go beyond economic management to issues of governance, which is central in aid management.

8. The modality of aid need also be based on partnership and shared vision of donors and recipients, with clear defined nation wide program (such as medium term budget or expenditure framework) with focus on social sectors.

9. Finally, the international community has also its share of the task such as: ensuring ownership of policies by recipients, capacity building in the formulation and maintaining of medium term planning and budgeting approaches such as the Medium Term Expenditure Framework (MTF) which is increasingly informing macroeconomic management in many African countries. Moreover, donors need also to give at least equal significance to regional or continental undertakings, which may help to create long run enabling environment. African governments may need to put collective pressure on international donor institutions to realize the afromentioned objectives.

Foreign Direct Investment (FDI)

10. Critical analysis of the theoretical literature on FDI in African context suggests the importance of understanding Multinational Corporations (MNC) to understand FDI flows to Africa. Thus, policy makers need to invest in understanding on how MNC do operate. This is helpful both to positively influence them and to regulate their operation when such regulation is invaluable.

11. FDI related empirical studies in Africa underscore the importance of relative market size, mining activity as well as the historical pattern of FDI flows in attracting FDI. These findings have two policy implications. In the short to medium term appropriate macro policy is important. In the long run, however, structural transformation to bring about meaningful change in market size, availability of skilled labour force and enabling infrastructure is vital to attract FDI.

12. The recent surge in FDI in extractive sectors of African economies might be explained by the profitability of such sectors owing to the existence of fairly secured world market for such commodities (such as oil) and the relative (low) cost of creating an enabling environment for such industries (such as security and site-to-port infrastructure). The policy implication is that recipients need to offer enabling condition by investing on public goods such as security and infrastructure so as to make investing in their country less costly. From a long run perspective, a concerted effort to direct FDI flows to other sectors that have a potential world market through appropriate (and graduating) incentive structure is important. Moreover, countries need to encourage joint ventures to strengthen local entrepreneurs and domestic technological capacity.

.

13. Finally, as much as African countries need FDI, they also need to develop appropriate regulatory framework to counteract possible adverse effects of FDI. This may include dealing with the development of an enclave sector, detrimental transfer pricing and possible adverse effects of FDI on the infant and often less competitive domestic producers.

Other Private Capital Flows

14. Drawing policy implications about other capital flows requires appropriate data. This is a major problem with regard to private capital flows. Thus, compiling, and analyzing such data both at regional and specific country level is central to make informed policy. Investing on capacity building in such area could attain this.

15. In the face of this acute shortage of external resources, encouraging such flows is important. This can be done by designing appropriate macroeconomic and financial policies. This may include creating stable exchange rate regime, competitive interest rate policy, provision of foreign currency account facilities and a framework for appropriate financial sector supervision and regulation (see below). Moreover, policies aimed at improving the credit rating of African countries such as proper debt management, low repayment arrears and investment guarantees as well as provision of accurate (unbiased) information about Africa are important instruments. At regional level, regional integration and standardization of financial sector operations across regional groupings are important policy instruments to attract private capital flows.

16. Recent studies also show that suppliers of such capital flows have no information about opportunities in Africa (because of partly negative media coverage). Studies also show that equity flows to developing countries are much more affected by country specific factors (while bond flows by global factors). African policy makers, thus, need to bridge this information gap in a sustained manner.

17. The empirical literature also points to the importance of push factors (such as trends of opportunities in home country of the investors) in influencing private capital flows to developing countries in general African countries in particular. The observed low correlation between yields in African and emerging markets is an advantage to Africa since such funds can find their way in Africa when such yields are low in emerging markets. The policy implication is that African countries need to closely monitor the trend and pattern of push factors in source countries and exploit (or design ways of exploiting) such opportunities. This requires developing the capacity to monitor and analyze such trends both at country and regional level.

18. Finally, attracting such flows need to be carried cautiously. This is because of the notorious impact of such flows in destabilizing economies when investors suddenly pull them back. In the short to medium term the following policies could be very important:

1. To develop prudential regulatory and supervisory body at central banks and relevant regional bodies,

2. To develop the human capital required to carry out point (a) above,

3. Gradual (step-by-step) implementation and appropriate sequencing of financial sector liberalization programs (especially the liberalization of the capital account), and

4. In the medium and long term, implementing fiscal and financial polices aimed at discouraging sudden withdrawals of such funds (as has been done in some Latin American countries) from recipient countries.

Domestic Saving Mobilization

19. Notwithstanding the importance of domestic saving, the empirical evidence about the determinants of saving is largely inconclusive. Thus, coming up with specific recommendation is difficult. This points to the need for conducting further research on the issue. We this caveat, the following pointers are singled out.

20. First, policy makers need to focus not only on growth but also (and perhaps more importantly) on its sustainability so as to bring about a positive impact on domestic saving.

21. Second, according to the existing empirical evidence, there is a policy dilemma of raising external resources (in particular aid) and its negative impact on domestic saving. Policy makers, thus, need to find creative ways of neutralizing the possible negative impact of aid on saving.

22. The existing African empirical evidences also emphasize institutional and structural factors that could determine the level of domestic saving. This list includes:

1. Structural transformation of the economy,

2. Expansion of saving mobilizing institutions such as banks,

3. Developing basic infrastructure, especially transport, and

4. Tapping the potential in the informal saving which comprises the bulk of household saving in Africa.

23. Finally, demographic and social polices aimed at reducing dependency ratio, such as family planning and basic (female) education, as well as encouraging labour-intensive and off-farm employment are instrumental to raise domestic saving.

Capital Flight and External Debt

24. The high incidence of capital flight in Africa is explained by return differential, risk (such as nationalization, implicit taxes for instance inflation and overvaluation of the currency) and indebtedness. Addressing these problems may have a positive impact on reversing capital flight. Moreover, polices aimed at macroeconomic and political stability, debt management and appropriate incentive structure as well as provision of foreign currency denominated banking services are important.

25. In terms of debt, although collective voice for debt cancellation is important, appropriate policy designed to mitigate debt-creating flows is important in the long run. Moreover, efficient use of such flows (e.g. local ownership of technical assistance as well as raising productivity of public expenditure) needs to be an integral part of such strategy.

26. In the long run, since trade in primary commodities is the main source of debt repayment problems and indebtedness, a structural transformation of African trade is fundamental to tackle the debt problem.

The New International Financial Architecture

27. The impetus for designing the new international financial architecture came from the Asian crisis and focused on problems specific to that region. This may not that relevant to African countries where Asian-type volatile flows are not created yet. Thus, African policy makers need to articulate their own need and attempt to bring such issues on board when a new financial architecture is being designed. An in-depth study of the reason for the new initiative (such as the Asian crisis), its content and relevance for African countries need to be a priority for formulating a common position by African countries.