Information about Sub-Saharan Africa África subsahariana
Chapter

12 Private Capital Flows to Sub-Saharan Africa: An Overview of Trends and Determinants

Editor(s):
Zubair Iqbal, and S. Kanbur
Published Date:
September 1997
Share
  • ShareShare
Information about Sub-Saharan Africa África subsahariana
Show Summary Details
Author(s)

The surge in private capital flows to developing countries during the 1990s has largely bypassed sub-Saharan Africa. In sharp contrast to the earlier lending boom of 1977–82, when sub-Saharan Africa accounted for 8.9 percent of total private flows to developing countries, the region accounted for only 1.6 percent of such flows during the period 1990–95. However, this low share of international private flows during the 1990s masks significant differences among countries and types of flows. This paper provides an overview of trends in private flows to highlight both common features and differences within the region, and undertakes an analysis of the macroeconomic factors that have influenced private flows to the region.

Overview of Trends

Official and Private Finance in Sub-Saharan Africa

Official finance accounts for a higher proportion of external financial flows to sub-Saharan Africa than any other developing region (Figure 1). Despite the sharp increase in official finance to Europe and Central Asia in the 1990s, sub-Saharan Africa continues to account for the largest and indeed growing share of official development finance; during 1990–95, its share amounted to 26 percent of total official development finance to developing countries. Almost 95 percent of this was on highly concessional or grant terms.

In contrast, the share of long-term private capital flowing to sub-Saharan Africa is lower as a percentage of GNP than other developing regions with the exception of South Asia. Private transfers and other private flows (including capital flight) play a relatively important role in sub-Saharan Africa as they do in other regions like South Asia and the Middle East and North Africa. However, the inclusion of these flows does not change the picture. In fact, when these flows are included, private flows to Africa are lower as a percentage of GNP than all other developing regions.

Figure 1.Structure of External Finance: Annual Averages, 1990–95

(Percentage of GNP)

Source: World Bank Debtor Reporting System.

Trends in Private Flows Compared with Other Regions

Along with Latin America, sub-Saharan Africa saw the sharpest decline in private flows in the aftermath of the debt crisis (Figures 2 and 3). Private flows began to recover in the second half of the 1980s, but in contrast to the experience of most other developing regions, they declined again in the early 1990s before recovering modestly in the 1993–95 period. For most of the years during 1982–95, annual long-term private capital flows were less than half the peak of $5.5 billion reached in 1982. If South Africa is included, the decline in flows in the early 1980s was more pronounced and the level of flows even lower throughout the decade.

Figure 2.Private Capital Flows by Region

(In billions of U.S. dollars)

Sources: World Bank and IMF data.

1 Excluding South Africa.

Differences in Overall Trends Between Country Groups

Figures 4 through 9 depict trends in private flows (with and without private transfers) for three alternative sets of country groupings: CFA countries compared with non-CFA countries; countries registering positive per capita growth over the period 1988–95 compared with those that had negative per capita growth over the same period; and low-income countries compared with middle-income countries (excluding Angola and South Africa).1

CFA countries suffered larger and more sustained declines in private flows than did non-CFA countries. When private transfers are included, the difference is even more striking. Overall private flows (including private transfers) to the CFA group remained negative throughout the 1985–95 period. In contrast, private flows showed a significant recovery for the non-CFA countries in the second half of the 1980s, and a further increase in the 1993–95 period. Contrary to the aggregate ratios for Africa, private flows to some non-CFA countries are not much lower as a percentage of GDP than other developing regions.

Figure 3.Private Capital Flows by Region

(In billions of U.S. dollars)

Sources: World Bank and IMF data.

In a similar fashion, countries that recorded positive per capita growth during the period 1988–95 showed higher levels of private flows than countries that experienced negative per capita growth in the same period. Growing economies also showed an improving trend, especially when private transfers are taken into account. In terms of income groups, the middle-income countries (excluding Angola and South Africa) display a very erratic but long-term declining trend in private flows, which appears to have been arrested in the last two years (1994–95). The low-income countries experienced a recovery in private flows in the second half of the 1980s and a further increase during the 1993–95 period.

Composition of Flows

Underlying the aggregate trend in private flows are quite marked differences in trends between different types of flows (Figure 10). Long-term private capital flows are divided into three components: foreign direct investment, private loans and portfolio equity flows. Private loans, in turn, consist of loans from commercial banks, bond finance, and other flows. However, bond issuance has so far been limited to very modest amounts in the case of sub-Saharan Africa.

Figure 4.Private Capital Flows to Sub-Saharan Africa,1 CFA and Non-CFA Countries

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa

Private loans, which were the dominant component of private flows during the commercial bank lending boom of 1977–82, saw a sharp decline following the debt crisis, a brief recovery in the second half of the 1980s, and a subsequent decline. This component has been negative or close to zero over most of the 1990s for sub-Saharan Africa in the aggregate. Foreign direct investment, conversely, has been on an upward trend throughout the 1980s and the 1990s and now dominates aggregate private flows. Finally, recorded portfolio equity flows to sub-Saharan Africa (excluding South Africa) were nonexistent until 1992, amounted to $17 million in 1993, jumped to $641 million in 1994, but fell back to $297 million in 1995. Portfolio equity flows to South Africa—which account for 16 percent of the IFC Emerging Market Index—have seen a more spectacular increase, skyrocketing from $144 million in 1992 to $4.6 billion in 1995, the largest such flow to any developing country in that year.

Figure 5.Private Capital Flows (Including Net Transfers) to Sub-Saharan Africa,1 CFA and Non-CFA Countries

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa

Foreign Direct Investment

Foreign direct investment has shown a significant increase since the late 1980s for the non-CFA countries as well as for countries registering positive per capita growth (Figures 11 and 12). For some countries in this group, foreign direct investment as a percentage of GDP in 1994–95 compares favorably with recipients in Asia and Latin America. For instance, Ghana, Mozambique, and Nigeria all received more FDI as a percentage of GNP than Brazil, India, Mexico, and the Philippines. By contrast, FDI flows have been stagnant or shown only a modest increase for the CFA countries and for countries with negative per capita growth.

Figure 6.Private Capital Flows to Sub-Saharan Africa,1 Countries with Positive and Negative Per Capita Growth

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa.

The major recipients of FDI flows to Africa can be placed in three broad groups. The first consists of the long-term recipients, including Botswana, Mauritius, Seychelles, Swaziland, and Zambia. Since these countries were early recipients of FDI flows, net FDI flows have tended to plateau and even decline in the case of Botswana and Zambia. The second group consists of countries that recorded large increases in FDI flows during the 1990s; Angola, Cameroon, Gabon, Ghana, Guinea, Lesotho, Madagascar, Mozambique, Namibia, Nigeria, and Zimbabwe, A large proportion, though not always the majority, of FDI inflows to these countries has been directed to the oil and mining sectors. The third group of countries are those that saw low and declining levels of foreign direct investment through much of the 1980s and early 1990s, but where there have been encouraging trends of a turnaround in the last year or two. In some cases, the turnaround has been spectacular. For instance, FDI flows to Uganda are estimated to have reached $112 million in the 1995/96 fiscal year, or 23 percent of total net official inflows and 2 percent of GNR2

Figure 7.Private Capital Flows (Including Net Transfers) to Sub-Saharan Africa,1 Countries with Positive and Negative Per Capita Growth

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa.

Despite these promising trends, most countries in sub-Saharan Africa receive very modest magnitudes of foreign direct investment. This is despite the fact that rates of return on FDI have generally been much higher in Africa than in other developing regions. During the period 1990–94, rates of return on FDI in Africa averaged 24–30 percent, compared with 16–18 percent for all developing countries.3 This suggests that risks are perceived to be higher in sub-Saharan Africa than in other regions of the world.

Figure 8.Private Capital Flows to Sub-Saharan Africa,1 Low- and Middle-Income Countries

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa.

Several economic and political factors are put forward as constraining foreign direct investment in Africa:4

(1) Civil strife. On the one hand, during the past 15 years, a relatively large number of countries in the region have been affected by civil strife, which in the most extreme cases (Liberia, Rwanda, Somalia, Sudan, and Zaïre) brought FDI inflows to a standstill. On the other hand, several countries that have seen an end to civil conflicts (e.g., Angola, Mozambique, Namibia, South Africa, and Uganda) have benefited from a significant increase in FDI inflows during the 1990s.

Figure 9.Private Capital Flows (Including Net Transfers) to Sub-Saharan Africa,1 Low- and Middle-Income Countries

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa.

(2) Macroeconomic instability. Large structural fiscal deficits, erratic monetary and exchange rate policies, and weaknesses in financial systems in many countries have contributed to high and variable inflation and interest rates and a high degree of variability in real effective exchange rates. These factors have all worsened the general investment climate.5 In the next section, it is shown that countries that have made some progress in reducing macroeconomic instability have been more successful in attracting FDI inflows.

(3) Slow economic growth and small domestic markets. Although FDI investments in the primary sectors in Africa have, on average, earned a high rate of return, the poor growth performance of sub-Saharan Africa and the limited size of domestic markets are seen as deterring more broadly based foreign direct investment. GNP growth in sub-Saharan Africa (excluding South Africa) averaged 2.3 percent during the period 1983–59 and 1.4 percent during 1990–95 compared with 3.8 percent and 5.1 percent for all other developing countries (excluding the former Soviet Union).

Figure 10.Composition of Private Capital Flows to Sub-Saharan Africa

(In billions of U.S. dollars)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa.

(4) Inward orientation and burdensome regulations. Compared with other developing regions, which have seen dramatic shifts to more outward-oriented, market-based investment regimes since the mid-1980s, sub-Saharan Africa has remained relatively more inward oriented, with foreign investment often subject to excessive and discriminatory regulation.

(5) Slow progress on privatization. In contrast to many Latin American and Eastern European countries, which have used aggressive privatization programs to secure a large boost in foreign direct investment, progress in privatizing state-owned enterprises has been slow in Africa. During the period 1988–94, proceeds from privatization amounted to $2.4 billion in the case of sub-Saharan Africa compared with $63.4 billion for Latin America and $16.3 billion for Europe and Central Asia.6

Figure 11.Foreign Direct Investment In Sub-Saharan Africa,1 CFA and Non-CFA Countries

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa.

(6) Poor infrastructure, Sub-Saharan Africa’s physical, financial, human, and institutional infrastructure is all generally less developed than other regions’ and in many cases has actually deteriorated since the early 1980s. This reflects low and declining investment in all areas of infrastructure, heavy state intervention but with poor implementation capacity, and limited success thus far in expanding private provision of basic infrastructure.

(7) High wage and production costs. As a result of the macroeconomic and microeconomic factors listed above, and in some cases, countries’ labor market policies, wage costs in the region tend to be high relative to productivity levels. Overall costs of production are generally higher than else-where—for example, almost double that of low-income Asian countries.

Figure 12.Foreign Direct Investment in Sub-Saharan Africa,1 Countries with Positive and Negative Per Capita Crowth

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa.

Private Loans

Private loans have been on a declining trend for all country groups in sub-Saharan Africa (Figures 13 and 14). Unlike other developing regions where commercial bank lending has shown a sharp turnaround in the 1990s, commercial bank lending remains negative or at very low levels. In part this is because most African countries have not yet restored their access to financial markets. In contrast to other regions, where creditworthiness ratings have shown a marked improvement in the 1990s, creditworthiness ratings for sub-Saharan Africa have remained much lower, on average, and are only just beginning to improve (Figure 15). The main factors that are believed to have contributed to sub-Saharan countries’ generally low levels of creditworthiness are high political risk, weak growth and export performance, macroeconomic instability, and high levels of indebtedness. Low levels of commercial bank borrowing also reflect decisions made by many countries to restrict the level of such borrowing, especially for general budgetary support.

Figure 13.Private Loans to Sub-Saharan Africa,1 CFA and Non-CFA Countries

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa.

Portfolio Equity Flows

Although portfolio investment into sub-Saharan Africa (with the notable exception of South Africa) is still extremely small compared with flows into other emerging markets, there are encouraging signs of growing investor interest. Since 1994, more than 12 Africa-oriented funds have been set up with a total size of more than $1 billion. Initially, the focus of these funds was primarily the South African market, but the base has been broadening to encompass a growing (though still limited) number of countries, including Botswana, Côte d’Ivoire, Ghana, Kenya, Mauritius, Zambia, and Zimbabwe. This growing pool of portfolio investment is already perceived to bring important benefits including liquidity, incentives for privatization, and pressure for policy reforms and improvement of the financial infrastructure.

Investors, looking ahead, express guarded optimism about making portfolio investments in Africa. In sharp contrast to only a few years ago, virtually all stock markets on the continent have now been opened up to foreign investment, and in many countries there has been a shift away from state-centered ideologies. A number of factors are, however, still seen as constraining portfolio investment: investors view political instability and weak macroeconomic fundamentals as the most important impediments.

Figure 14.Private Loans to Sub-Saharan Africa,1 Countries with Positive and Negative Per Capita Growth

(Percentage of GDP)

Sources: World Bank and IMF data.

1 Excluding Angola and South Africa.

Many structural weaknesses are also viewed as inhibiting investment. A reduction in the transactions costs of doing business will be critical. The setting up of an efficient trading and settlement system and the presence of international custodians are important elements of the financial infrastructure that is needed to attract foreign investors. Corruption in the public sector, including the judiciary, is cited by many investors as not only increasing transactions costs but acting as a deterrent in its own right.

The supply of assets is still very limited and in addition to the public companies already listed on stock exchanges, the number of private firms listed needs to be increased. In some cases, privatization of public assets offers the best avenue for increasing the supply of assets and attracting foreign investors. While foreign investment can play a valuable role in stimulating capital markets in Africa, growth and stability of these markets will require the development of a healthy base of domestic investors. Pension reform and the promotion of mutual funds could encourage domestic investment in fledgling stock markets. Over the long term, deficiencies in human capital, infrastructure, and institutions need to be addressed if more African countries are to attract a larger share of global portfolio flows.

Figure 15.Trends in Country Creditworthiness

Source: Institutional Investor credit rating.

Note: Rating is on a scale of 1 to 100.

Macroeconomic Determinants of Private Capital Flows

This section uses panel data for the period 1980–95 to examine empirically the effect of domestic and external factors on the inflow of long-term private capital to 31 countries in sub-Saharan Africa.7 This sample was constructed from the total of 47 sub-Saharan countries by excluding: (1) “small” countries (Cape Verde, Comoros, Djibouti, Equatorial Guinea, São Tomé and Príncipe, and Seychelles);8 (2) countries that experienced protracted civil unrest in the period under consideration (Angola, Congo, Ethiopia, Liberia, Rwanda, Somalia, Sudan, and Zaïre); (3) Guinea-Bissau and Namibia, because information on real effective exchange rates from 1980 to 1995 was not available; (4) South Africa, because although the economic issues it is confronting may be similar to those faced by other middle-income African countries, the post-apartheid transition it is undergoing is of a different nature and the time it has spent in this new phase is too short to evaluate the trends in financial flows and macroeconomic variables. The final sample included Benin, Botswana, Burkina Faso, Burundi, Cameroon, Central African Republic, Chad, Congo, Côte d’Ivoire, Gabon, The Gambia, Ghana, Guinea, Kenya, Lesotho, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Niger, Nigeria, Senegal, Sierra Leone, Swaziland, Tanzania, Togo, Uganda, Zambia, and Zimbabwe. Below, besides examining the volume of private capital flows, we also analyze the behavior of the foreign direct investment and private loans components of these flows.

In some sub-Saharan African countries the volume of unrecorded flows can be relatively large and recorded flows may be subject to misclassification. For example, since the late 1980s Uganda has seen a considerable expansion of net transfers in the current account.9 It is very difficult, if not impossible, to assess to what extent such transfers represent a misrecording of items in the current account, return of flight capital, or inward movements of capital by foreign investors. Ideally, we would like to isolate the last category and use it to correct the recorded flows for conducting our analysis. However, pending information that may allow us to make such a judgment, our empirical work below uses the recorded data on international capital flows.

The general theoretical model underlying the regression specifications below is that long-term private capital flows into a particular country are determined by relative rates of return at home and abroad and the relative risks associated with such investments.10 Rates of return, risk perceptions of foreign investors, and the climate for foreign investment are affected by certain domestic characteristics of the countries and the international environment. The domestic factors are proxied by the growth rate of the economy, the rate of investment, the openness of the economy, the ratio of external debt to GDP, and volatility of the real effective exchange rate. The most important external factor is international interest rates, which provides a proxy for the opportunity cost of investing funds in developing countries. The channels through which these factors affect private capital flows and the expected qualitative impact of these variables is summarized thus:11

Real GDP growth: A rapidly or steadily growing economy is likely to offer higher rates of return and lower risks on investment. The expected sign on the coefficient is positive.

Investment rate: High investment rates serve as a proxy for expected future growth of an economy and hence higher expected returns. The expected sign on the coefficient is positive. However, as Hernandez and Rudolf (1995) note, to the extent that domestic investments are financed by foreign savings, there will be a positive contemporaneous correlation between high investment rates and capital inflows. This simultaneity issue is handled by using lagged investment rates as a proxy for future returns.

Commercial openness: A large traded goods sector also signals increased ability to compete in the international market place and a greater capacity to repay external debt obligations. The expected sign on the coefficient is positive.

External indebtedness: The larger the existing burden of external debt the greater the debt-service obligations and the greater the vulnerability of the economy to increases in international interest rates. The expected sign on the coefficient is negative.

Real effective exchange rate variability: High and variable inflation rates and erratic exchange rate policies are likely to result in a larger variance of the real effective exchange rate. A highly variable real effective exchange rate is likely to affect the country’s traded goods sector adversely as well as make the returns to foreign investors more uncertain. The expected sign on the coefficient is negative.

The variables used in the panel regressions are defined below:

PFprivate capital flows (percentage of GDP)
FDIforeign direct investment (as a percentage of GDP)
PLprivate loans (as a percentage of GDP)
RGDPCannual percentage change in real GDP
GFCFgross fixed capital formation (as a percentage of GDP)
OPENindex of openness, defined as sum of exports and im-
ports divided by GDP (as a percentage)
XDEBTtotal external debt (as a percentage of GDP)
REERCVcoefficient of variation of the monthly real effective
exchange rate index
USB3YU.S. three-year government bond yield (in percent)

The regression results are reported in Tables 1 through 4. The estimated specifications are linear and in the first three tables the two-factor panel regressions allow for fixed country effects and include annual time dummy variables that capture effects common to all countries but specific to each year.12Table 4 reports the effect of international interest rates using a one-factor panel model that uses country dummy variables only, since introducing international interest rates that are common across countries into the specification in the presence of time effects would lead to an identification problem.

Table 1 presents the estimation results for private capital flows. The coefficients for the output growth rate, investment rate, openness index, and external debt ratio have the expected sign and are statistically significant at the 5 percent level. These results suggest that output growth, gross fixed capital formation, and the openness of the economy positively affect the volume of private capital flows while a large external debt relative to GDP adversely affects inward movements of capital. The coefficient for real effective exchange rate variability is not statistically significant in this regression, which uses aggregate flows and does not distinguish between the components of private flows. The test statistics presented in the continuation of Table 1 indicate that the time effects are not jointly significant; their exclusion does not have any appreciable impact on the coefficients of the explanatory variables.

Tables 2 and 3 present the estimation results for the foreign direct investment and the private loans component of private flows, respectively The key factors influencing FDI are the GDP growth rate, the openness of the economy, and the variability of the real effective exchange rate. The first two factors exert a positive influence on private capital inflows, while real exchange rate fluctuations have a negative effect. The domestic investment ratio and the external debt ratio do not seem to have significant effects. In contrast, the regressions reported in Table 3 show that, in addition to a growing economy, the pivotal factors for obtaining private loans are the domestic investment rate and the ratio of external debt to GDP—both have coefficients that are statistically significant at the 1 percent level. The coefficients have the expected signs with an increase in the investment rate and a lowering of the external debt ratio making it easier to borrow abroad. As in Table 1, the test statistics in Table 2 and Tables 3 suggest that the time effects can be excluded from the estimated regressions.

Table 1.Total Private Capital Flows (PF)
Model IAModel IBModel ICModel IDModel I
PFt-10.330.280.270.250.25
(0.04)(0.04)(0.04)(0.04)(0.04)***
RGDPCt-1/1007.126.666.456.26
(2.40)(2.44)(2.41)(2.43)**
GFCFt-1 /1006.265.837.126.97
(1.99)(2.03)(2.04)(2.05)***
OPENt-1 /1001.032.142.13
(0.95)(1.00)(1.00)**
XDEBTt-1 /100–1.41–1.44
(0.42)(0.42)***
REERCVt /100–1.18
(1.56)
LOG-L–1,074.21,061.6–1,060.9–1,054.8–1,054.4
R20.310.350.350.370.37
Adjusted R20.240.270.270.290.29
Model 1: Test Statistics
LOG-LR2
(1) Constant term only–1,161.50
(2) Country dummies only–1,116.480.18
(3) Covariates: only–1,084.300.28
(4) Covariates and country dummies–1,061.010.35
(5) Covariates, country dummies, and time effects–1,054.990.37
Chisquaredd.f.Probability

value
FF-testDenominatorProbability

value
(2)vs.(1)90.02303.0930429
(3) vs. (1)154.40630.106454
(4) vs. (1)200.98364.7636425
(4) vs. (2)110.96621.056425
(4) vs. (3)46.59300.031.50304250.45
(5) vs. (4)12.04140.60.78144100.70
(5) vs. (3)58.62450.081.24454100.15

denotes significance at 1 percent level.

denotes significance at 5 percent level.

denotes significance at 10 percent level.

Note: Panel data regression with country dummy variables. Dependent variable = total private capital flows as a ratio of GDP. Time period = 1980–95. Number of countries = 31; number of observations = 461. Log-likelihood (LOG-L) with no explanatory variables = -1,161.50.

denotes significance at 1 percent level.

denotes significance at 5 percent level.

denotes significance at 10 percent level.

Note: Panel data regression with country dummy variables. Dependent variable = total private capital flows as a ratio of GDP. Time period = 1980–95. Number of countries = 31; number of observations = 461. Log-likelihood (LOG-L) with no explanatory variables = -1,161.50.
Table 2.Flows of Foreign Direct Investment (FDI)
Model IIAModel IIBModel IICModel IIDModel II
FDIt-10.290.270.220.220.21
(0.05)(0.05)(0.05)(0.05)(0.05)***
RGDPCt-1 /1003.973.003.002.71
(1.42)(1.41)(1.42)(1.42)*
GFCFt-1 /1000.23–0.82–0.82–1.07
(1.16)(1.17)(1.18)(1.18)
OPFNt-1 /1002.252.262.27
(0.56)(0.59)(0.59)***
XDEBt-1 /1000.000.06
(0.24)(0.24)
REERCVt /1001.92
(0.92)**
LOG-L–824.1–819.3–810.5–810.5–808.1
R20.400.420.440.440.44
Adjusted R20.340.350.370.370.37
Model II: Test Statistics
LOG-LR2
(1) Constant term only–943.61
(2) Country dummies only–853.990.32
(3) Covariates only–851.700.33
(4) Covariates and country dummies–817.220.42
(5) Covariates, country dummies, and time effects–808.620.44
Chisquaredd.f.Probability

value
FF-testDenominatorProbability

value
(2) vs. (1)179.24306.8130429
(3) vs. (1)183.83637.086454
(4) vs. (1)252.78366.3836425
(4) vs. (2)73.54614.816425
(4) vs. (3)68.95302.2830425
(S) vs. (4)17.20140.251.11144100.34
(5) vs. (3)86.15451.8745410

denotes significance at 1 percent level.

denotes significance at 5 percent level.

denotes significance at 10 percent level.

Note: Panel data regression with country dummy variables and time effects. Dependent variable = foreign direct investment flows as a ratio of GDP. Time period = 1980–95, Number of countries =31; number of observations = 461. Log-likelihood (LOG-L) with no explanatory variables = -943.6.

denotes significance at 1 percent level.

denotes significance at 5 percent level.

denotes significance at 10 percent level.

Note: Panel data regression with country dummy variables and time effects. Dependent variable = foreign direct investment flows as a ratio of GDP. Time period = 1980–95, Number of countries =31; number of observations = 461. Log-likelihood (LOG-L) with no explanatory variables = -943.6.
Table 3.Flow of Private Loans (PL)
Mlodel IIIAModel IIIBModel IIICModel IIIDModel III
PLt-10.310.260.270.240.23
(0.04)(0.04)(0.04)(0.04)(0.04)***
RGDPCt-1 /1002.963.683.453.64
(1.85)(1.88)(1.84)(1.85)**
GFCFt-1 /1005.846.517.817.98
(1.54)(1.57)(1.57)(1.58)***
OPENt-1 /100–1.42–0.40–0.39
(0.73)(0.75)(0.75)
XDEBTt-1 /100–1.37–1.34
(0.32)(0.32)***
REERCVt/1001.11
(1.19)
LOG-L–954.6–943.5–941.4–931.4–930.9
R20.270.310.310.340.34
Adjusted R20.190.230.230.260.26
Model III: Test Statistics
LOG-LR2
(1) Constant term only–1,028.870.00
(2) Country dummies only–1,003.930.10
(3) Covariates only–965.790.24
(4) Covariates and country dummies–940.320.32
(5) Covariates, country dummies, and time effects–931.510.34
Chisquaredd.f.Probability

value
FF-testDertominatorProbability

value
(2) vs.(1)49.89300.011.64304290.02
(3) vs. (1)126.17623.826454
(4) vs. (1)177.10364.2136425
(4) vs. (2)127.22625.116425
(4) vs. (3)50.94300.011.65304250.18
(5) vs. (4)17.63140.221.14144100.32
(5) vs. (3)68.56450.011.46454100.03

denotes significance at 1 percent level.

denotes significance at 5 percent level.

denotes significance at 10 percent level.

Note: Panel data regression with country dummy variables and time effects. Dependent variable = flow of private loans as a ratio of GDP. Time period = 1980–95. Number of countries = S1; number of observations = 461. Log-likelihood (LOG-L) with no explanatory variables = -1,028.9.

denotes significance at 1 percent level.

denotes significance at 5 percent level.

denotes significance at 10 percent level.

Note: Panel data regression with country dummy variables and time effects. Dependent variable = flow of private loans as a ratio of GDP. Time period = 1980–95. Number of countries = S1; number of observations = 461. Log-likelihood (LOG-L) with no explanatory variables = -1,028.9.

The disaggregated analysis presented in Tables 2 and 3 shows that the two main components of capital flows to sub-Saharan Africa are affected by different factors and that the results obtained at the aggregate level (Table 1) are a convoludon of these factors. Our results suggest that FDI, the component whose share in private flows during the period under consideration has been increasing, is attracted to growing open economies with relatively stable real effective exchange rates. Access to private loans, however, seems to be easier for growing countries with low levels of external debt relative to their GDP and higher rates of investment, presumably signaling appropriate utilization of resources in previous years and hence better growth prospects.

The effect of including international interest rates in the regression specifications for private flows as well as its two main components is reported in Table 4. In all three regressions, the coefficient on the three-year U.S. Treasury bond yield, our proxy for international interest rates, is not statistically significant. We experimented with other proxies—the three-month U.S. Treasury bill rate and the ten-year U.S. Treasury bond rate—but obtained similar results. This suggests that, at least for the sub-Saharan African countries, movements in international interest rates have not played an important role in determining foreign direct investment and the flow of private loans.13 It is possible that changes in international interest rates and the return on equities in industrial countries have considerable effect on portfolio equity flows to developing economies. However, in the case of sub-Saharan Africa where many borrowers are credit rationed (that is, have not reached a minimum level of creditworthiness) portfolio flows are not likely to be interest sensitive. As sub-Saharan African countries become creditworthy and begin to attract larger volumes of equity capital, movements in international interest rates may begin to exert greater influence on private flows to these countries.14 Further, since sub-Saharan Africa (excluding South Africa) has only recently seen any (recorded) portfolio equity flows, and that too, relative to other regions of the world, at a mere trickle, it will take some years before crosscountry data can be used to shed light on the determinants of such investments.

Table 4.Effect of International Interest Rates
Model IE (PF)Model IIE (FDI)Model IIIE (Pi)
PFt-10.24
(0.04)***
FDIt-10.20
(0.05)***
PLt-10.23
(0.04)***
RGDPCt-1 /1006.373.023.65
(2.32)***(1.37)**(1.78)**
GFCFt-1 /1006.82–1.458.30
(1.99)***(1.16)(1.55)***
OPENt-1 /1002.052.39–0.65
(0.97)**(0.58)***(0.74)
XDEBTt-1 /100–1.210.05–1.25
(0.40)***(0.23)(0.31)***
REERCVt /100–1.01–1.640.85
(1.51)(0.89)*(0.93)
USB3Yt /100–1.290.260.04
(5.26)(3.08)(4.06)
LOG-L–1,061.0–817.2–940.3
0.350.420.32
Adjusted R20.300.370.26

denotes significance at 1 percent level.

denotes significance at 5 percent level.

denotes significance at 10 percent level.

Note: Panel data regressions with country dummy variables. Time period = 1980–95. Number of countries = 31; number of observations = 461.

denotes significance at 1 percent level.

denotes significance at 5 percent level.

denotes significance at 10 percent level.

Note: Panel data regressions with country dummy variables. Time period = 1980–95. Number of countries = 31; number of observations = 461.

Concluding Remarks

Aid fatigue and fiscal pressures in the industrial countries have made it more difficult for developing countries to attract official capital flows. In such an environment, sub-Saharan Africa has no recourse but to tap private foreign capital to raise productivity levels necessary for sustained increases in living standards. With many Asian and Latin American countries growing rapidly and far ahead of most African countries in terms of putting in place financial infrastructure needed to efficiently absorb foreign capital, sub-Saharan African countries will have to undertake speedy policy and structural reform to attract private flows. Market discipline is likely to be severe in the initial stages, and countries that backtrack on reform will find that access to international capital is limited and that what is available will be on costlier terms.

While microeconomic factors—poor infrastructure, shaky banking systems, thin financial markets, weak regulatory frameworks, dearth of human capital, slow pace of privatization, corruption—are difficult to quantify, the macroeconomic factors used in our empirical analysis yield clear-cut conclusions. Using data for the period 1980–95, the paper demonstrates that in sub-Saharan Africa characteristics like output growth, openness, relative stability in real effective exchange rates, low external debt, and high investment rates have been favored by private capital flows. It further shows that the first three of these factors have been crucial for drawing in foreign direct investment and the last two factors combined with output growth have been particularly important for obtaining foreign private loans.

References

    CorboVittorio and PatriciaRojas1995“Exchange Rate Volatility Investment and Growth: Some New Evidence” (unpublished; Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    EasterlyWilliam and RossLevine1996“Africa’s Growth Tragedy: Policies and Ethnic Divisions” HIID Discussion Paper No. 536 (Cambridge, Massachusetts: Harvard Institute for International Development).

    • Search Google Scholar
    • Export Citation

    Fernandez-AriasEduardo1996“The New Wave of Private Capital Inflows: Push or Pull?” Journal of Development EconomicsVol. 48 (December) pp. 389418.

    • Search Google Scholar
    • Export Citation

    Fernandez-AriasEduardo and Peter J.Montiel1996“The Surge in Capital Inflows to Developing Countries: An Analytical Overview” World Bank Economic ReviewVol. 10No. 1 pp. 5177.

    • Search Google Scholar
    • Export Citation

    GhuraDhaneshwar and Michael T.Hadjimichael1996“Growth in Sub-Saharan Africa” Staff Papers International Monetary FundVol. 43No. 3 pp. 60534.

    • Search Google Scholar
    • Export Citation

    HadjimichaelMichael T.GhuraDhaneshwarMartin MühleisenRoger Nord and E. MuratUçer1995Sub-Saharan Africa: Growth Savings and Investment 1986–93 IMF Occasional Paper No. 118 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    HernandezLeonardo and HeinzRudolf1995“Sustainability of Private Capital Flows to Developing Countries: Is a Generalized Reversal Likely?”Policy Research Working Paper No. 1518 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    KasekendeLouisDamoniKitabire and MatthewMartin1996“Capital Inflows and Macroeconomic Policy in Sub-Saharan Africa” Jerome Levy Economics Institute Working Paper No. 158 (Annandale-on-Hudson, New York: Bard College).

    • Search Google Scholar
    • Export Citation

    MontielPeter j.1995“Financial Policies and Economic Growth: Theory Evidence and Country Specific Experience from Sub-Saharan Africa” AERC Special Paper No. 8 (Nairobi: African Economic Research Consortium).

    • Search Google Scholar
    • Export Citation

    PierceR.A.Islam and K.Sauvant1992“The Determinants of Foreign Direct Investment A Survey of Empirical Evidence” report for United Nations Centre on Transnational Corporations (New York: United Nations).

    • Search Google Scholar
    • Export Citation

    SachsJeffrey D. and Andrew M.Warner1996“Sources of Slow Growth in African Economies” HIID Discussion Paper No. 545 (Cambridge, Massachusetts: Harvard Institute for International Development).

    • Search Google Scholar
    • Export Citation

    SinghHarinder and KwangJun1995“Some New Evidence on Determinants of Foreign Direct Investment in Developing Countries,”Policy Research Working Paper No. 1531 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    UNCTAD1996“Foreign Direct Investment in Africa” UNCTAD/DTCI/19 Current Studies Series ANo. 28 (Geneva: United Nations).

    World Bank1994Adjustment in Africa: Reforms Results and the Road Ahead (New York: Oxford University Press).

    World Bank1996World Debt Tables 1995–96: External Finance for Developing Countries (Washington: World Bank).

List of Participants*

Ahmed, Masood

Director, International Economics Department, The World Bank

Ajayi, S. Ibi

Professor of Economics, University of Ibadan

Bates, Robert H.

Professor of Government, Harvard University

Bhattacharya, Amar

Economic Advisor, International Economics Department, The World Bank

Boorman, John

Director, Policy Development and Review Department, International Monetary Fund

Boote, Anthony

Division Chief, Policy Development and Review Department, International Monetary Fund

Buffie, Edward F.

Professor of Economics, Indiana University

Bradford, Colin

Chief Economist, Office of the Administrator, U.S. Agency for International Development

Camdessus, Michel

Managing Director, International Monetary Fund

Claessens, Stijn

Principal Economist, East Asia and Pacific Regional Office, The World Bank

Cline, William R.

Deputy Managing Director, Institute of International Finance, Washington

Desai, Lord M.

Professor, London School of Economics

Detragiache, Enrica

Economist, Research Department, International Monetary Fund

Dollar, David

Chief, Macroeconomics and Growth Division, Policy Research Department, The World Bank

Dooley, Michael

Professor, Department of Economics, University of California, Santa Cruz

Elbadawi, Ibrahim

Research Coordinator, African Economic Research Consortium Nairobi

Fellgert, Robin

Head, Developing Countries, Debt, and Export Finance, Her Majesty’s Treasury, United Kingdom

Fisher, Matthew

Division Chief, Policy Development and Review Department, International Monetary Fund

Gwin, Catherine

Senior Vice President, Overseas Development Council, Washington

Haas, Jerome

Secretary General, Paris Club

Iqbal, Zubair

Advisor, IMF Institute

Kanbur, Ravi *

Principal Economic Advisor, Office of the Senior Vice President and

Chief Economist, The World Bank

Kasekende, Louis

Executive Director of Research, Bank of Uganda Kampala

Khan, Mohsin

Director, IMF Institute

Kilby, Frederick

Economic Advisor, International Economics Department, The World Bank

Killick, Anthony

Senior Research Fellow, Overseas Development Institute, London

Lin, Soe

Chief Economist, Canadian International Development Agency, Ottawa

Montiel, Peter

Professor of Economics, Williams College

Mwanza, Jacob

Governor, Bank of Zambia, Lusaka

Ndulu, Benno J.

Executive Director, African Economic Research Consortium, Nairobi

Ndung’u, Njuguna

Lecturer, Economics Department, University of Nairobi

Niepold, John

Portfolio Manager for Africa, Emerging Market Investors, Arlington, Virginia

Ouattara, Alassane

Deputy Managing Director, International Monetary Fund

Pfeffermann, Guy

Director and Chief Economic Advisor, International Finance Corporation

Sandström, Sven

Managing Director, The World Bank

Schadler, Susan

Senior Advisor, Policy Development and Review Department, International Monetary Fund

Sharma, Sunil

Economist, Research Department, International Monetary Fund

Stevens, Simon

Research Officer, Overseas Development Institute, London

Thugge, Kamau

Senior Economist, Policy Development and Review Department, International Monetary Fund

van Trotsenburg, Axel

Head, HIPC Implementation Unit, Africa Regional Office, The World Bank

West, Gerald

Senior Advisor, Guarantees, Multilateral Investment Guarantee Agency

Wickham, Peter

Division Chief, Research Department, International Monetary Fund

Williamson, John

Chief Economist, South Asia Regional Office, The World Bank

Wong, Chorng-Huey

Assistant Director, IMF Institute

The authors thank Edward Buffie and Tom Krueger for comments. Brooks Calvo, Jill Dooley, Manzoor Gill, and Jos Jansen provided excellent research assistance.
1Of the 47 sub-Saharan African countries, those with positive average rates of per capita real GDP growth during 1988–95 were Botswana, Burkina Faso, Cape Verde, Chad, Equatorial Guinea, Gabon, Ghana, Guinea, Guinea-Bissau, Lesotho, Mali, Mauritius, Mozambique, Namibia, Nigeria, Seychelles, Sudan, Swaziland, Tanzania, and Uganda.
2A large part of the increase in 1995/96 is due to a reclassification of private transfers. Although data for previous years have not been adjusted, the increase in FDI is believed to have taken place largely after 1993.
3Survey of Current Business (various issues).
4This discussion is partly based on interviews with selected commercial bankers, investment bankers, and mutual fund managers in New York and London. See UNCTAD (1996) fur a detailed review of foreign direct investment in Africa, and also Pierce and others (1992) and Singh and Jun (1995) for reviews of and evidence for determinants of foreign direct investment in developing countries more broadly.
7The annual data on capital flows was obtained from the World Bank’s Debtor Reporting System and the information on the macroeconomic variables from the IMF’s World Economic Outlook and International Financial Statistics databases.
8”Small” countries were defined to be those with population in 1991 of less than 500,000. As in World Bank (1994), these countries were excluded because external aid has a disproportionate effect on their macroeconomic performance.
9See, Kasekende, Kitabire, and Martin (1996) and footnote 1 above.
10For a stock adjustment interpretation of the regressions, see for example, Hernandez and Rudolf (1995) and Fernandez-Arias (1996).
11For a discussion or” trends in growth, savings, and investment in sub-Saharan Africa, see Hadjimichael and others (1995), Easterly and Levine (1996), Ghura and Hadjimichael (1996), and Sachs and Warner (1996).
12The usual Hausman tests for fixed versus random effects decisively rejected the latter.
13ldeally, we would like to include the differential between a representative domestic interest rate and an international interest rate. However, for most of the sub-Saharan African countries in our sample such a series is difficult, if not impossible, to construct.
*This list includes only authors of papers, speakers, presenters of papers, and discussants at the conference. The titles and affiliations are those that were in effect at the time of the conference (December 1996).
*Ravi Kanbur has become professor of world affairs at Cornell University.

    Other Resources Citing This Publication