2. Nigeria and South Africa: Spillovers to the Rest of Sub-Saharan Africa

International Monetary Fund. African Dept.
Published Date:
October 2012
  • ShareShare
Show Summary Details

Introduction and Summary

This chapter examines the economic linkages between sub-Saharan Africa’s two largest economies, Nigeria and South Africa, and the rest of the region, and explores how developments in these countries can affect other countries in the region. Nigeria is an important export market only for a few neighboring countries, but financial linkages with countries further afield are growing with the regional expansion of Nigerian banks. Porous borders mean that trade with its neighbors is heavily influenced by tax and subsidy policies in Nigeria; and closely linked food markets ensure that inflation in neighboring countries is significantly affected by inflation developments in Nigeria. South Africa’s linkages to the rest of sub-Saharan Africa are much larger and steadily intensifying, but are currently of macroeconomic significance mainly within the Southern African Development Community (SADC).

In the past few years, the IMF has conducted extensive research on how economic shocks and policy changes in each of the world’s major economies affect other countries in the international economy.1 This ‘spillover analysis’ has shed significant light on the importance of interconnections among the different economies of the world and the mechanisms through which shocks in one country are propagated to other countries, often at high speed.

This chapter seeks to extend the analysis to sub-Saharan Africa, examining the linkages between the region’s two largest economies and the rest of the region, and exploring the significance of spillovers from these regional giants to their neighbors. This analysis is a prerequisite for addressing such issues as: (i) how does a slowdown in South Africa, originating domestically or abroad, affect the rest of the region, who is affected, and why? (ii) how close are the economic links between Nigeria and its neighbors, and, given these links, how do developments in Nigeria affect the rest of West Africa? and (iii) how important are counter-cyclical policies in South Africa in insulating other parts of the continent from adverse shocks? Between them, the economies of Nigeria and South Africa account for one-half of sub-Saharan Africa’s GDP, and are potentially major drivers of growth for the region as a whole.

Historically, the economic structures of the colonial era yielded relatively few intra-African economic linkages, with the key economic connections being those between the colony and the colonial power. Since the 1960s, many efforts have been made to promote economic integration within Africa, including through a proliferation of free trade agreements and, more recently, moves toward promoting monetary integration beyond the existing common currency zones. Although sub-Saharan Africa’s trade links with Europe, the United States, and, increasingly, developing Asia still far outweigh those inside the region, intra-African trade and financial linkages have been expanding more rapidly in recent years and should grow even faster in the years ahead. It is a good time to explore how these developments are affecting the macroeconomic interconnections within the region.

Spillover Channels

Developments in one economy can spillover to other countries through several channels, depending on the depth of the underlying economic linkages. Key channels include: (i) trade in goods and services, both formal and informal; (ii) financial sector interconnections; (iii) flows of capital, whether in the form of foreign direct investment, portfolio flows, or loans; and (iv) labor movements and (in the reverse direction) remittance flows.

Institutional factors can also play an important role: examples include the revenue-sharing arrangements in the Southern African Customs Union (SACU); the regional bond market that has been established in the West African Economic and Monetary Union (WAEMU); and the exchange rate arrangements of the Common Monetary Area in southern Africa, where the three smaller member countries have long-standing exchange rate pegs to the rand.

Quantifying the significance of these channels in sub-Saharan Africa is challenging given data limitations: trade statistics fail to capture what are often large volumes of unrecorded informal trade; data on capital flows and stocks are often of very poor quality; and information on labor flows and remittances typically understate the scale of activity involved by sizable margins. That said, some facts can be noted:

  • South Africa plays a significant role in the structure of intra-sub-Saharan African trade. Recorded exports to South Africa exceed ++1 percent of domestic GDP for at least a dozen countries, with links most noticeable for countries in the SADC sub-region (Box 2.1).
  • Some clustering of trade flows can also be seen between Nigeria and its closest neighbors and within eastern Africa. The large amount of informal (unmeasured) cross-border trade in these sub-regions, particularly in agricultural goods, suggests closer ties and linkages than indicated by official trade statistics.
  • Although trade within the region remains modest as a share of countries’ total trade, the ratio of intra-regional trade to national GDP has generally risen significantly in the past decade (Figure 2.1). Going forward, improved regional infrastructure and vigorous implementation of existing free trade agreements—including the use of less restrictive rules of origin and reductions in non-tariff barriers—would likely produce a further sharp increase in the scale and importance of such trade.
  • The expansion of investment within the region by South African companies and institutions—both financial and nonfinancial—has brought with it a deepening of trade and other linkages within sub-Saharan Africa, while also helping to diversify the market orientation of South African exports.
  • Banking groups headquartered in South Africa and Nigeria have rapidly expanded their operations across the region in recent years—although the business models used (focusing on local funding and lending) may act to contain the scope for financial contagion within the region.2
  • Although available data on remittances suggest quite modest financial contributions from migrant workers to their home countries, estimates of migration across African borders point to large, mostly informal flows. Adverse shocks to host countries, such as South Africa and Nigeria, would likely lead to a fall-off in transfers in cash and in kind to the migrants’ home countries and at least some return of migrants.
  • Domestic policies such as highly distortive tax and subsidy regimes, and institutional arrangements such as customs revenue-sharing arrangements, can play important roles in transmitting shocks between countries.

Figure 2.1.Sub-Saharan Africa: Intra-Regional Exports, 1981–2010

Source: IMF Direction of Trade Statistics.

Box 2.1.The Role of Nigeria and South Africa in the Network of Sub-Saharan African Merchandise Trade

Tools for network analysis can shed some light on the complex underlying patterns of bilateral trading relationships across sub-Saharan Africa and the relative importance of sub-Saharan African countries.1 One such tool—used in the two figures here—is the specialized software for network analysis called NodeXL, which facilitates the computation of network statistics and helps visualize underlying structures.

In Figure 1, all intraregional exports larger than 1 percent of the exporter’s GDP are identified by lines connecting the exporter to the relevant importer (indicated by an arrow). Trade linkages of this magnitude point to the potential for a non-negligible macroeconomic impact on the exporter if economic conditions in the importing country deteriorate. The size of each country’s bubble indicates the number of countries for which it is a significant export destination, while the thickness of each line indicates the size of its bilateral exports relative to GDP. The figure confirms the key importance of South Africa as an intraregional export destination, with Nigeria a fairly distant second. Exports to South Africa exceed 1 percent of GDP for 13 sub-Saharan African countries—most of which are members of the Southern Africa Development Community. South Africa’s partners in the Southern Africa Customs Union figure prominently; Swaziland shows the largest intraregional exposure, with exports to South Africa at close to 30 percent of GDP.

Figure 1.Intraregional Trade Linkages

Figure 2 includes global as well as intraregional trade linkages, with South Africa’s main connections shown in blue (global exports) and red (intra-sub-Saharan African imports). Countries within the euro area are the most important export destination for 34 sub-Saharan African countries, making the euro area the center of sub-Saharan Africa’s external networks; China and the United States are key markets for 19 sub-Saharan African countries, while India plays a significant role for 15 economies.

Figure 2.Global and Intraregional Trade Linkages

Sources: IMF Direction of Trade Statistics; United Nations Comtrade database; country authorities; and IMF staff estimates.

For each country, bubble size is proportional to the number of countries for which it is an important export destination.

The arrows point to the receiving country.

Line thickness indicates export values relative to the exporter’s GDP.

Red (blue) lines indicate exports to (from) South Africa.

Green lines represent other bilateral trade above 1 percent of exporter’s GDP.

This box was prepared by Jorge Iván Canales-Kriljenko and Estelle Liu.1 Network analysis is a series of techniques to characterize and visualize the ways in which elements of a system are interconnected (Errico and Massara, 2011).

Key Findings

Examination of the linkages between South Africa and the rest of the region point to large spillover effects from South Africa to its immediate neighbors in the SACU—but the rich tapestry of linkages among SACU member countries ensures that the net impact of faster growth in South Africa on other SACU members is not straightforward. For a set of countries close to the SACU area, export demand, direct investment, and, in some cases migration flows ensure some spillover effects from South Africa, although the impact of these spillovers is typically modest. Further afield, developments in South Africa appear to have limited impact on the rest of sub-Saharan Africa.

Nigeria’s development model has resulted in little demand for products from other sub-Saharan African countries, aside from immediate neighbors—where porous borders produce complex trade flows that are heavily influenced by tax and subsidy policies in Nigeria. Inflation in these neighbors is sensitive to inflation developments in Nigeria.

South Africa’s Impact on Sub-Saharan Africa

South Africa is the largest economy in sub-Saharan Africa, accounting for nearly 30 percent of its GDP. This section looks at the main economic linkages between South Africa and the rest of sub-Saharan Africa and seeks to assess the significance of spillovers from developments in South Africa to the rest of the region.

Trade Linkages

South Africa is an important export destination for a sizable number of countries in sub-Saharan Africa: recorded merchandise exports to South Africa in 2010 amounted to more than one percent of GDP for 13 countries.3 South Africa’s partners in the SACU—neighbors in a single customs union—are the most reliant on South African markets as an export destination, but the South African market is also important for several countries further afield, mainly in the broader SADC region (Figure 2.2). For products sold in integrated global markets (e.g., many commodities, including oil), the product’s sale to South Africa does not imply significant dependence on South African demand, because demand contraction there would merely result in diversion of sales to other countries.4 By contrast, for goods that are differentiated (many manufactured products) or sold in markets segmented by transport costs or trade barriers, reliance on South African markets does create exposure to fluctuations in South African demand—and, by extension, to adverse spillovers from a downturn in the South African economy.

Figure 2.2.Sub-Saharan Africa: Exports to South Africa from Neighboring Countries

(Percent of GDP)

Source: IMF, Direction of Trade and Statistics; IMF, Worl Economic Outlook; United Nations Comtrade database; and countries authorities.

The absence of large trade flows with South Africa does not imply the absence of spillover effects to countries that produce commodities for which South Africa constitutes an important source of global supply. In such cases (for example, platinum, gold), supply developments in South Africa will influence export prices for other producers of these goods in sub-Saharan Africa—with potentially significant macroeconomic effects on these economies.

South Africa features more strongly as a supplier of imports than as a destination for exports for almost all sub-Saharan African countries.5 In aggregate, South African exports to sub-Saharan African countries amount to 6.4 percent of South Africa’s GDP, about double its imports from the region. Looking at this trade in more detail:

  • In nearly half of the region’s economies (21 countries), imports from South Africa amount to more than 1 percent of the importer’s GDP. Imports from South Africa account for the bulk of imports in other SACU members (BLNS: Botswana, Lesotho, Namibia, Swaziland) and amount to at least 10 percent of domestic GDP in Malawi, Zambia, and Zimbabwe.
  • Exports to the rest of sub-Saharan Africa account for roughly 25 percent of South Africa’s merchandise exports, consisting in large part of manufactures. That said, exports to the region are quite diversified across countries, implying that adverse shocks to individual sub-Saharan African economies (as distinct from, say, the sub-Saharan Africa or SADC regions as a whole) would have only a very modest effect on South Africa.6

Direct Investment Linkages

South Africa is an important source of foreign direct investment (FDI) in sub-Saharan Africa, having increased sharply over the past decade. By 2011, the value of South Africa’s direct investment in the region had reached 6 percent of South African GDP—almost one-fourth of total recorded foreign investment by South Africa (Figure 2.3). Investments in Mauritius (in large part as a gateway to third countries) and Nigeria have been particularly important in increasing South Africa’s exposure in sub-Saharan Africa.

Figure 2.3.South Africa: Outward Direct Investment

Source: South African Reserve Bank.

South African FDI stocks are large (at least 10 percent of host country GDP) in Mauritius, Mozambique, Swaziland, and Zimbabwe (United Nations, 2005); significantly more modest in such countries as the Democratic Republic of the Congo, Ghana, Malawi, Nigeria, Tanzania, Uganda, and Zambia. The range of South African companies with a presence in the region is diverse, and includes banks and other financial institutions, large-scale retailers, mining groups, and selected manufacturing companies (Tables 2.1 and 2.2).

Table 2.1.Sub-Saharan Africa: Key Nonfinancial South African Firms Operating in the Region
Source: South Africa financial firms’ annual reports.
Source: South Africa financial firms’ annual reports.
Table 2.2.Sub-Saharan Africa: Key South Africa-based Financial Groups Operating in the Region(Share of deposits 2010, where available)
Sources: South Africa financial firms’ annual reports, BankScope; and Bankers Almanac.

ABSA is a South Africa-based institution that is majority-owned by the Barclays group (UK).

Nedbank is a South Africa-based institution that is majority-owned by the Old Mutual group (UK).

Sources: South Africa financial firms’ annual reports, BankScope; and Bankers Almanac.

ABSA is a South Africa-based institution that is majority-owned by the Barclays group (UK).

Nedbank is a South Africa-based institution that is majority-owned by the Old Mutual group (UK).

South African-based banks have opened subsidiaries in 16 countries in sub-Saharan Africa, in part following their clients; in 11 of those countries, South African subsidiaries are among the five largest banks. Most of these banks fund their operations with local deposits; cross-border loans are typically modest, with a few exceptions. The scale of both cross-border lending to and deposit-taking from other sub-Saharan African countries by South Africa-based banks is small in relation both to South African GDP (about 0.5 percent) and to the assets and liabilities of the South African banking system.

Nonbank financial services, including insurance and wealth management, are also provided across the region by South African firms, while the spread of large South African retail companies has provided a significant conduit for goods and services from South Africa, mainly in the SADC region.

South African-based mining companies have also expanded operations into several countries in the region, in part reflecting declining mineral stocks in South Africa itself. Ownership in these firms, typically listed on several exchanges, including elsewhere in sub-Saharan Africa, is highly diversified.

Portfolio Investment

According to the IMF’s portfolio investment survey, the stock of cross-border portfolio investments between South Africa and the rest of sub-Saharan Africa exceeds 1 percent of the partner’s GDP only in Mauritius and Namibia. The correlation between stock prices in the Johannesburg Stock Exchange (JSE) and other price indices in the region is positive. All price indices tend to respond in a similar fashion to changes in global risk aversion, although the impact is larger in South Africa than elsewhere. However, despite dual listing of some key stocks on the JSE and other stock exchanges in the region, the only stock exchange that (statistically) reacts to shocks originating in South Africa is that of Namibia (Box 2.2).

Remittances from South Africa

South Africa houses many immigrants and temporary workers from sub-Saharan Africa.7 However, as elsewhere in the world, up-to-date information on their numbers, and the remittances they send back to their families, is incomplete. Available estimates only identify macroeconomically significant remittances to Lesotho (20 percent of its GDP in 2010) and Swaziland (3 percent).8

In Lesotho, the remittances stem from the compensation of Basotho employees working in the mining sector. Although large, these income receipts have fallen substantially relative to GDP in the last five years, reflecting policies to reduce the structurally high unemployment in South Africa that favor local employment, and also the trend decline in gold production. Although hard data are not available, remittances from South Africa to Zimbabwe are likely to be sizable, given the number of Zimbabweans estimated to be living and working in South Africa.

The Sharing of SACU Customs Revenue

The mechanism for sharing pooled customs revenue among the five members of the SACU results in strong linkages between South African import levels (accounting for about 95 percent of total SACU imports) and budget revenue in the other SACU member countries (BLNS). The mechanism in place provides large revenue flows to these countries (about half of total SACU tariff revenue) that are closely linked to South African import levels, typically with a lag of 1 to 2 years.

Because imports are usually more volatile than overall economic activity, the revenue sharing formula contributes to significant volatility in budgetary revenue in the BLNS—most marked in the cases of Lesotho and Swaziland, where the SACU revenues play a central role in the revenue base (Figure 2.4). Failure to manage this revenue volatility appropriately—the behavior of which is akin to budget receipts from the mineral sector in resource-rich countries—can create severe fiscal stresses. Measures to handle this volatility can include fiscal buffers during good times, the establishment of stabilization funds, or the use of revenues to prefund contingent obligations (see Cuevas, A., L. Engstrom, V. Kramarenko, and G. Verdier, forthcoming.).9

Figure 2.4.BLNS: Customs Revenue payments1

Sources: Country authorities.

1 Payments received based on forecast customs receipts and shares of interegional trade.

The Common Monetary Area

With the exception of Botswana, BLNS countries have long-standing one-to-one currency pegs to the rand. With effectively free capital movements and wide circulation of the rand, the BLNS countries effectively import monetary and exchange rate policy from South Africa10—benefiting from the credibility of South Africa’s monetary policy regime and associated interest rate levels, but also experiencing significant currency volatility vis-a-vis the rest of the world. South African policy actions, through the impact on both interest rates and the value of the rand, have an immediate impact on economic conditions in the smaller CMA members—as indeed do developments in global financial markets, which have significant effects on the rand.11

Spillover Effects: Are they Significant?

The discussion above has identified various economic linkages between South Africa and the rest of sub-Saharan Africa—linkages that are closest within the southern Africa sub-region, but extend somewhat further, including via terms of trade effects, to countries exporting (mineral) products for which South Africa is an important producer. How important are these linkages in propagating shocks between South Africa and the rest of sub-Saharan Africa?

In terms of connectedness to South Africa, the region can arguably be split into three groupings:

  • South Africa’s immediate neighbors in the SACU—Botswana, Lesotho, Namibia, and Swaziland (BLNS)—where trade, investment, financial, and institutional ties are strong.
  • A group of countries mainly in the SADC that includes Malawi, Mauritius, Mozambique, Zambia, and Zimbabwe, for whom direct investment ties are of some significance; South Africa is a market destination of some importance for exports of goods and services; and migration flows and remittances are important in some cases (Malawi, Zimbabwe).
  • The rest of sub-Saharan Africa, where there are some investment linkages (especially in anglophone countries), but where monetary conditions are unaffected by developments in South Africa; where dependence on the South African market as an export destination is minimal;12 and where labor and remittance flows are not significant.

Box 2.2.Linkages among Sub-Saharan Africa Stock Exchanges

Stock price indices in sub-Saharan Africa are subject to spillovers from changes in global investor sentiment, but do not seem to have a significant influence on stock price indices in other parts of the region. The exceptions are stock prices from the Namibia stock exchange, perhaps reflecting dual listings.

The Johannesburg Stock Exchange (JSE) is not only the largest stock exchange in sub-Saharan Africa, but is also among the 15 largest in the world. It has close connections to the London stock exchange, including dual listings of some of its largest stocks. U.S. investors hold a significant amount of the stock value on the JSE. Although global investors have started paying closer attention to the stock exchanges in other sub-Saharan African frontier markets (Figure 1), the market capitalization and trading in these exchanges remains small compared to that of the JSE, in which stock trading exceeded 90 percent of South African GDP in 2011 (Table 1).1

Figure 1.Sub-Saharan Africa: Equity Flows, 2004–121

Sources: EPFR Global.

1 Other Sub-Saharan African countries for which data is available include Botswana, Ghana, Côte D’Ivoire, Kenya, Malawi, Mauritius, Namibia, Nigeria, Zambia, and Zimbabwe.

Table 1.Sub-Saharan Africa: Selected Indicators of Regional Equity Markets, 2011
Market capitalization

(Percent of GDP)
Stocks traded total


(Percent of GDP)


Firms listed at

the JSE

South Africa21091.239.8
Source: World Bank, World Development Indicators.
Source: World Bank, World Development Indicators.

To assess whether stock prices at the JSE play an independent role in determining stock prices in the other sub-Saharan Africa stock exchanges, we estimated, for 1999M1–2012M7, a system of equations to model sub-Saharan Africa stock prices as in Saadi and Williams (2011). We found that stock prices in all sub-Saharan African exchanges tend to increase when global risk aversion (as proxied by the Chicago Board of Options Exchange Volatility Index, VIX) falls, and vice versa. The only stock exchange that also reacts to regional shocks originating from the JSE is the Namibia stock exchange, which likely reflects the high degree of cross-listings between these two exchanges.

This box was prepared by Jorge Iván Canales-Kriljenko and Oral Williams.1 For a description of developments, issues, and challenges in stock market developments in sub-Saharan Africa, see Yartey and Adjasi (2007).

Box 2.3.Spillovers within the Southern Africa Customs Union

Outward spillovers from South Africa to the other members of the Southern Africa Customs Union1––Botswana, Lesotho, Namibia, and Swaziland (BLNS)––appear to be substantial. Customs revenue is heavily affected by the level of South African imports and accounts for a substantial amount of fiscal revenue in BLNS. Lesotho, Namibia, and Swaziland peg their currencies to the South African rand, in practice adopting monetary policy decisions from the South African Reserve Bank. South Africa is the main provider (80–90 percent) of imports to BLNS and the main destination of the exports of Swaziland (44 percent), Lesotho (42 percent), and Namibia (27 percent). Remittances from South Africa to Lesotho, and to a lesser degree Swaziland, amount to a substantial proportion of GDP. Moreover, South African financial groups are the dominant players in the financial markets of the region, spanning banking, pension, insurance, and wealth management services, whose treasury and risk management decisions tend to be centralized in Johannesburg. These institutions may help channel foreign direct investment and exports from South Africa into the rest of the region.

However, it is difficult to identify systematic linkages between fluctuations in growth rates in South Africa and in BLNS. Simple bilateral contemporaneous correlations of growth rates are low for Botswana and Namibia and even negative for Lesotho and Swaziland.2 Furthermore, dynamic panel regressions that control for global developments and fiscal policies, such as those reported in Table 2, find no systematic evidence that South Africa’s growth helps explain the evolution of growth in BLNS.3 The results show instead that economic activity in BLNS depends heavily on real and financial external conditions, and that national macroeconomic policies make a difference. On average, a 1 percentage point change in world growth is associated with a change of 0.6 percentage point in growth in BLNS and a 1 percentage point decrease in the fiscal balance increases growth by 0.25 percent point, consistent with the successful implementation of countercyclical fiscal policies.

Table 1.South Africa’s Share in the BLNS Total exports and Imports (2000–11)
(percent of total)
(percent of GDP)
Source: United Nations COMTRADE database.

Based on 2000–08 trade data.

Based on 2000–07 trade data.

Source: United Nations COMTRADE database.

Based on 2000–08 trade data.

Based on 2000–07 trade data.

Table 2.Panel-GMM Regression on Real GDP Growth in BNLS Countries, 1986-2010
Real GDP growth (t-1)-0.16(-1.24)
World growth0.63(3.78)***
Terms of trade0.03(6.58)***
Fiscal deficit0.26(6.19)***
TED spread-1.37(-3.56)***
AR(1) test: p-value0.07
AR(2) test: p-value0.14
Sargan test: p-value0.49
Source: IMF staff calculations.Note: *, **, and *** indicate significance at the 10, 5, and 1 percent levels, respectively.
Source: IMF staff calculations.Note: *, **, and *** indicate significance at the 10, 5, and 1 percent levels, respectively.

Vector-autoregressive (VAR) models for the individual BLNS countries produce similar results in terms of the impact of world growth, but shed a little more light on the possible role of developments in South Africa. Impulse responses from these VARs imply that events affecting world growth spill over onto GDP growth in Botswana, Namibia, and Swaziland, but not Lesotho. As in the panel regressions, after controlling for world growth, South Africa’s real GDP growth does not seem to contribute much to GDP growth in Botswana, Namibia, and Swaziland. But, in Lesotho, growth seems to fall when it increases in South Africa. While this may sound counterintuitive, it could reflect the fact that workers from Lesotho migrate to South Africa when the South African economy is booming (lowering domestic production), and come back when South Africa is facing rough times. The impulse responses also suggest that a real appreciation in South Africa tends to reduce growth in Botswana—perhaps by hurting Botswana’s competitiveness—but to increase it in Swaziland. It appears insignificant for Lesotho and Namibia. These results also hold for real demand growth.

In short, although developments in South Africa have significant spillovers in the region through a variety of channels, direct growth linkages with BLNS are difficult to identify. The financial and external environments (as well as national policies) are much more closely linked to output fluctuations. Given that financial systems in BLNS are much less developed than in South Africa, one possibility is that South Africa acts as a conduit in the transmission of external shocks, but it does not amplify or cushion them.

This box was prepared by Jorge Iván Canales Kriljenko, Farayi Gwenhamo, and Saji Thomas.1 The Southern African Customs Union is the oldest customs union in the world, and it includes Botswana, Lesotho, Namibia, South Africa, and Swaziland.2 Correlations with other countries in SADC, including Angola and Tanzania, are much larger than with BLNS, perhaps reflecting other common factors, such as world growth.3 The controls include world growth, oil prices, interest rate spreads, fiscal balances, and country fixed effects.

The significance of spillovers from South Africa to these three country groupings differs quite markedly.

  • For BLNS, spillover effects are significant, albeit less straight forward than simple intuition would suggest (Box 2.3). BLNS depend in good part on export markets beyond South Africa, which may or may not be closely linked to shocks to the South African economy. Labor links can contribute to negative output correlations, as the “pull” of workers into a growing South African economy reduces labor supply, and potentially output, in neighboring countries. Finally, as noted earlier, the SACU revenue sharing formula introduces close links between South African import levels and BLNS fiscal positions, but these links operate with a lag.
  • For the second grouping of countries, spillovers from South Africa are of some significance, via export demand effects, corporate direct investment, and, in some cases, migration flows (Table 2.3).
  • For the third grouping of countries, spillover effects from South Africa are likely of a second or third order, given the lack of propagation mechanisms—except for countries that compete in export markets where South Africa is a significant producer that can directly influence world market prices.
Table 2.3.Sub-Saharan Africa: Key Linkages with South Africa, 2010(Percent of GDP, unless otherwise indicated)
Sources: IMF’s Direction of Trade Statistics, World Economic Outlook, United Nations Comtrade database, and country authorities.Color code: White (below 1 percent), green (between 1 and 10 percent), orange (between 10 and 20 percent), red (above 20 percent).

The sum of the first three columns as a percentage of total current account revenues in 2010. Because bilateral data on trade services and other current account revenues are not available, the ratio should be taken to be a lower bound.

Sources: IMF’s Direction of Trade Statistics, World Economic Outlook, United Nations Comtrade database, and country authorities.Color code: White (below 1 percent), green (between 1 and 10 percent), orange (between 10 and 20 percent), red (above 20 percent).

The sum of the first three columns as a percentage of total current account revenues in 2010. Because bilateral data on trade services and other current account revenues are not available, the ratio should be taken to be a lower bound.

We use statistical techniques here to examine more closely the links between output movements in South Africa and the rest of sub-Saharan Africa. First, analysis of bilateral contemporaneous correlations in sub-Saharan Africa (Figure 2.5) sheds little light, as it is clear that many of the high correlation cases can be explained by external factors, such as the strength of world commodity prices (e.g., the relatively high correlations between output growth in South Africa and, respectively, Angola and the Democratic Republic of the Congo). By contrast, output correlations between South Africa and its SACU partners are low, notwithstanding the exceptionally close linkages and varying spillover effects from South Africa to BLNS.

Figure 2.5.Bilateral Correlations of Output in Sub-Saharan Africa

Real GDP

Sources: IMF World Economic Outlook database, and Fund staff estimates.

The figure shows the bilateral correlations of annual GDP growth in sub-Saharan Africa that exceed 0.5. For each country, the bubble size is proportional to the number of countries for which the bilateral correlation exceeds the threshold. Red lines indicate countries correlated with South Africa.

To move beyond a simple comparison of correlations, we examine econometrically the determinants of growth in income per capita among the countries in the first two groups above which were identified as having significant direct economic linkages with South Africa. As discussed in Appendix, the econometric analysis suggests that per capita income growth in these economies is, unsurprisingly, influenced by both global developments and domestic policy variables. But there is no strong evidence that the inclusion of variables relating to South African developments and policies adds to the explanatory power of the regressions. The assertion that South African developments have significant contemporaneous effects on growth in the rest of sub-Saharan Africa should therefore be treated with caution.

Nigeria’s Impact on its Neighbors

Nigeria is the second-largest economy in sub-Saharan Africa, accounting for almost 20 percent of the region’s GDP, and 18 percent of its population. It is also the region’s largest oil exporter. This section looks at the various economic links between Nigeria and the rest of sub-Saharan Africa focusing on Nigeria’s closest neighbors (Benin, Burkina-Faso, Cameroon, Chad, Côte d’Ivoire, Ghana, Niger, and Togo), with which the linkages appear to be most significant.

Formal Trade Linkages

Although Nigerian imports account for only a small fraction of sub-Saharan Africa’s aggregate imports, Nigeria is a significant export market for a number of countries in West Africa. Nigeria was the destination for only 1 percent of the region’s total exports during 2008–10, but exports to Nigeria were sizable in relation to GDP for Niger (6.3 percent), Togo (3.8 percent), Côte d’Ivoire (3.3 percent), and Benin (1.7 percent) (Figure 2.6). Niger’s exports to Nigeria consist mainly of food and animals; Togo exports manufactured goods; and Côte d’Ivoire exports chemicals and other manufactured products. Nigeria is a significant supplier of goods—mainly petroleum products—to Equatorial Guinea (11.9 percent of GDP), Côte d’Ivoire (7.5 percent), Ghana (4.6 percent), and Cameroon (4.2 percent), as well as to the Democratic Republic of the Congo (17.5 percent).

Figure 2.6.Western Africa: Exports to Nigeria

(Percent of GDP)

Source: United Nations COMTRADE database.

From a Nigerian perspective, official data suggest that the economic importance of trade with sub-Saharan Africa is small but increasing—its imports from sub-Saharan Africa accounted for around 5 percent of its total imports during 2008–10, a fraction that has doubled over the past decade. Its exports, mainly oil, to sub-Saharan Africa (Figure 2.7) accounted for about 13 percent of total exports during 2008–10, up from 9 percent a decade earlier. Nigeria has traditionally enjoyed a positive, albeit small, merchandise trade balance with the rest of sub-Saharan Africa.

Figure 2.7.Relevance of Imports from Nigeria

Sources: Nigerian authorities and author’s calculations.

Note: Data for ECOWAS imports from Nigeria excludes Nigeria.

These data do not tell the complete story. For the most part, the trade data represent formal sector activity only.13 However, based on other sources, there is reason to believe informal sector links between Nigeria and its neighbors are stronger and offer a channel through which shocks in Nigeria are transmitted to other countries.

Informal Trade Linkages

Various studies have documented strong informal sector trading links between Nigeria and its neighbors. In particular, studies have highlighted largely unrecorded trade in agricultural goods, especially cereals and grains, and petroleum products; and re-export (entrepôt) trade in a wide variety of goods.

  • Agricultural goods. Nigeria is at the center of the cereals trade in West Africa, covering Niger, Chad, Côte d’Ivoire, Cameroon, and Benin (Balami and others, 2011). Nigeria’s cereal production is an important component of food security in the region. It meets 70 percent of the cereals needs of Niger and Chad. Well-organized trading networks facilitate the transfer of cereals from surplus zones to deficit zones, thereby facilitating the food security of the region and affecting prices at the regional level. The direction of trade shifts on a seasonal basis. Only a small proportion of this trade is recorded in merchandise trade data.
  • Petroleum products. A highly subsidized fuel (gasoline, kerosene) and electricity regime is in place in Nigeria. Across West Africa, but most notably in bordering states (Benin, Cameroon, Chad, Niger), the smuggling of petroleum products from Nigeria is a major source of fuel imports (Golub, 2008). In Benin, for example, more than 80 percent of domestic consumption is estimated to come from smuggled Nigerian gasoline. In 2011, the regulated gasoline price in Nigeria was, on average, about 50 percent below the levels prevailing in neighboring countries, which contributed to the substantial informal trade in fuel products. Box 2.4 describes in more detail the energy market linkages between Nigeria and its neighbors, including the supply of electricity to Niger at highly subsidized prices.
  • Re-export trade. Benin and Niger operate as “entrepôt states” for products that are highly taxed in Nigeria or subject to import bans, including particularly second-hand cars, textiles and garments, rice, and cigarettes. These countries are platforms for onward transport to Nigeria (Golub, 2012). It is estimated that the value of this re-export trade is equivalent to more than 50 percent of Benin’s GDP. Although this trade is partially captured in Benin’s official data, it is not recorded in Nigeria’s statistics (Box 2.5).

Box 2.4.Energy Linkages between Nigeria and Surrounding Countries

Fuel and power prices in Nigeria are substantially below the levels prevailing in surrounding countries. Whereas Nigeria heavily subsidizes many fuel products, most of the neighboring countries impose taxes on these products while adopting more market-sensitive pricing regimes. In addition, transportation costs and other markups are higher for the landlocked countries. These factors contribute to the large price differential between Nigeria and other countries. For example, as shown in Figure 1, gasoline prices in countries around Nigeria are about double the regulated prices observed in Nigeria. These large differentials have led to a variety of rent-seeking behaviors and strong informal trade between Nigeria and neighboring countries.

Figure 1.Comparison of Gasoline prices

Source: Various GTZ reports.

The price differentials and weak management of the subsidy regime in Nigeria have generated strong spillovers for surrounding countries. While it is hard to quantify, informal trade in fuel products has considerable implications for fuel imports and tax revenue in Nigeria’s neighboring countries. For example, it is estimated that official gasoline sales accounted for only 10 to 15 percent of total sales in Benin at end-2011. In addition, in Togo, official gasoline sales declined steadily between 2008 and 2011 (Figure 2). Then, in January 2012, there was a sharp spike following the fuel price increase in Nigeria, which was only partially reversed after the Nigerian authorities reduced the price increase.

Figure 2.Togo: Official Sales of Premium Gasoline

Source: Togolese authorities.

Strong electricity sector linkages exist or are developing between Nigeria and some of the surrounding countries. Under an agreement dating back to 1977, Nigeria provides close to 90 percent of Niger’s power consumption at a highly subsidized price of US$0.03 per kWh, whereas Niger’s own generation cost is on average around US$0.16–US$0.20 per kWh (World Bank, June 2011). In addition, Nigeria will be supplying gas at a reduced price to electricity generating units in Benin and Togo.

Given these strong informal and formal fuel and electricity linkages, changes in Nigeria’s pricing policies for these products would have significant spillovers. IMF staff estimated that the increase in Nigerian fuel prices in January 2012 had a substantial impact on inflation in Benin (Box 2.5). Similarly, World Bank staff estimated that an electricity price increase to US$0.08 (Nigeria’s long-run marginal cost of power generation) would increase Niger’s annual electricity costs by US$26 million.

This box was prepared by Mumtaz Hussain.

Financial Sector Linkages

Another important link between Nigeria and other sub-Saharan African countries is the expanding cross-border activity of Nigerian-based banks.14 The cross-border expansion of Nigerian banks largely took off after the consolidation and strengthening of the Nigerian banking system in 2004.15 The number of foreign subsidiaries increased from a total of just three in 2002, to 44 now operating in 21 sub-Saharan African countries outside Nigeria. The Nigerian banks with the largest number of subsidiaries are United Bank for Africa with 16 sub-Saharan African subsidiaries; Access Bank with 8; and Guaranty Bank with 5. This increases the potential for financial system shocks in Nigeria to be transmitted across sub-Saharan Africa. Including the Togo-incorporated group, Ecobank, the number of bank subsidiaries in sub-Saharan African countries that are strongly linked to Nigeria rises to 75 across 32 countries (Figure 2.8). According to Bankscope data (available for only about half of the subsidiaries), in at least 15 of these countries there is at least one Nigerian subsidiary that accounts for more than 10 percent of systemwide deposits or credit.

Figure 2.8.Number of Subsidiaries of Nigerian-based Banks

Sources: Central Bank of Nigeria; and BankScope.

Nigerian-based banks are present mainly in anglophone West African countries. There are six subsidiaries in The Gambia, six in Sierra Leone, and five in Ghana. Moreover, the assets of Nigerian banks (excluding Ecobank) constitute around 40 percent of the banking system assets in the Gambia and Sierra Leone, about one-fourth in Liberia, and approximately one-sixth in Ghana (see Nnaji, forthcoming). Ecobank has subsidiaries in these four countries, and also has a significant presence in francophone West Africa. In fact, its subsidiaries account for more than 10 percent of system-wide credit or deposits in 11 francophone countries.

The 2009–10 Nigerian banking crisis was a test of the importance of this transmission channel for financial system shocks. The crisis occurred following two years of a credit boom, amidst poor risk management, connected lending, and weak supervision. As a result, the Central Bank of Nigeria (CBN) intervened in 10 out of the 24 banks in 2009, following audits that revealed that most of the intervened banks were insolvent.16 Four of the intervened and insolvent banks had significant cross-border expansion. The CBN audits also revealed that a few non-intervened banks with cross-border subsidiaries experienced sharp increases in non-performing loans (NPLs) or had weak capital. There were thus initial concerns in some sub-Saharan African countries that the Nigerian banking crisis could spread beyond Nigeria.

Box 2.5.Informal Trade between Benin and Nigeria

Trade with Nigeria has always played a major role in Benin’s economy. IMF Direction of Trade data show that Nigeria accounts for only about 5 percent of Benin’s recorded exports. Different trade policies, however, have given impetus to a large informal trade, only part of which is captured in official statistics. It is still lucrative to smuggle goods to Nigeria, even after paying import duties in Benin (World Bank, 2009). This box presents (i) the factors explaining the intensive informal trade between Benin and Nigeria, (ii) its composition and an estimate of the volume, and (iii) the spillover implications.

Factors explaining informal trade. Informal trade has grown mainly on the backbone of different trade policies. Nigeria is a highly protected economy with around 30 products currently subject to import bans and import tariffs hovering around 50 percent on average for consumer goods, in particular for rice and sugar. Unlike Nigeria, Benin’s import tariffs are low—maximum of 20 percent. Nigeria’s trade restrictions provide a large incentive for informal trading. Other policy decisions, especially on fuel subsidies, encourage large scale smuggling of petroleum products from Nigeria to Benin. Finally, cultural affinities between the people of Benin and Nigeria, especially the presence of the Yoruba ethnic group across the two countries, have contributed to informal trading.

Composition and volume. According to Geourjon, Chambas, and Laporte (2008), about half of imports going through the Port of Cotonou to Benin’s domestic market have Nigeria as their final destination. The top imports that transit through the Port of Cotonou include frozen poultry, used cars, and textiles, most of which are on Nigeria’s list of prohibited imports. Golub (2012) estimates Benin’s informal trade, mostly with Nigeria, at around 52 percent of GDP (average between 2006 and 2008), even after excluding cotton and petroleum products, which are not accounted for by Benin customs in those categories. Table 1 provides a comparison of the imports of key products heavily informally traded between Benin and Nigeria with similar products in Togo.1 Because Togo’s and Benin’s domestic demands are similar—given the similarities of their economies and population—large discrepancies in imports for domestic use generally reflect differences in informal trade. On gasoline trading, more than 80 percent of the gasoline consumed in Benin is smuggled from Nigeria, curtailing the provision of gasoline in the formal market and the number of gas stations.

Table 1.Selected Key Imports in Togo and Benin, 2008(Billions of CFA francs)
Imports for domestic useTransit and re-exportTotalImports for domestic useTransit and re-exportTotal
Goods facing import bans in Nigeria
Cotton Cloth17.9350.5368.42.330.432.7
Frozen chicken58.
Used clothes22.433.
Palm oil35.
Vegetable oil0.311.711.
Goods facing high tariffs in Nigeria
Goods facing lower tariffs in Nigeria
Frozen fish4.
Sources: Golub, 2012 (Benin and Togo Customs and author’s calculations).
Sources: Golub, 2012 (Benin and Togo Customs and author’s calculations).

Spillover implications. Informal trade has important spillovers from Nigeria to Benin. The transit and re-export activities are among Benin’s most important industries. They are a significant source of employment, and tax revenue on goods going to Nigeria is estimated at about 14 percent of total tax revenue (2.4 percent of GDP). At the same time, government loses revenue because of illegally imported petroleum products from Nigeria, estimated at between 0.1 and 0.3 percent of GDP. Recently, in January 2012, Nigeria’s decision to cut fuel subsidies by half was immediately and fully passed through to the informal gasoline market in Benin, increasing average inflation from 1.8 percent in the last quarter of 2011 to 6.5 percent for the first quarter of 2012, and dampening domestic demand as households adjusted to higher fuel prices.

This box was prepared by Cheikh Gueye, Carla Macario, and Kevin Wiseman.1 Even goods legally brought into Benin for re-export are not recorded when they enter Nigeria.

The financial turmoil in Nigeria did not lead to systemic contagion across the region. This partly reflects that the two banks with the largest cross-border operations, accounting for more than half of the number of foreign subsidiaries, were not among the troubled banks. Another factor that minimized systemic contagion was that the banking systems of the host countries are mainly funded by local deposits and therefore do not significantly depend on Nigerian funding.17 The foreign subsidiaries of the intervened Nigerian banks were eventually acquired by the banks that acquired the parent banks as part of the crisis resolution process; the problems of the troubled banks had originated at home rather than abroad.

However, a number of cross-border subsidiaries of Nigerian banks did experience turbulence during those years.18 According to Bankscope data, there were instances in which Nigerian subsidiaries in Benin, Burkina Faso, Cameroon, the Democratic Republic of the Congo, Guinea-Bissau, Niger, Rwanda, and Uganda experienced significant declines in either deposits, credit, or assets. Importantly, except in the case of Benin, none of these host countries experienced a systemic decline in these indicators during those years, suggesting that some of the problems of Nigerian subsidiaries may have reflected contagion from Nigeria.19 Similarly, in Ghana, Kenya, and Liberia, NPLs in two Nigerian subsidiaries rose to levels substantially above the average in the banking system of their host countries.

Other Linkages

Remittances from migrant workers

Recorded data on worker remittances from Nigeria to neighboring countries are small relative to GDP, but many remittances go unrecorded. World Bank estimates of migration from neighboring countries to Nigeria suggest that the share of such migrants is modest, viewed from the standpoint of both Nigeria and the source countries (Figure 2.9). The World Bank estimates that there were about 1 million of sub-Saharan Africa migrants in Nigeria in 2010. The main home countries of migrants were Benin (0.2 million, 3 percent of population); Ghana (0.2 million, 1 percent); Togo (0.1 million, 2 percent) and Mali (0.1 million, 1 percent). Recorded remittances from Nigeria are also modest, with only Benin and Togo showing significant receipts (Table 2.4).

Figure 2.9.Share of Migrants to Nigeria to Total Population

Table 2.4.Main Recipients of Remittances from Nigeria
Remittances from Nigeria
Millions of

U.S. dollars
Percent of recipient

Sources: Ratha and Shaw (2007) updated with additional data for 71 destination countries as described in the Migration and Remittances Factbook 2011 (World Bank).
Sources: Ratha and Shaw (2007) updated with additional data for 71 destination countries as described in the Migration and Remittances Factbook 2011 (World Bank).

Foreign Direct Investment

Nigeria’s investment in the sub-Saharan African region, outside the financial sector, is of little significance. Concrete data on Nigeria’s investment in sub-Saharan Africa is difficult to come by. However, even if a substantial portion of Nigeria’s investment was in sub-Saharan Africa, it would still be negligible in terms of aggregate foreign direct investment in the region: the stock of Nigeria’s global outward FDI stood at US$5.9 billion in 2011, compared with an estimated stock of FDI in sub-Saharan Africa of US$290 billion. Anecdotal evidence does not point to a noteworthy Nigerian non-bank corporate presence across sub-Saharan Africa, or even in the western Africa sub-region.

Co-Movements of Macroeconomic Variables

Correlations of GDP between Nigerian and neighboring countries are generally weak. Using annual data from 1989 to 2011, correlations were calculated between (i) the (de-trended) level of real GDP in Nigeria and its neighboring countries,20 and (ii) the corresponding growth rates of real GDP. Only for Chad did the data consistently point to a strong positive correlation, which probably reflects the important share of oil in Chad’s GDP (Table 2.5).

Table 2.5.Nigeria and Neighboring Countries’ GDP
Business cycle1 co-movement of Nigeria

with its neighbors, 1989-2011
Growth rates co-

movement of Nigeria

with its neighbors,

Burkina Faso-0.0950.00.478Burkina Faso0.079
Côte d’Ivoire-0.153-0.2-0.061Côte d’Ivoire-0.370
Source: Authors’ calculations.

Cyclical components of GDP obtained using the Hodrick-Prescott (Nigeria is at t).

Source: Authors’ calculations.

Cyclical components of GDP obtained using the Hodrick-Prescott (Nigeria is at t).

On inflation dynamics, correlations between Nigeria and its neighboring countries are stronger. Contemporaneous correlations are high with Niger, Chad, and Burkina-Faso, while the results suggest that lagged inflation in Ghana and Togo co-move with that of Nigerian inflation. Correlation between food prices is even stronger, pointing to the key role of food prices in inflation co-movements (Table 2.6).

Table 2.6.Nigeria and Neighboring Countries’ CPI
Inflation co-movement of Nigeria with its neighbors,

Food inflation co-movement of Nigeria with its

neighbors, 200X1:Q1–2011:Q4
Burkina Faso0.3140.5-0.042Burkina Faso0.2540.30.212
Côte d’Ivoire0.3370.1-0.397Côte d’Ivoire0.4520.5-0.365
Source: Authors’ calculations.

According to data availability (around 2000).

Source: Authors’ calculations.

According to data availability (around 2000).

Although correlations can be the basis for a preliminary presumption on the size of spillovers, they may be misleading.21 Instead, there is a need to dig further for more insight into the potential for “true” spillovers from Nigeria to its neighboring countries. For this, vector autoregression (VAR) analysis is used.

Shocks to gross domestic product

To estimate output spillovers arising from Nigeria to its neighboring countries, country-specific structural vector autoregression (SVAR) models are estimated for each of the eight countries (Benin, Burkina Faso, Cameroon, Chad, Côte d’Ivoire, Ghana, Niger, and Togo). The variables in the SVAR include (i) domestic GDP; (ii) Nigeria’s GDP; (iii) global demand, proxied by a weighted average of real GDP of the Group of Seven countries and China, with weights proportional to their purchasing power parity adjusted GDP; and (iv) commodity prices, measured by a broad price index (a variant of the model replaces the broad commodity price index with just oil prices, measured by the World Economic Outlook oil price).

The SVAR analysis essentially reinforces the results from the correlation analysis. After controlling for common global factors, shocks to Nigerian GDP have a negligible impact on GDP in neighboring countries.

Shocks to inflation

The SVARs estimated to analyze the effect on neighboring countries’ inflation of shocks to Nigerian inflation include the following variables: (i) domestic inflation, measured by the CPI; (ii) Nigerian inflation, measured by its CPI; and (iii) global inflation, measured by the all-items consumer price index.

The results show that inflation in neighboring countries is quite responsive to inflation dynamics in Nigeria. The impulse-response functions (IRFs) show that a one standard deviation (3 percent) shock to inflation in Nigeria would raise inflation by at least 1 percent in all the neighboring countries, with the exception of Cameroon (Appendix Figures 2.1 and 2.2). Shocks are also mostly transmitted fairly quickly, with the impact taking place within the same year. The spillovers are strongest in Niger (2 percent) and Chad (2 percent), which is consistent with their close informal sector trading relationships with Nigeria that were described in the previous section. Controlling for global food inflation instead of global inflation would not change the results significantly.

The integrated grain market within the region seems to be one of the main transmission channels of Nigeria’s inflation spillovers to neighboring countries. Another set of SVARs was estimated for food inflation. Impulse response functions in Figure A2.2 show that the food prices of Togo, Niger, Côte d’Ivoire, and Chad—which are at the heart of the regional grain market—react strongly to one standard deviation shock of food inflation in Nigeria. Inflation in Benin reacts with a one period lag.

Spillover Effects from Nigeria: Are They Significant?

To summarize:

  • Although Nigeria is an important trading partner for some of its neighbors, traditional indicators (e.g., trade, foreign direct investment, remittances) suggest that spillovers from Nigeria to its neighbors are small, and Nigeria’s role as an engine or catalyst of regional economic activity is relatively unimportant.
  • Other indicators—informal and institutional—point to a somewhat stronger influence of Nigeria in sub-Saharan Africa. During the past decade, Nigeria’s banking sector has expanded throughout the region, and Nigerian-owned banks now operate in 21 sub-Saharan African countries. Informal trade flows with border states and in regional markets are important, particularly in cereals and grains, re-exports from Benin and Niger unrecorded in Nigeria, and illegal petroleum product exports from Nigeria to neighboring countries.
  • The quantitative analysis in this section suggests that, although spillovers from Nigerian GDP to neighboring countries are negligible, inflation shocks in Nigeria have a strong impact on inflation in other countries.

Concluding Remarks

We have examined how developments in South Africa and Nigeria, the two largest economies in sub-Saharan Africa, affect the rest of the region. The main conclusions are that: (i) South Africa’s linkages with the rest of the continent have been growing steadily and look set to deepen over time, but, for now, these linkages are of macroeconomic significance mainly for southern Africa; and (ii) Nigeria still has weak links with the rest of the region, except with its near-neighbors, where its tax and subsidy policies have magnified normal incentives for trade. Macroeconomic shocks in these two economies have spillover effects on several countries—but the spillover effects can be complex (as in the case of SACU), with a longer reach in the case of South Africa, much less so in the case of Nigeria.

Intra-regional trade and financing links within sub-Saharan Africa have been expanding significantly in recent years—but it is widely recognized that there is a long road to travel in terms of achieving close economic integration at the regional and sub-regional level. As this integration proceeds and economic linkages deepen, the importance of spillover effects from large countries to the rest of sub-Saharan Africa, and within their own sub-region, will grow: closer economic linkages inevitably imply increased exposure to shocks, both favorable and unfavorable, in partner countries.

Appendix: Econometric Results

To assess the relative importance of global developments, domestic policies, and South African developments and policies in determining per capita GDP growth in South Africa’s key partners in sub-Saharan Africa, this section presents four pooled regressions (Appendix Table 2.1). These are estimated over two subperiods (1980–2010 and 1990–2010) using annual data drawn in the main from the IMF’s WEO database. The pooled regressions include fixed effects to accommodate country-specific growth differences, including different starting income levels.

Appendix Table 2.1.Pooled Regressions: Per Capita GDP Growth in South Africa’s Main Sub-Saharan African Partners1
Regression 1Regression 2Regression 3Regression 4
Global developments
Real commodity prices2.30 ***2.74***2.09***2.65 ***
U.S. TED spread2-0.79 *-1.06-0.63-1.02-2.19 **
Lagged panel real GDP per capita growth0.23 ***0.18 ***0.28 **0.34 **-0.11 *
World real per capita GDP growth0.53 ***0.47*** **0.360.43 *0.36 **
Regional developments
Public Revenue to GDP0.08 ***0.10 ***0.07 **0.08 **
Real government consumption growth0.04 ***0.04 ***0.04 ***0.03 ***
Inflation-0.05 ***-0.05 ***-0.05 ***-0.02 ***
Real exchange rate-0.04 **-0.03-0.04 **
Lagged panel real GDP per capita growth-0.06-0.10
South African developments
South Africa real per capita GDP growth0.28 **0.34 **
South Africa real exchange rate
South Africa short-term interest rates0.18 **0.18 *
South Africa inflation
South Africa fiscal revenue to GDP0.50 ***0.89 ***
South African real government consumption growth
Regression statistics
Adjusted R-squared0.
Sources: IMF, World Economic Outlook; Haver Analytics Inc.; and IMF staff estimates.

Based on the analysis of real and financial linkages, South Africa’s main trading partners are Angola, Botswana, Congo (DR), Ghana, Lesotho, Malawi, Mauritius, Mozambique, Namibia, Nigeria, Seychelles, Swaziland, Tanzania, Uganda, Zambia, and Zimbabwe. The unbalanced pooled regressions include fixed effects. The asterisks represent significance at the 10 (*), 5 (**), and 1 (***) percent levels.

The TED spread, an indicator of financial risk, is the difference between interest rates on 3-month LIBOR and U.S. treasury bills.

Sources: IMF, World Economic Outlook; Haver Analytics Inc.; and IMF staff estimates.

Based on the analysis of real and financial linkages, South Africa’s main trading partners are Angola, Botswana, Congo (DR), Ghana, Lesotho, Malawi, Mauritius, Mozambique, Namibia, Nigeria, Seychelles, Swaziland, Tanzania, Uganda, Zambia, and Zimbabwe. The unbalanced pooled regressions include fixed effects. The asterisks represent significance at the 10 (*), 5 (**), and 1 (***) percent levels.

The TED spread, an indicator of financial risk, is the difference between interest rates on 3-month LIBOR and U.S. treasury bills.

Key Implications

  • Global developments are, unsurprisingly, an important determinant of growth. In particular, as illustrated in the first regression, per-capita growth increases with real commodity prices, falls with increases in U.S. banking spreads (a proxy for tighter global financial conditions), and increases with the pace of world growth.
  • Domestic policies matter. Fluctuations in per-capita growth are affected by available fiscal resources (the ratio of fiscal revenue to GDP), the pace of growth of government consumption, and the pace of inflation (which, in turn, reflects both the exchange rate arrangement and policy reactions).
  • There is no compelling evidence that aggregate growth in South Africa’s main partners in sub-Saharan Africa is affected by South African developments or policies. When all possible candidate variables (external and domestic policy, including South African policies) are initially included and the variables that contribute the least are sequentially removed until only significant variables remain, the pace of world growth is a significant explanatory factor but the various South African explanatory variables drop out. More mixed results are obtained when one excludes either (i) the role of domestic policy variables, or (ii) the role of global developments.

Appendix Figure 2.1.CPI Inflation Response to a Shock to Nigeria’s Overall inflation

(One standard deviation)

Source: IMF staff estimates.

Note: Red line indicates response to shock. Green lines indicate plus and minus two standard deviations.

Appendix Figure 2.2.Food Prices’ Inflation Response to a Shock to Nigeria Food Price Inflation

(One standard deviation)

Source: IMF staff estimates.

Note: Red line indicates response to shock. Green lines indicate plus and minus two standard deviations.

This chapter was prepared by Trevor Alleyne, Jorge Iván Canales-Kriljenko, Cheikh Gueye, Calvin McDonald, Gonzalo Salinas, and Jon Shields.


For a thorough discussion, see IMF, 2012a.


This analysis only includes actual Liberian exports, not all cargo carried by Liberian-registered ships.


Thus, shifts in South African demand for oil would not have a noticeable effect on Angola or Nigeria, notwithstanding the significant amounts of oil they export to South Africa.


The main exceptions are Angola and Nigeria, which have positive trade balances with South Africa because of sizable oil exports.


Only in the cases of Botswana and Namibia do South African exports to an individual sub-Saharan African country exceed 1 percent of South Africa’s GDP.


Including a reported 860,000 migrants from Zimbabwe and perhaps 1 million from elsewhere in the SADC region in 2010 (using the methodology of Ratha and Shaw, 2007).


World Bank estimates that factor in census information, using the methodology described in Ratha and Shaw (2007) suggest that the number of migrants working in South Africa in 2010 from Lesotho amounted to about 30 percent of its labor force.


The SACU members are currently reviewing their revenue sharing formula.


Asonuma and others (forthcoming) find that the use of shocks to South African inflation improves the forecasting of inflation for the BLNS countries, especially in regard to food prices.


The Botswana pula is linked to a currency basket in which the rand is given a very large weight: consequently, there are also significant, if less marked, monetary spillovers from South Africa to Botswana.


Sizable exports to South Africa from Angola and Nigeria are in fact oil products—standardized commodities, where the influence of demand in the market destination on prices is minimal.


Benin and Niger are notable exceptions. Their official trade data include estimates for informal sector trade.


Ecobank, which is incorporated in Togo and operates in 30 sub-Saharan African countries, is treated here as a Nigerian-based bank; Ecobank-Nigeria accounts for more than 40 percent of Ecobank’s total earning sources (Lukonga and Chung, 2010).


The consolidation of the Nigerian banking system was prompted by an increase in the minimum capital requirement from N2 billion to N25 billion. As a result, the number of banks declined from 89 in 2003 to 24 in 2009, and 20 at the end of 2011.


The assets of the intervened banks accounted for 38 percent of total bank assets in Nigeria.


This was the case, for example, in Liberia, where there were reports of poor risk management in the subsidiary of a Nigerian bank that resulted in very high NPLs. The banking system was not affected, partly because Liberian banks are funded mainly by domestic deposits, and their assets are limited to cash and loans to the domestic private sector.


This is just a cursory analysis, because data are available mainly for subsidiaries of non-intervened banks and, moreover, for only 46 out of 75 subsidiaries.


Benin’s banking system experienced a decline in assets and private credit in 2011. However, these declines were not related to the difficulties of the Nigerian subsidiaries.


The de-trended GDP series were computed using the Hodrick-Prescott (HP) filter.


Correlations do not imply causation. Moreover, correlations may pick up the effect of the two variables having similar responses to common shocks, such as global factors that affect systemic and small economies alike.

    Other Resources Citing This Publication