Journal of African Economies

Journal of African Economies Advance Access published online on August 24, 2009
Journal of African Economies, doi:10.1093/jae/ejp017

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Google Scholar Articles by Arbache, J. S. Articles by Page, J. Social Bookmarking          What's this?

© The author 2009. Published by Oxford University Press on behalf of the Centre for the Study of African Economies. All rights reserved. For permissions, please email: journals.permissions@oxfordjournals.org

How Fragile Is Africa's Recent Growth?

Jorge Saba Arbache^* and John Page¹

Office of the Chief Economist, Africa Region, The World Bank, The Brookings Institution, Washington, D.C. 20433, USA

^* Corresponding author: Jorge Saba Arbache. E-mail: jarbache{at}worldbank.org.

JEL classification: O11, O55

Has Africa finally reached the path to sustained growth? We find that much of the improvement in economic performance in Africa after 1995 is attributable to a substantial reduction in the frequency and severity of growth declines in all economies and an increase in growth accelerations in mineral-rich economies. We find, however, that growth accelerations have not been generally accompanied by improvements in variables often correlated with long run growth, such as investment. We also fail to find evidence that substantial policy and governance improvements were associated with the post-1995 accelerations. We conclude that Africa's growth recovery remains fragile.

	1. Introduction

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 
Economic performance in Sub-Saharan Africa (Africa) has markedly improved. Average GDP per capita growth (purchasing power parity (PPP)) increased from –0.07% during 1975–94, to 1.88% in 1995–2005. Per capita income in Africa has been rising in tandem with the rest of the world since the mid-1990s, and Africa's top performers are doing well compared with fast-growing countries in other regions (Arbache et al., 2008).

Has the region finally turned the corner on the path to sustained growth? Africa's growth over the past three decades has not only been low; it has been highly volatile (see for example, Ndulu et al., 2007; Arbache and Page, 2008a; Raddatz, 2008). In light of this volatility we first ask whether growth has accelerated in a statistically meaningful sense since 1995. We find that there is a structural break in the rate of growth—signalling an acceleration—beginning in the mid-1990s.

We then ask whether the higher rate of growth since 1995 is sustainable. To answer this question we use a variant of the methodology developed by Hausmann et al. (2005) for analysing growth accelerations. Our approach differs from theirs in two ways that we believe make it better suited to Africa's volatile, low growth environment. First, it identifies both growth accelerations and decelerations in a cross section of countries. Second, it does not use a common threshold growth rate to identify growth episodes. Instead, it defines acceleration and deceleration relative to each country's long run economic performance.

We find that much of the improvement in economic performance in Africa after 1995 is attributable to a substantial increase in the frequency and country coverage of growth accelerations combined with reductions in the frequency and severity of growth decelerations. We also find that growth accelerations have not been evenly distributed across countries. Resource-rich economies have had a significantly higher frequency of growth accelerations than non-resource-rich economies.

To assess the durability of growth we test whether there have been significant changes in such long run ‘determinants’ of growth as investment, trade openness and macroeconomic stability during growth episodes before and after 1995. We find little evidence to support the view that significant changes in investment, trade or macroeconomic management underpin the increased frequency of growth accelerations. We also find little evidence that better policies and institutions have spurred growth accelerations, although they appear to be associated with fewer growth declines in resource poor economies. This raises important questions with respect to the durability of Africa's growth performance.

The remainder of this paper is organised as follows. The next section discusses Africa's growth experience and examines whether there were structural breaks in the GDP growth series in the 1990s. Section 3 describes the methodology. Section 4 presents the main results. Section 5 evaluates the sustainability of current growth and Section 6 concludes the paper.

	2. Africa's growth during 1975–2005

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 
In our growth analytics we use the most recent time-series of PPP income data for forty-five African countries from 1975 to 2005.² This period follows the first oil shock and includes the commodity prices plunge, the introduction of structural reforms and the recent growth recovery. Table 1 shows GDP per capita growth rates for 1975–2005 for all of Africa and for various subsets of countries.³ At the aggregate level, growth increased substantially in 1995–2005 to 1.88% from –0.23% in the previous decade and –0.07% in 1975–94. This shift implies an increase of 2% in growth, which is about three-times the long-term growth rate of 0.70%.

View this table: [in this window] [in a new window]   Table 1. GDP per capita growth–aggregate and by country subsets

 
Figure 1 presents the trend and actual paths of growth. Trend growth declined until the mid-1980s, accelerated until 2002–2003 and then had a modest slowdown. Actual growth has been above the trend more frequently since 1995.

View larger version (6K): [in this window] [in a new window] [Download PowerPoint slide]   Figure 1. GDP per capita growth.

 
The significant changes in the rate and volatility of per capita growth after 1995 suggest that some type of a break in the time series may have taken place at about that time. Figure 2 shows the results of recursive residual estimations on the GDP growth series. A statistically significant structural break occurs during the period 1995–97: both the Chow breakpoint test and the Chow forecast test reject the hypothesis of no structural break in the growth series during the mid-1990s.

View larger version (7K): [in this window] [in a new window] [Download PowerPoint slide]   Figure 2. Recursive residual test of stability of the growth series.

 
The region's post-1995 growth has been far from uniform across countries.⁴ Since 1995 geology has trumped geography in Africa (Table 1). Resource-rich economies as a group grew significantly faster than all other African economies. Oil exporters especially have prospered: they grew at 4.5% per annum since 1995, more than twice the region's mean growth rate. Other resource-rich economies in contrast grew at about the same rate as non-resource-rich economies, 1.4% per year in 1995–2005.

Resource-poor landlocked and coastal economies grew at approximately the same rate. Landlocked countries without natural resources grew by 1.2% per year in 1995–2005, a substantial improvement over their growth rate in 1975–94 of 0.10%. Non-resource-rich coastal countries grew at 1.3% after 1995 in contrast to 0.2% during 1975–95.

Conflict retarded growth both before and after 1995. Countries in major conflicts experienced a –1.4% per year growth decline in 1995–2005, similar to their rate of decline between 1975 and 1994 of –1.1%. Countries with minor conflicts, however, managed to grow at about 1% per year and improved substantially in relation to their very rapid rate of decline in 1975–94 (–2.2%).

	3. Understanding the turnaround

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 
This section examines factors underlying the shift in Africa's growth since 1995. We begin by employing a variant of the methodology first developed by Hausmann et al. (2005) to identify growth accelerations and decelerations (Arbache and Page, 2007). To identify a growth acceleration we require that the following three conditions are met in each of at least three consecutive years:

Condition 1: The forward four-year moving average growth rate minus the backward four-year moving average growth rate exceeds 0.⁵

Condition 2: The forward four-year moving average growth rate exceeds the country's average growth rate, meaning that the pace of growth during acceleration is higher than the country's trend.

Condition 3: The forward four-year moving average GDP per capita exceeds the backward four-year moving average.

A sign change from (+) to (–) in Condition 1 suggests a growth trend shift. A deceleration episode occurs when in three consecutive years: the forward three-year moving average growth rate minus the backward three-year moving average growth rate is less than zero (Condition 1); the forward three-year moving average growth rate is below the country's average growth rate (Condition 2); and the forward three-year moving average GDP per capita is below the backward three-year moving average (Condition 3). A growth acceleration or deceleration episode is defined to include the three years following the last year that satisfies Conditions 1–3.⁶

Condition 2 makes our definition of a growth acceleration or deceleration endogenous to each country's long run rate of growth. There is clearly a risk that by identifying a period of modest, sustained growth in a low growth economy as a growth acceleration we will assign too much significance to a minor change in economic performance. However, it is also true that a period of relatively modest per capita growth may signal an important economic change in a country with very low growth rates, and a decline in per capita income of equal magnitude could spell a serious economic collapse in a stagnant economy.⁷ Condition 3 ensures that the growth acceleration episode is not a recovery from a recession.

Our definition of a growth acceleration is less restrictive than that of Hausmann et al. (2005). The main differences between their methodology and ours are that: they employ both a longer moving average window—seven versus four years—and a higher threshold—two versus zero—percent growth rate. They further require that average growth must be at least 3.5% during the acceleration episode. We define acceleration (and deceleration) relative to each country's trend growth rate.⁸

Table 2 shows the relative frequency of accelerations and decelerations, and their respective growth rates, for different periods. Between 1975 and 2005, there was a slightly higher probability that the representative African economy was in a growth acceleration than a deceleration: 25% of the 1,243 total country-year observations (per country per year) identify growth accelerations, while 22% identify growth decelerations.⁹ The remaining 53% of country-year observations reflect normal economic times in which countries were growing at about their trend growth rate. Between 1975 and 2005, countries in Africa that experienced growth accelerations managed to grow on average by 3.6% per year during the acceleration. During decelerations, countries contracted on average by 2.7%.

View this table: [in this window] [in a new window]   Table 2. Frequency of growth acceleration and deceleration, growth rates and GDP per capita

 
In 1975–94 growth decelerations were twice as frequent as accelerations: 29 versus 14% of all country-year observations. In contrast between 1995 and 2005, accelerations were 3.5 times more likely than decelerations: 42% of the 494 country-year observations for 1995–2005 were growth accelerations, and 12% of the country-year observations were growth decelerations. In 1995–2005, the average growth rate for countries during acceleration episodes was 3.8% The average growth rate for countries experiencing growth decelerations in 1995–2005 (–1.3%) was less than half that in previous decades (–3.1%).

Table 3 shows the frequency of growth acceleration and deceleration episodes by country type for the period 1995–2005. There is no statistical difference in the probabilities of growth acceleration and deceleration episodes for different geographical locations. Landlocked countries fared about the same as their coastal neighbours: t-tests between the mean frequency for each group's acceleration and deceleration episodes and the Africa-wide mean reveal no statistically significant differences.

View this table: [in this window] [in a new window]   Table 3. Frequency of growth acceleration and deceleration by country subset – 1995–2005

 
Resource-rich economies as a group had a significantly higher frequency of growth accelerations (0.51, significant at the 0.05 level) than the regional average. They also had a significantly lower frequency of growth declines (0.08, significant at the 0.10 level). Countries without natural resources experienced growth accelerations less frequently (0.38, significant at the 0.05 level) than the regional mean and had a slightly higher frequency of growth declines (0.14, significant at the 0.10 level). Making paired t-tests we find statistically different probabilities of growth accelerations and decelerations between the resource-rich and non-resource-rich groups.

	4. What underpins recent growth?

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 
Not all of the growth turnaround, however, can be attributed to resource-rich economies and fewer growth declines. Africa's non-resource-rich economies have also experienced a major improvement in growth performance; growth accelerations in the resource-poor countries were more than two and a half times more frequent in 1995–2005 than in 1975–94. This section addresses the question of the extent to which this turnaround is associated with changes in three types of ‘growth determinants’—investment and trade, macroeconomic management and policies and institutions—that have been linked to long run growth in the cross-country literature.¹⁰ We test for significant differences in the means of growth related variables for the representative African economy during growth episodes before and after 1995, the year of the structural break in the growth series.¹¹

Clearly, such an approach does not allow us to identify causal relationships between changes in growth determinants and growth episodes. From the point of view of assessing sustainability, this is not necessarily a limitation: we do not need to know whether, say, an increase in the investment rate led to more frequent growth accelerations or more frequent accelerations of growth prompted a rise in investment. Regardless of the direction of causality, if there is evidence that the structural characteristics of Africa's economies strengthened during the period of more rapid or frequent growth, it increases our expectation that the post-1995 growth acceleration may be sustained.

4.1 Investment and openness
Table 4 compares sample means of a set of variables reflecting investment and trade openness in 1995–2005 with sample means in 1975–94. In panel 1 the comparison covers all growth episodes—growth accelerations, growth decelerations and trend growth—in the two periods. Panel 2 focuses on growth accelerations only. Table 4 also presents the comparison of means separately for non-resource-rich and resource-rich countries.¹²

View this table: [in this window] [in a new window]   Table 4. Differences of means of investment and trade variables before and after 1995

 
We do not find much evidence that for all countries and all growth episodes investment has changed very much since 1995. The small increase in the GDP share of aggregate investment for all countries and all episodes between 1975–94 and 1995–2005 was not statistically significant, although private investment increased from 11.5% to 12.5% of GDP (significant at the 0.10 level). The largest change was in foreign direct investment (FDI), which increased to 5% of GDP in 1995–2005 from 1.5% in the previous period.

The increase in investment was primarily in the resource-rich economies. Aggregate investment increased (by about 2% of GDP, significant at the 0.10 level) in resource-rich economies during all growth episodes between 1975–94 and 1995–2005. It remained essentially constant in the non-resource-rich economies. Private investment rose by about 2% of GDP in resource-rich economies and by about 0.6% in the non-resource-rich, but neither increase was significant. FDI increased about five fold in resource-rich economies (to 7.9%, significant at the 0.05 level) to a level that was more than twice the share of GDP in resource poor countries (3.6%).

When we focus on growth accelerations alone, most investment variables declined for the region as a whole in 1995–2005 in comparison with the previous period. Indeed, the only investment variable to increase significantly was FDI as a share of GDP. There was no statistically significant change in the mean values of investment variables in resource-rich economies during growth accelerations before and after 1995. The non-resource-rich economies show a pattern that is similar to that for the region as a whole: total investment and private investment were significantly lower as a share of GDP during growth accelerations in 1995–2005 than during 1974–94. FDI was unchanged.

These results are partly due to compositional changes in the country-year sample of growth accelerations, particularly in the non-resource-rich economies. Because the frequency of growth accelerations was much higher during 1995–2005 than in 1975–94 (0.42 and 0.14, respectively), a wider variety of countries are included in the sample. In 1975–94 the small number of countries that experienced growth accelerations stood out markedly from their peers with respect to their investment rates. After 1995 a wider range of countries experienced accelerated growth, including some with low investment rates.

Much of the empirical literature on growth accelerations suggests that there is a strong association between trade openness and accelerated growth.¹³ During all growth episodes and for all countries, trade as a share of GDP increased by 8% points after 1995 (significant at the 0.05 level). The expansion of trade was almost evenly divided between increases in exports and imports. While trade as a share of GDP increased significantly in both resource-rich and non-resource-rich economies between 1975–94 and 1995–2005, the expansion of exports in resource-rich economies was substantially larger than for resource poor economies (5.5 versus 3.1% of GDP).¹⁴

Looking at growth accelerations alone, we find a pattern similar to that for investment. Trade as a share of GDP was significantly lower for all economies and for non-resource-rich economies during growth accelerations after 1995. Trade as a share of GDP was essentially unchanged between the two periods in resource-rich economies undergoing growth accelerations. The same compositional effect is clearly at work: the sample of non-resource-rich economies undergoing growth accelerations is substantially larger in 1995–2005 and includes some countries that have low trade ratios.

4.2 Macroeconomic management
Macroeconomic management has clearly improved since 1995. The average fiscal deficit as a percentage of GDP in African countries declined from 5.7% during the 1980s and 1990s to 2.9% during 2000–06.¹⁵ The number of countries able to keep inflation below 10% a year increased from 11 to 26 in the early 1990s to about 31–35 since 2000. Our data reflect the improvement in macroeconomic performance, but are statistically not very informative.

Statistically, inflation did not decline for all growth episodes between 1975–94 and 1995–2005 (Table 5). Comparing resource-rich and non-resource-rich economies we observe an increase in inflation among the former and a substantial reduction among the latter, but neither change is statistically significant. These results are strongly influenced by outliers. Had we removed outliers, we would have observed a statistically significant reduction in inflation during 1995–2005.¹⁶

View this table: [in this window] [in a new window]   Table 5. Differences of means of macroeconomic variables before and after 1995

 
Inflation also declined during growth accelerations, but again the change was not statistically significant. Inflation in resource poor economies was lower during growth accelerations in 1995–2005 compared with 1974–94 and higher in resource-rich economies, but again neither change is statistically significant. The presence of outliers strongly influences these results as well.

The real exchange rate depreciated after 1995 both for all countries and all episodes and for the country-year observations during growth accelerations (Table 5). Both resource-rich and non-resource-rich economies experienced improvements in exchange rate competitiveness. Differences in the real exchange rate between the two periods were much less, however, for the sample of countries undergoing acceleration episodes. This reflects the fact that their exchange rates were substantially more competitive than the regional average during 1975–95 while they were marginally more appreciated than the regional average in 1995–2005.

4.3 Policy and institutional reforms
The empirical literature on growth accelerations suggests that policy and institutional reforms can lead to accelerated growth, even in circumstances where underlying economic variables remain largely unchanged (Hausmann et al., 2005; Jones and Olken, 2005; International Monetary Fund, 2005). The World Bank's Country Performance and Institutional Assessment (CPIA) score gives a measure of policy and institutional performance that is available in comparable form since 1997.¹⁷ The average CPIA score for Africa rose from 2.8 in 1997 to 3.2 in 2006. The number of African countries with CPIA scores equal to or greater than the international good performance threshold of 3.5 also rose from five countries in 1997 to 17 in 2006.

The most striking improvement in the policy components of the CPIA is observed in macroeconomic management (Figure 3), a component whose rating is consistently above the overall mean rating. The public sector management and institutions component, which is composed of subcomponents such as transparency, accountability and corruption, quality of public administration, property rights and rule-based government has also improved, but at a slower pace. Its scores remained consistently below the overall rating.

View larger version (14K): [in this window] [in a new window] [Download PowerPoint slide]   Figure 3. CPIA, selected components and overall rating 1997–2006, means (rating scale ranges from 1 [low] to 6 [high]). Source: World Bank.

 
Due to the short length of the CPIA series, we are not able to compare levels of institutional and policy performance before and after 1995. Progress in improving policies and institutions since 1995, however, appears to have been more closely related to the reduction in growth declines than to the increase in growth accelerations. The correlation between the CPIA and acceleration episodes (1997–2005) is not statistically significant. It is highly significant and negative in the case of deceleration episodes. When CPIA scores rise, the likelihood of a deceleration decreases.

Surprisingly, despite the improvement in the CPIA, other measures of Africa's governance, such as the World Bank's Worldwide Governance Indicators, worsened between 1996 and 2005 (Kaufmann et al., 1999). Resource-rich—and oil-rich countries in particular—were among the main causes of the deterioration in average regional measures of governance. These indicators were particularly low during growth accelerations in resource-rich economies and among oil exporters in particular (Table 6).

View this table: [in this window] [in a new window]   Table 6. Governance indicators during growth acceleration—1996–2005 (–2.5 to +2.5)

 

	5. How durable is growth?

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 
What does our analysis suggest about the durability of Africa's growth recovery? Although a number of other studies have found that growth accelerations can occur without corresponding increases in investment rates, the lack of change in aggregate and private investment is worrisome.¹⁸ In the past private investment has been endogenous to growth in Africa.¹⁹ The absence of any increase in the private investment rate during growth accelerations in our data suggests that much of the growth following 1995 consists of movement from inside the average economy's production possibility frontier towards the boundary. While such movement would be registered as an improvement in total factor productivity, and marks a reversal from past low yields on investment, it does raise concerns regarding the sustainability of growth once Africa's economies begin to reach their production frontiers.²⁰ Africa's aggregate and private investment rates lag other regions, and even with an improvement in productivity, the lack of change in investment poses problems from the point of view of the sustainability of further growth.

The fact that trade has expanded overall since 1995—although not for countries undergoing growth accelerations—should be interpreted with some caution. Exports and openness do not appear to be leading growth since 1995. Exports as a share of GDP were lower in non-resource-rich economies undergoing growth accelerations after 1995 than during all episodes, a marked reversal from 1975–94 when higher exports were correlated with growth accelerations. Compared with Latin America, the Middle East and North Africa, and South Asia, Africa has always had a high share of trade in its national income and a high ratio of exports to GDP. However, African non-oil exports are growing relatively slowly, and non-traditional exports are a tiny share of the region's output. Africa's exports remain heavily concentrated in a few traditional commodities. Our data suggest that the expansion of trade after 1995 was primarily driven by rising exports from the resource-rich countries.

Changes in structural variables may have played a greater role in reducing the frequency and severity of growth declines than in increasing the frequency of growth accelerations. We find statistically significant improvements between 1975–94 and 1995–2005 in most indicators related to investment and trade during growth deceleration episodes (Table 7). Aggregate and private investment increased, as did FDI. Trade as a share of GDP increased substantially; exports rose by 6.5% of GDP between the two periods. These results are consistent with our finding that growth contractions were far less severe in 1995–2005 than during 1974–94.

View this table: [in this window] [in a new window]   Table 7. Differences of means of key economic variables during growth decelerations before and after 1995

 
Better macroeconomic management appears to be an important factor explaining the region's better growth performance. While our measure of average inflation is problematic, we find that it is uniformly lower—on average and for both growth acceleration and deceleration episodes—in the period after 1995. In countries experiencing growth declines after 1995, inflation is less than half of its 1975–94 rate, although the difference in the means of the GDP deflator is not statistically significant. We also find that countries experiencing growth accelerations show evidence of lower inflation and countries in growth declines have higher inflation than the regional average in both periods.

The move to more flexible and competitive exchange rates may also have spurred growth. There has been a convergence in the real exchange rate across different country types and growth episodes since 1995. In contrast to 1975–95, there is very little difference in the real exchange rate between the overall regional average and the average for countries during their growth accelerations. The real exchange rate for countries in growth declines also depreciated significantly after 1995 and is similar to the regional average.

The convergence of both inflation and exchange rates across growth episodes suggests that the region may have already received much of the growth pay-off from improved macroeconomic management. Macroeconomic stability, moreover, has been found to play a larger role in reducing the probability of growth declines than in raising the probability of growth accelerations (Arbache and Page, 2007). This suggests that while maintaining improved macroeconomic management should be a priority for avoiding a return to lower growth, it is unlikely alone to be sufficient to sustain growth in the future.

Our evidence on the role of improvements in policies and institutions in spurring growth after 1995 is incomplete. We are unable to compare the CPIA or the World Bank governance indicators after 1995 with earlier periods. Thus, it is possible that a major improvement in policies and governance in the mid-1990s accounts for much of the recent improvement in growth performance. This certainly appears to have been the case for macroeconomic policy.

Some of the evidence we do have on policies and institutions raises concerns for growth sustainability, however. The public sector management and institutions component of the CPIA has shown little change since 2002. Governance indicators have declined since 1996 for the region as a whole and for the resource-rich economies in particular. They are also markedly lower in resource-rich economies compared with non-resource-rich economies during growth accelerations. Given that the resource-rich economies have led the growth boom, these results are worrisome. They suggest the possibility that in mineral-rich economies booms are accompanied by adverse governance outcomes that may eventually reduce future growth.

	6. Conclusions

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 
The improvement in economic performance in Africa after 1995 turns out to be due largely to two factors: a substantial reduction in the frequency and severity of growth declines in all economies and an increase in growth accelerations in mineral-rich economies. The growth recovery has been stronger for the resource-rich economies than for those with poor natural resources. The resource-rich economies had significantly higher frequencies of growth accelerations and lower frequencies of growth decelerations than their neighbours without natural resources.

We found evidence of modest improvements in some structural correlates of growth, such as FDI and trade openness, between 1975–94 and 1995–2005. However, we failed for the most part to find significant increases in the shares of total investment and private investment in GDP. Where investment did increase, it was confined largely to the resource-rich economies. With the exception of macroeconomic policy, we failed to detect a significant relationship between improvements in the overall policy stance or governance of countries and growth accelerations, although better policies and governance did appear to help countries avoid growth declines. In sum, our evidence suggests that Africa's growth surge was propelled by the rapid growth of demand for natural resources and by movement towards the region's existing production possibility frontier.

The role of rising resource revenues in Africa's growth recovery has been widely discussed. What is perhaps less widely recognised is that Africa's growth turnaround has also been largely a matter of learning how to avoid economic declines.²¹ However, fewer economic mistakes and rising resource revenues may not be sufficient to sustain growth. Avoiding economic mistakes is an important achievement, but with the frequency of growth declines now in the range of 10% there are limits to how much more avoiding growth collapses can contribute to long run growth. The concentration of good economic times in the resource-rich economies raises the possibility that growth is vulnerable to declines in commodity prices and to the natural resource ‘curse’.²² Africa's growth recovery is a cause for celebration but not for complacency.

	Appendix A

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 

Table A1. Country categories assignment

Country Oil exporter Coastal Resource rich Angola 1 1 1 Benin 0 1 0 Botswana 0 0 1 Burkina Faso 0 0 0 Burundi 0 0 0 Cameroon 1 1 1 Cape Verde 0 1 0 Central African Republic 0 0 0 Chad 1 0 1 Comoros 0 1 0 Congo, Democratic Republic 0 0 1 Congo, Republic 1 1 1 Cote d'Ivoire 0 1 0 Equatorial Guinea 1 1 1 Eritrea 0 1 0 Ethiopia 0 0 0 Gabon 1 1 1 Gambia, The 0 1 0 Ghana 0 1 0 Guinea 0 1 1 Guinea-Bissau 0 1 0 Kenya 0 1 0 Lesotho 0 0 0 Madagascar 0 1 0 Malawi 0 0 0 Mali 0 0 0 Mauritania 0 1 0 Mauritius 0 1 0 Mozambique 0 1 0 Namibia 0 1 1 Niger 0 0 0 Nigeria 1 1 1 Rwanda 0 0 0 Sao Tome and Principe 0 1 1 Senegal 0 1 0 Seychelles 0 1 0 Sierra Leone 0 1 1 South Africa 0 1 0 Sudan 1 1 1 Swaziland 0 0 0 Tanzania 0 1 0 Togo 0 1 0 Uganda 0 0 0 Zambia 0 0 1 Zimbabwe 0 0 0

	Acknowledgements

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 
This paper benefited from the comments and suggestions of two anonymous referees.

	Notes

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 
¹ The findings and interpretations of this paper are those of the authors. They do not represent the views of the World Bank, its Executive Directors or the countries it represents.

² Data on per capita income (PPP at 2000 international prices) and its growth rate are taken from the World Bank's World Development Indicators (WDI) (various years) unless otherwise specified. Data on conflicts are from the UCDP/PRIO Armed Conflict Database. The WDI's GDP per capita PPP series starts in 1975. The sample includes all Sub-Saharan countries except Liberia and Somalia, for which there are no data. Thus, there is an unbalanced panel of data with T = 31 and N = 45. We use un-weighted country data in reporting regional or sub-regional averages.

³ The country assignment is in Table A1 in the appendix. It follows the groupings proposed by Collier and O'Connell (2006), with the exception of Democratic Republic of Congo, which is here classified as resource-rich. We acknowledge that country classifications are often controversial.

⁴ Arbache and Page (2008b) show that Africa's cross-country income distribution is becoming less equal over time, that inequality reached a peak in 1995–2005, and that Africa can be divided into stable rich and poor income clubs.

⁵ This requires the forward moving average window (t, t + 1, t + 2, t + 3) to be higher than the backward window (t, t – 1, t – 2, t – 3) and above 0.

⁶ As an example, if Conditions 1–3 identify a growth acceleration during, say, 1991–1995, the years 1996–98 are included as part of the episode. The growth acceleration episode comprises a period that starts in 1991 and ends in 1998.

⁷ Condition 2 also helps to limit the number of identified accelerations in countries with sustained, long run growth. If a country is growing rapidly, it will lift the growth trend, reducing the number of estimated accelerations. This is particularly significant for countries experiencing very low or very high growth rates.

⁸ For more details on the methodology and how it compares with Hausmann et al. (2005), see Arbache and Page (2007).

⁹ As a means of checking the robustness of the results, growth accelerations and decelerations were also identified by replacing 0 with +1 and –1% for acceleration and deceleration, respectively, in Condition 1, but the results did not change substantially. Therefore, only the base-case results are reported, because they are less restrictive.

¹⁰ See for example, Hausmann et al. (2005), Ndulu et al. (2007), Johnson et al. (2007) and Berg et al. (2008). This literature identifies a long list of factors associated with sustained growth: political and economic institutions, inequality, fractionalisation, social and physical indicators, export structure, macroeconomic stability, costs of doing business, trade liberalisation, exchange rate overvaluation, education, health, terms of trade, financial liberalisation, among others. The most common are the ones we use in Section 4.

¹¹ We also split the series in before and after 2000, when some commodity prices started to increase. Interestingly, we did not find substantial differences in the means of our economic indicators compared with 1995–2005.

¹² We use the resource-rich category instead of resource-rich, non-oil because it provides a better sense of the influence of natural resources, including oil, on economic performance. Results for resource-rich, non-oil are, however, similar to those for resource-rich.

¹³ Jones and Olken (2005, p. 2) assert for example that ‘growth takeoffs are primarily associated with large and steady expansions in international trade’.

¹⁴ Although the terms of trade merit some analysis, lack or limited availability of data for several countries, notably the mineral rich such as Angola, Equatorial Guinea, Sierra Leone and Sudan, was found to bias the results. Therefore, we did not include it in our analysis.

¹⁵ Low-income countries in Africa are generally constrained from borrowing in international capital markets. As a result, the current account balance is not a good indicator of macroeconomic management. Imports in these countries tend to adjust to export revenues and aid inflows.

¹⁶ Average inflation is a difficult variable to interpret because of the sizeable standard deviations observed in many African countries and the presence of extreme outliers. There are many country-year outliers including Angola in 1996 (5,400%), DRC in 1996 (638%) and Zimbabwe, more recently. Despite these limitations, we found it worthwhile to discuss inflation because of the substantial difference in means observed during accelerations and all episodes.

¹⁷ The CPIA scale ranges from 1 (low) to 6 (best practice) on a range of indicators reflecting economic and structural policies and institutions.

¹⁸ Brazil and the Philippines, for example, have also experienced a similar puzzle in recent years (Bocchi, 2008). Jones and Olken (2005) using growth accounting exercises for 125 countries find that physical capital accumulation plays a negligible role in growth take-offs.

¹⁹ See for example Devajaran et al. (2003) and Hoeffler (2002).

²⁰ See Devarajan et al. (2003) for a discussion of the low returns to investment in Africa.

²¹ This is similar to the conclusion drawn by Ndulu et al. (2007) and Fosu and O'Connell (2006) who also examine the impact of growth declines on economic performance.

²² See Collier (2007) for a discussion of the resource ‘curse’.

	References

TOP 1. Introduction 2. Africa's growth during... 3. Understanding the turnaround 4. What underpins recent... 5. How durable is... 6. Conclusions Appendix A Notes Acknowledgements References

 

Arbache J. S., Page J. More Growth or Fewer Collapses? An Investigation of the Growth Challenges of Sub-Saharan African Countries. (2007) World Bank Policy Working Paper No. 4384.

Arbache J. S., Page J. Patterns of Long Term Growth in Sub-Saharan Africa. In: Africa at a Turning Point? Growth, Aid and External Shocks.—Go D., Page J., eds. (2008a) Washington, DC: The World Bank.

Arbache J. S., Page J. Hunting for Leopards: Long Run Country Economic Dynamics in Africa. (2008b) World Bank Policy Working Paper No. 4715.

Arbache J. S., Go D., Page J. Is Africa's Economy at a Turning Point? In: Africa at a Turning Point? Growth, Aid and External Shocks.—Go D., Page J., eds. (2008) Washington, DC: The World Bank.

Berg A., Ostry J. D., Zettelmeyer J. What Makes Growth Sustained? (2008) IMF Working Paper WP/08/59.

Bocchi A. M. Rising Growth, Declining Investment: The Puzzle of the Philippines. (2008) World Bank Policy Research Working Paper No. 4472.

Collier P. The Bottom Billion (2007) Oxford: Oxford University Press.

Collier P., O'Connell S. Opportunities and Choices, in Explaining African Economic Growth. (2006) Oxford: Oxford University and Center for the Study of African Economies.

Devajaran S., Easterly W., Pack H. Low Investment Is Not the Constraint on African Development. Economic Development and Cultural Change (2003) 51:547–71.[CrossRef][Web of Science]

Fosu A., O'Connell S. Explaining African Growth: The Role of Anti-Growth Syndromes. In: Proceedings: Annual Bank Conference of Development Economics—Bourguignon F., Plescovitch B., eds. (2006) Washington, DC: World Bank.

Hausmann R., Pritchett L., Rodrik D. Growth Accelerations. Journal of Economic Growth (2005) 10:303–29.[CrossRef][Web of Science]

Hoeffler A. The Augmented Solow Model and the African Growth Debate. Oxford Bulletin of Economics and Statistics (2002) 64:135–58.[CrossRef][Web of Science]

International Monetary Fund. Regional Economic Outlook: Sub-Saharan Africa (2005) Washington, DC.

Johnson S., Ostry J. D., Subramanian A. The Prospects for Sustained Growth in Africa: Benchmarking the Constraints. (2007) IMF Working Paper WP/07/52.

Jones B. F., Olken B. A. The Anatomy of Start-Stop Growth. (2005) NBER Working Paper Series No. 11525.

Kaufmann D., Kraay A., Zoido P. Governance Matters. (1999) World Bank Policy Research Working Paper No. 2196.

Ndulu B. J., Chakraborti L., Lijane L., Ramachandran V., Wolgin J. Challenges of African Growth – Opportunities, Constraints, and Strategic Directions (2007) Washington, DC: The World Bank.

Raddatz C. Have External Shocks Become More Important for Output Fluctuations in African Countries? In: Africa at a Turning Point? Growth, Aid and External Shocks—Go D., Page J., eds. (2008) Washington, DC: The World Bank.

CiteULike    Connotea    Del.icio.us    What's this?

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Google Scholar Articles by Arbache, J. S. Articles by Page, J. Social Bookmarking          What's this?

Online ISSN 1464-3723 - Print ISSN 0963-8024

Copyright © 2009 Centre for the Study of African Economies

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics