Policy Research Working Paper 11028 The Changing Landscape of Africa’s Growth Cesar Calderon Andrew Dabalen Ayan Qu Africa Region Office of the Chief Economist January 2025 Policy Research Working Paper 11028 Abstract This paper examines the main features of real growth per suggests that the region underperforms on duration during capita of Sub-Saharan Africa over the past six decades, episodes of expansions/recessions, and the contribution of before uncovering the sources of growth—including those total factor productivity remains small, although it has of growth miracles. Three distinct growth phases before improved over the past two decades. A few countries in the recent “lost decade” are observed. The swinging pat- the region, for example, Botswana, Ethiopia, and Mauritius, tern of income per capita shows that the region has not have sustained growth for decades, on par with the world’s converged with most benchmarks and is not resilient to best performance. These exceptional “growth miracles” are shocks, although resilience has improved since the begin- distinguished by a common set of factors that explain how ning of the twenty-first century. Furthermore, growth such miracles start and are sustained: leadership, economic across countries in the region is heterogeneous and falls diversification, market expansion, and investment for the into three broad groups. Analysis of the sources of growth future. This paper is a product of the Office of the Chief Economist, Africa Region. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at ccalderon@worldbank.org; adabalen@worldbank.org; and aqu@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team The Changing Landscape of Africa’s Growth1 Cesar Calderon, Andrew Dabalen, Ayan Qu Keywords: Sub-Saharan Africa, Growth Accounting, Growth Miracles. JEL Codes: O40, O47, O55 1 This paper is a product of the Office of the Chief Economist, Africa Region. It is part of a larger effort by the World Bank to provide open access to its research and contribute to development policy discussions around the world. The author(s) can be contacted at ccalderon@worldbank.org; adabalen@worldbank.org; aqu@worldbank.org. The views expressed in this paper are those of the authors, and do not necessarily reflect those of the World Bank or its Boards of Directors. 1. Introduction Over the past two decades, there was great optimism about economic performance in the African continent. During the period 2000-14—known as the Africa Rising period—nearly half of the 25 countries in the world with the fastest GDP growth were in Sub-Saharan Africa, and the growth rates of economic activity for all these countries in the region surpassed an average rate of 6.5 percent per year. In half of the years during this period, Africa’s GDP grew at a faster pace than that of East Asia. Economies in the region were flourishing and most countries were at peace. Life expectancy in the region rose from 51 years in 2000 to 59 years in 2014 while the incidence of HIV and malaria dropped sharply by nearly 60 and 33 percent, respectively. Foreign direct investment into the continent increased from $10.4 billion in 2000 to about $54.5 billion in 2014. Technology appeared to be spreading faster in the region. Mobile cellular subscriptions rose 60-fold between 2000 and 2014, with mobile penetration soaring from less than 1 mobile phone per 100 people in 2000 to about 13 per 100 people in 2014. There was hope that livelihoods would continue improving in the years ahead. A surge in economic activity in the region was broad-based: resource rich nations rode the wave of commodity prices and had protracted periods of rapid GDP growth (e.g., Equatorial Guinea, Chad, Mozambique, Nigeria and Zambia), while non-resource rich low-income countries experienced growth spurts supported by sound macroeconomic and structural policies (e.g., Rwanda, Uganda, Tanzania, and Ethiopia). Broadly, the growth record of the region during this period was attributed to both external and domestic factors. On the external front, the commodity super cycle, the emergence of China as trade and investment partner, and the massive inflow of foreign capital helped propel growth in the region. Domestically, improved macroeconomic management supported rising consumption and investment across resource-intensive sectors (extractives) and non- resource sectors (telecommunications, finance, retail, real estate and transportation). Economic performance during the Africa Rising period, however, was less impressive when accounting for population growth and not adequately inclusive. After hitting a trough in the mid-1990s, the growth spurt during the period lifted the region’s per capita income to a level that surpassed its previous peak in economic activity—circa 1974 (Figure 1).2 After the plunge in commodity prices, however, growth per capita returned to negative territory, i.e., -0.33 percent per year in 2015-19.3 Additionally, lower growth elasticity of poverty resulted in a reduction in the poverty headcount ratio that was considerably slower than that in East Asia and South Asia.4 For instance, at $2.15 (in 2017 PPP), the headcount poverty ratio in Sub-Saharan Africa dropped from 56.5 percent in 2000 to 37.6 percent in 2014. In contrast, headcount poverty declined in East Asia and the Pacific (EAP) from 39.5 in 2000 to 3.6 in 2014 while it fell in South Asia (SA) from 39.8 in 2002 to 17.9 in 2014.5 2 It took more than three decades for the region to surmount the previous peak in living standards. In 2008, Sub- Saharan Africa’s GDP per capita (US$ 1,539 at 2015 constant prices) finally exceeded its previous peak (US$ 1,533 at 2015 constant prices) in 1974 (Figure 1). 3 The economic fallout from the COVID-19 pandemic resulted in the first contraction in economic activity—as measured by the gross domestic product (GDP)—in the region over the past 25 years. In per capita terms, the contraction was even larger (-4.5 percent). 4 The growth elasticity of poverty has also been historically lower in Africa than elsewhere (World Bank 2018 b). This is attributed to the inequalities in the access to education, basic services (safe water and sanitation), the lack of structural transformation, and economic diversification (Wu, Atamanov, and Bundervoet 2023). 5 Growth in the region after the plunge in commodity prices was insufficient to reduce poverty significantly. The poverty headcount ratio declined to 34.9 in 2019. See Baah et al. (2023). Figure 1. Real and Relative GDP per capita in Sub-Saharan Africa, 1960-2022 1700 0.35 1600 0.30 GDP per capita (constant 2015 US$) 1500 relative GDP per capita 0.25 1400 1300 0.20 1200 0.15 1100 1000 0.10 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 GDP per capita (LHS) Relative GDP per capita (RHS) Source: World Development Indicators (accessed on 10/26/2023) and authors’ calculations. Notes: Relative GDP per capita is the ratio of real GDP per capita of the Sub-Saharan African region to that of the world. The grey area denotes the “Africa Rising” period from 2000 to 2014 when both real and relative GDP per capita trend up, though much less pronounced in terms of relative GDP per capita. Neither has the region converged vis-à-vis the global economy. A comparison of the evolution of the GDP per capita in the region vis-à-vis the world average shows that Sub-Saharan Africa has lost ground over the past six decades: the region’s relative income per capita roughly halved from 0.3 of the world’s average in 1960 to 0.15 in 2022 (Figure 1). In particular, the region’s relative income has significantly diverged from that of upper- and lower-middle income countries over the past decades (Figure 2) as well as that of East Asia and South Asia (Figure 3). A rough estimate of the cost of falling behind the rest of the world suggests that income per capita in 2022 could have been between 1.5-fold of its actual level (if it were to keep pace with the world’s output per capita) and 3-fold (if it were to keep pace with East Asia and the Pacific).6 Furthermore, the evolution of income per capita over the past six decades exhibits long swings, shifting between rising and declining growth (Figure 1, Table 1). Three distinct periods can be identified in the evolution of the real and relative GDP per capita of Sub-Saharan Africa during the period 1960-2022: (i) growing but at a slower pace than that of the world economy during the 1960s and 1970s; (ii) declining in 6 This is worrisome as the economic fallout from the different (covariate) shocks affecting the world economy—say, the pandemic, climate shocks, the war in Ukraine and global inflation—is partly deactivating the global growth engine for African economies, thus limiting their growth prospects over the near future. both real and relative terms in 1980-2000 as a result of inadequate economic policies, poor human capital development, and low levels of private investment,7 and (iii) rising and slightly faster than the world average during the Africa Rising period of 2000-14—with political stability, external tailwinds, robust investment and improved management delivering an increase in economic activity that kept pace with growth in the global economy. In the most recent decade, however, the upward trend in GDP per capita has reversed due to the economic fallout from a series of global shocks (commodities, COVID-19, and conflicts) and the limited macroeconomic policy space to stave off these shocks. Figure 2. Relative income per capita of SSA, Figure 3. Relative income per capita of SSA, 1990-2022: Compared with income groups 1990-2022: Compared with world regions Source: World Development Indicators (WDI). Notes: This figure depicts the relative real GDP per capita of Sub-Saharan Africa and other income and country groups at constant international PPP US dollars at 2015 prices (World=100). This paper examines the main features of real growth per capita in Sub-Saharan Africa over the past six decades to understand what drives growth of the region – including the miracle episodes – and why it has not been sustained. We will look at the different growth regimes across time and assess multiple statistical properties of growth in comparison with different benchmarks. Next, we look at the wide variety of country experiences in the region and provide a taxonomy of growth according to their empirical patterns. Additionally, the paper examines episodes of growth including the main properties of expansions and contractions. Furthermore, we will identify sources of growth using the growth accounting approach. Finally, we identified the incidence of growth miracles that have sustained growth over prolonged time periods and conducted case studies on three exceptional ones during their onsets (i.e., Botswana in the 1960s, Mauritius in the 1980s, and Ethiopia in the 2000s) to understand how a miracle episode might have started. We summarize the key findings of our paper below: First, trend growth in the region is low and presents patterns of long swings. Real GDP per capita of the region has experienced multiple regimes in the past six decades: growing during the 1960s and 1970s, declining in the following two decades, and rising again during the “Africa Rising” period from 2000 to the mid-2010s, before entering the recent “lost decade”. Second, growth patterns across countries in the region are heterogeneous and they could be classified into three broad groups: roughly one-third of the countries have been growing, one-third exhibits long swings, and the rest are stagnant, if not declining. 7 See Basu et al. (2000). Third, the region underperforms on duration, i.e., it experiences shorter expansions and longer recessions, while the speed of expansion/recession, i.e., growth per year, is more comparable with other regions. Fourth, the contribution of total factor productivity growth to economic performance in the region is small but has turned positive over the past two decades, especially among high-performing countries during the “Africa Rising” period. Fifth, a few exceptional growth miracles in the region are on par with the world’s best performance. They share common characteristics during the onsets that might have explained how a miracle starts: excellent leadership, economic diversification, market expansion, and investment for the future. The rest of the paper is divided into the following sections. Section 2 documents the main features of growth per capita in the Sub-Saharan Africa region from three perspectives: (a) analyze the basic statistical moments in the region’s aggregate GDP per capita across sub-periods; (b) present heterogenous patterns of growth across countries; and (c) investigate the main properties of growth episodes, i.e., expansions (trough- to-peak) and contractions (peak-to-trough). In Section 3, we will use the growth accounting approach to estimate the contribution of different factors and total factor productivity to growth per capita. In Section 4, we identify the incidence of growth miracles in the continent and examine best performers with case studies to draw insights on how a miracle episode started. Section 5 concludes. 2. Growth in Sub-Saharan Africa: Stylized Facts Economic performance in Sub-Saharan Africa, as measured by per capita growth and risk-adjusted growth, has been poorer than other regions or country income group benchmarks during most time periods in the past six decades with a few exceptions—most notably since the beginning of the 21st century when it grew at a faster pace than industrial countries. Aggregate regional growth performance hides wide heterogeneity across countries: roughly one-third of the countries have grown, one-third have presented patterns of long swings, and the rest have been stagnant if not declining over the past six decades. When examining growth by episodes of expansions and recessions, we see that the region spends less time in expansions, especially the oil-abundant and conflict-affected countries, when compared with the rest of the world. Meanwhile, a typical expansion is also shorter and shallower while a typical recession is more comparable in speed with that of other developing countries on average but slightly longer. 2.1 Aggregate growth in the region is low and volatile Growth in Sub-Saharan Africa is considerably slower than that of industrial and developing countries on average over the past six decades. The (weighted) average growth rate for the region was 0.8 percent for the period 1960-2019, which is about one-fourth of the average growth recorded by developing countries (3.3 percent) and nearly one-third of that of industrial countries (2.2 percent). The living standards of countries in Sub-Saharan Africa have continued diverging over time relative to other groups. For instance, the median income of industrial countries was more than 7-fold that of Sub-Saharan Africa in 1960 and this gap widened to 16-fold in 2019. The same trend can be observed relative to developing countries. The median income of this group was 2.5-fold that of Sub-Saharan Africa in 1960 and it became 4.5-fold in 2019 (see Table 1).8 Over the past 60 years, growth per capita in Sub-Saharan Africa has experienced long swings, shifting from periods of positive to negative growth and vice versa. For instance, (weighted average) growth per capita in the region was low but positive at 1.1 percent per year in 1960-80, then slowed sharply to an annual rate of -0.85 percent over the next two decades (1981-2000), wiping out the gains of the previous two decades. The Great Moderation and the super cycle of commodity prices boosted incomes in the region over the following two decades when the (weighted average) growth per capita increased to 2.1 percent in 2001- 2019—still slower than that of developing countries (3.7 percent) but greater than that of industrial countries (1.0 percent). Overall, Table 1 suggests that the narrative on the growth record of the region is not one of persistently poor performance but instead a case of short-lived, unsustained growth. Growth per capita in Sub-Saharan Africa is more volatile than that of industrial countries, and slightly lower than that of the developing countries throughout the 1960-2019 period. In fact, the average standard deviation of growth rates in the region is 1.4 percent as opposed to 0.5 percent among industrial countries and 1.7 among developing countries. Fluctuations in growth per capita in the region were more volatile in the 1981-2000 period (2.4 percent), but still less oscillating than those of developing countries in the same period (3.3 percent). The stable macroeconomic environment experienced by the global economy during the Great Moderation (from the 1990s to the 2007-08 financial crisis) led to lower growth volatility for all country groups—including Sub-Saharan Africa. The standard deviation of growth per capita in the region dropped to 1.5 percent in the 2001-19 period. Additionally, the lower volatility of Sub-Saharan Africa vis- 8 Table 1 computes central moments of the distribution of growth across countries and over time such as the mean, the standard deviation, and selected percentiles (the median as well as the bottom and top deciles), as well as a measure of risk-adjusted growth. These moments are calculated not only throughout the entire period (1960-2019) but also across sub-periods (1960-1980, 1981-2000, and 2001-2019). à-vis the rest of the developing world could be attributed to the fact that the former group had a lower exposure to and transmission of international and domestic financial shocks—as a result of the region’s underdeveloped financial systems and its lower degree of international financial integration. Table 1. Economic Growth: Summary Statistics Source: Penn World Tables 10.1. Authors’ calculations. Note: 1/ Risk-adjusted growth is computed as the ratio of the average to the standard deviation of growth per capita. Greater average growth is usually accompanied by greater risk-as measured by the volatility of growth per capita. Hence, measures of growth adjusted for that volatility can provide a better measure of return to risks in economic activity. Risk-adjusted growth of the region is lower relative to that of industrial and developing countries. Defined by the ratio of the mean to the standard deviation of growth per capita, risk-adjusted growth in the Sub-Saharan Africa is 1.0 from 1961 to 2019, a rate that is significantly smaller than that of advanced countries (2.2) and other developing countries (1.4). Risk-adjusted growth in Sub-Saharan Africa has also been episodic: it improved from 0.2 in 1981-2000 to 1.5 in the period 2001-19. This jump in risk adjusted growth reflects not only the increase in average growth per capita but also a reduction in growth volatility (see Table 1). 2.2 Growth in the region is heterogeneous The aggregate evidence presented so far conceals various growth experiences across countries in the region over the past six decades—partially captured in Table 1 by the top and bottom deciles of the growth per capita distribution in Sub-Saharan Africa. For instance, seven countries in the region recorded average annual per capita growth rates that exceeded 3 percent in the period 1960-2019, whereas income per capita contracted in fourteen countries during the same period. The average growth differential between these two groups of countries was about 5 percent—a margin that can produce rapid shifts in relative income per capita over time. By looking at the growth performance across subperiods, Table 1 hints that the evolution of income per capita for the region as a whole cannot be characterized by a time trend deterministic model (either upward or downward sloping). Instead, growth per capita in the region swings from periods of fair to poor performance. Still, the great deal of heterogeneity in the evolution of GDP per capita across countries in the region signals that: (a) there is a wide dispersion of growth per capita between top and bottom performers, and (b) every country is unique in its evolution. Inspecting the evolution of real GDP per capita for each individual country in the region, in the spirit of Pritchett (2000), reveals a diverse landscape of growth patterns in the region. Specifically, the different growth topographies exhibited by Sub-Saharan African countries can be classified into three broad groups: (a) countries with an upward-sloping trend (uphill), (b) countries with long swings (mountain ranges), and (c) stagnant countries (plains)—if not declining. Figure 4 describes the criteria that we used to classify the growth patterns of Sub-Saharan African countries—which is derived after visually examining the trends and fluctuations of each country’s GDP per capita time series. Table 2 lists the countries that fall into each of the three groups. Unlike industrial countries where growth is steady with occasional recessions followed by quick recoveries, long swings and stagnant periods are prevalent in the region which have contributed to the low, short-lived, and volatile growth at the regional level. Figure 4. Three General Growth Patterns of Sub-Saharan Africa, 1960s to 2010s. Group 1: Growing (Uphill) Upward trend overall; Values during the last decade is on average more than double the level in the beginning decade; and Values do not shrink by over one half of historical high in any year. Group 2: Swinging (Mountain Ranges) Outside Group 1; and The minimum and maximum values do not fall within the beginning and ending decades; and The maximum value is more than double the size of the minimum value. Group 3: Stagnant (Plain) Outside Group 1 and 2; and The maximum value is less than double the size of minimum value throughout the six decades. Note: The value(s) in the diagram refer to real GDP per capita (at chained PPPs in 2017 US$) from Penn World Table version 10.01 (Feenstra, Inklaar and Timmer 2015). Table 2. List of Countries by Groupings of Empirical Growth Patterns Note: Only countries with complete data from 1960 to 2019 are listed. Central African Republic and Niger are two countries whose change in GDP per capita is around (slightly exceeding) the threshold between swinging and stagnation but display an overall declining trend. D.R. Congo and Guinea exhibit patterns of long swings, but their average GDP per capita is also lower in 2010s as compared with 1960s. More granular topographies emerge within each of the three broad patterns. Within the group of growing countries, Botswana seems to be reaching a plateau with relatively low growth in the recent decades after rapid growth from the 1960s to the 2000s. After a rather stagnant performance over the last four decades of the twentieth century, Ethiopia emerges as a fast growing economy in the mid-2000s. Mauritius, on the other hand, has been growing throughout the period with intermittent recessions. Among countries with long swings in real income per capita, Nigeria’s living standards have peaked twice, around 1980 and again in the 2010s. Zambia reached the bottom of a basin in the 1990s, while Zimbabwe’s per capita growth dropped sharply in the late 2000s, and still has not recovered to its GDP per capita level in the 1990s. While stagnant countries look more similar from afar, details vary when zooming in. Income per capita in some of these countries is flatter (e.g., Burundi) than others (e.g., Togo), and some have been sliding on a gradual downward slope, such as the Central African Republic. Figure 5 presents countries’ growth trajectories that are representative of each sub-category. It should be noted that as the classification branches out, commonality gives way to diversity and the uniqueness of each series emerges. Figure 5. Diverse Landscapes of Sub-Saharan African Countries Growing Swinging Stagnant Note: The figure plots the natural logarithm of GDP at chained PPPs per capita (in 2017US$) and are classified based on criteria in Figure 4. Source: Penn World Table 10.01 (Feenstra, Inklaar and Timmer 2015). 2.3 Characterizing expansions and recessions in Sub-Saharan Africa To assess the episodic nature of economic growth in the region, we estimate the main features of the expansions and recessions of 45 Sub-Saharan African countries from 1960 to 2021. We benchmark their main characteristics with those of 23 advanced countries and 115 developing countries outside the SSA region. Our analysis will focus on the frequency, duration, and depth of expansions and recessions in Sub- Saharan Africa vis-à-vis other world regions over the past six decades.9 We use the Bry-Boschan algorithm to identify turning points (peaks and troughs) in real GDP per capita—as implemented by Harding (2002) for long time series annual data.10 After computing these turning points, recessions or contractions are 9 We conduct this analysis for all expansions and recessions over the past 60 years, whether these phases are incomplete or complete, and their features before the COVID-19 pandemic. We also compute the features of expansions and recessions subgroups within the Sub-Saharan African region. 10 Annex II describes the Bry-Boschan algorithm that identifies turning points and the computation of the different features of expansions (trough-to-peak phase) and recessions (peak-to-trough phase). defined as episodes from peak to trough in GDP per capita, while expansions are the episodes from trough to the subsequent peak. Fact 1. Sub-Saharan African countries tended to spend less time in expansions or more time in recessions than other developing countries over the past six decades. Sub-Saharan African countries, on average, tend to spend a lower share of their time in expansion—and a greater share of their time in recession—relative to other developing countries during the period 1961-2021. Specifically, Sub-Saharan African countries spent 63 percent of their time in expansion—a proportion that is significantly smaller to that of developing countries (76 percent). This is a first hint at the shorter life of economic expansions across countries in the region. Note that the difference in the share of time spent in an expansion when comparing SSA countries vis-à-vis other developing countries is higher among lower- middle income countries (65 percent for SSA countries vs. 81 percent for developing countries). The time spent in an expansion among SSA low-income countries is comparable to that of other non-SSA low- income countries—that is, 60 and 61 percent, respectively (Figure 6).11 We also find differences in the share of the time spent in expansions within Sub-Saharan Africa. For instance, non-resource abundant countries, on average, spent more time in expansions (66 percent) than resource abundant ones (58 percent). Unsurprisingly, fragile countries spent an even shorter time in an expansion (55 percent)—thus, reflecting the fact that these countries might be more prone to volatile shocks (say, commodity prices, conflict) and have an inadequate risk management capacity to address such shocks. This is captured by the poor state capacity in fragile economies. Figure 6. Time spent in expansions in Sub-Saharan Africa vis-à-vis other developing countries, 1961- 2021: By income levels Source: Penn World Tables 10.01 (Feenstra, Inklaar, and Timmer 2015) and authors’ calculations. 11 Within the non-SSA developing countries, we find that East Asian and South Asian countries, on average, tend to spend more time in expansions (85 percent) over the past six decades. In contrast, countries in the Middle East and North Africa only spend 68 percent of their time in an expansion. Fact 2. The duration of a typical expansion in Sub-Saharan Africa is shorter than those in other developing countries, while recessions are more comparable in duration yet slightly longer. The average duration of an expansion in economic activity for Sub-Saharan Africa is 3.3 years—which is considerably smaller than the duration of these trough-to-peak phases in output per capita for other developing countries (5.1 years) as well as for advanced economies (7.2 years). Among developing regions, the duration of expansions is the longest in East Asia and South Asia (7.2 and 6.2 years, respectively)— which is comparable to that of advanced economies. Other than Sub-Saharan Africa, the Middle East and North Africa as well as Latin America experienced the shortest expansions, with 3.9 and 4.4 years, respectively (Figure 7). Within Sub-Saharan Africa, expansions tend to be slightly longer in non-resource abundant countries than among resource abundant ones (3.5 and 2.9 years, respectively), while fragile countries have slightly shorter expansions (2.7 years). On the other hand, the duration of recessions across Sub-Saharan African countries (2 years) is more comparable to that of other developing countries (1.7 years) and advanced economies (1.5 years). Across world regions, recessions are the shortest in South Asia (1.2 years) and East Asia (1.6 years). Within the SSA region, recessions tend to be slightly shorter among non-resource abundant countries (1.9) vis-à-vis resource abundant countries (2.2 years) and fragile countries (2.1 years). Figure 7. Average duration of expansions and recessions across Sub-Saharan Africa and the world, 1961-2021 (in years) Source: Penn World Tables 10.01 (Feenstra, Inklaar, and Timmer 2015) and authors’ calculations. Note: Duration is defined as the number of years from peak to trough in real GPD per capita during contractions and from trough to next peak during expansions Fact 3. The speed of expansions in Sub-Saharan Africa, on average, is comparable to that of other developing countries. The median speed of expansions in Sub-Saharan Africa, as computed by the cumulative growth divided by the duration of such expansion, is 3.1 percent per year—which is smaller than that of other developing countries (3.5 percent per year) but greater than that of advanced economies (2.1 percent per year). The accumulated growth is computed as the maximum increase (drop) of GDP from trough (peak) to peak (trough) during episodes of expansion (contraction). The regions with the fastest expansion are East Asia (4 percent per year) and Eastern Europe and Central Asia (4.1 percent) (Figure 8). The fastest expansion in the latter region may capture the strong rebound of countries that were transitioning out of the Soviet Union. Within the Sub-Saharan African region, expansions in resource abundant countries are as fast as those in non-resource abundant countries (3.1 percent per year). However, recessions tend to be deeper in resource abundant vis-à-vis non-resource abundant countries (-3.4 and -2.4 per year, respectively). Figure 8. Median speed of expansions and recessions across Sub-Saharan Africa and the world, 1961-2021 (growth per year, percent) Source: Penn World Tables 10.01 (Feenstra, Inklaar, and Timmer 2015) and authors’ calculations. Note: The speed is computed as the cumulative growth divided by the duration of such expansions. The distribution of central moments of duration and amplitude across episodes Looking at the average duration and median amplitude of expansions and recessions along the lifetime of these different phases of economic activity confirm the findings described above: Expansions (trough-to-peak phase of economic activity) • Most expansions in Sub-Saharan Africa are short-lived. Nearly three-quarters of trough-to-peak episodes in the region have no more than 3 years of duration. The share of short-lived expansions is significantly smaller for advanced countries (about one third) and other developing countries (more than half). • The (average) duration of trough-to-peak episodes along the lifetime of episodes converges to 3 years in Sub-Saharan Africa—as opposed to 5 years in other developing countries and nearly 7 years among advanced economies (Figure 9.a). • The (median) depth of expansions in Sub-Saharan Africa along the cumulative distribution of episodes is 6.4 percent—which is considerably lower than that of advanced economies (11.2 percent) and other developing countries (10.7 percent) (Figure 9.b). Figure 9. Duration and amplitude of expansions along the lifetime of trough-to-peak episodes Source: Penn World Tables 10.01 (Feenstra, Inklaar, and Timmer 2015). Note: The horizontal axis indicates the set of episodes of duration up to n (n=1, … , 24). Recessions (peak-to-trough phase of economic activity) • Most recessions in Sub-Saharan Africa have a shorter life than its expansions. Nearly 4 in 5 recessions experienced by Sub-Saharan African countries tend to be short-lived (no more than 2 years of duration). This proportion is comparable with that of other developing countries (82 percent), and smaller than that of advanced economies (90 percent). • SSA has notably more episodes with long recessions than industrialized countries. The (average) duration of peak-to-trough episodes along the lifetime of such episodes converges to 2 years in Sub-Saharan Africa, and this average length is slightly higher than the 1.8 years in other developing countries and 1.5 years in advanced economies. Note that the average duration of these peak-to- trough episodes converge at different horizons across groups (Figure 10.a). • The (median) depth of recessions tend to be deeper in Sub-Saharan Africa (4.5 percent) as compared to advanced economies (2.3 percent) and other developing countries (4 percent). The medians tend to converge to these values at different lifetime horizons of these trough-to-peak episodes (Figure 10.b). Figure 10. Duration and amplitude of recessions along the lifetime of peak-to-trough episodes Source: Penn World Tables 10.01 (Feenstra, Inklaar, and Timmer 2015). Note: The horizontal axis indicates the set of episodes of duration up to n (n=1, … , 16). 3. Explaining Long-Term Economic Performance in Sub-Saharan Africa: A Growth Accounting Approach This section examines the sources of growth per capita across Sub-Saharan African countries over the past 60 years. More specifically, we examine whether the long-term growth record of countries in the region is primarily driven by either: (a) factors accumulation (including demographics) and/or (b) total factor productivity (TFP) growth. Demographic factors—such as the labor force participation rate and the share of working age population—are included in our analysis following the framework suggested by Loayza and Pennings (2022). We estimate the contribution of factor accumulation vis-à-vis TFP growth in Sub- Saharan Africa and compare its performance to other world regions, country groups within the continent, and by subperiods—with additional analysis for the period of high growth in the continent (i.e., Africa Rising).12 We find that physical capital accumulation explains the bulk of growth (about four-fifths) while the contribution of TFP growth is negative overall. In contrast, TFP growth accounts for one-third of growth per capita among industrial countries and about 15 percent among developing countries excluding Sub- Saharan Africa. Finally, the contribution of demographics is small in Sub-Saharan Africa over the past six decades, reflecting shifts in demographic structure that first inhibits before it enhances growth. Figure 11 plots the Solow decomposition of growth per capita for Sub-Saharan Africa and other world regions from 1961 to 2019. 12 Annex I describes the methodology used to examine the relative importance of factor accumulation and TFP growth in driving long-term growth per capita. Figure 11. Growth decomposition in Sub- Figure 12. Growth decomposition across Saharan Africa vs. World, 1961-2019 subregions in Sub-Saharan Africa, 1961-2019 Source: Penn World Tables 10.01 (Feenstra, Inklaar, and Timmer 2015). Note: Annex I presents the methodology to calculate the sources of growth. Given the heterogeneity of growth per capita experiences in Sub-Saharan Africa, it is likely that the relative importance of factor accumulation vis-à-vis TFP growth differs across sub-regions and countries. Figure 12 plots the sources of economic growth for countries in the region classified according to their resource abundance and their condition of fragility. We find that: (a) the contribution of physical and human capital accumulation is comparable among resource abundant and fragile countries, while physical capital accumulation explains the bulk of growth per capita among non-resource abundant countries over the past six decades; and, (b) TFP growth contributes negatively to economic performance in resource abundant and fragile countries (-0.5 and -0.3 percent, respectively), while this contribution is negligible in non-resource abundant countries from 1961 to 2019. Note that the large negative contribution of TFP growth in the growth decomposition may signal an inefficient use of the factors of production. As we have shown that income per capita does not increase monotonically over time but tend to be short- lived and episodic, we distinguish three different growth per capita regimes over the past six decades in Sub-Saharan Africa. Figure 13 plots the growth decomposition for Sub-Saharan Africa and subregions according to their degree of resource abundance and fragility conditions across different time periods; namely, 1961-79, 1980-99, and 2000-19. A few takeaways on the contribution of factor accumulation and TFP growth across the three regimes at the regional level and by country groups: • Factor accumulation. Physical and human capital accumulation largely explains growth per capita in Sub-Saharan Africa in 1961-79 and 2000-19—in fact, factor accumulation explained nearly 60 percent of growth per capita in the region. The contribution of factor accumulation for the region throughout the three different periods is similar to that of non-resource abundant countries. For resource abundant countries, physical capital contributes negatively to growth in the periods 1980-99 and 2000-19; however, human capital accumulation more than offsets the loss of capital deepening in those periods. In fragile countries, factor accumulation accounts for half of growth per capita in 2000-19. • Demographics. Employment and population dynamics has a positive contribution to growth per capita in the region during the period 2000-19—as opposed to a negative contribution in 1961-79, and 1980-99. Analogously, demographics contribute negatively among resource and non-resource abundant countries in the first two periods while the contribution is positive in 2000-19. In this latter period, the contribution of demographics is stronger among non-resource abundant countries. In the case of fragile countries, demographics contribute negatively in all three periods—which might reflect declines in the employment to population ratio. Figure 13. Growth decomposition over three sub-periods, 1961-2019 (percent per year) Source: Penn World Tables 10.01 (Feenstra, Inklaar, and Timmer 2015). Note: Annex I presents the methodology to calculate the sources of growth. • TFP growth. Growth in Sub-Saharan Africa during the first two periods, 1960-79 and 1980-99, is characterized by allocative inefficiencies—as signaled by the negative contribution of TFP growth (-0.6 and -0.9 percent, respectively). In 2000-19, TFP growth accounted for about 30 percent of growth per capita in the region. The same pattern of contribution is found among resource and non-resource abundant countries, although the relative importance of TFP growth is higher in the former group in 2000-19.13 The contribution of TFP growth among fragile countries is comparable to that of resource abundant countries throughout the different regimes. Africa Rising: What drives growth among high performers? During the ‘Africa Rising’ period, the annual average growth per capita of the region was about 3.3 percent—with 14 of 37 countries with available data meeting the label of ‘high performers.’ All 14 countries registered growth per capita that exceeded 3.5 percent per year during this period.14 Figure 14 plots the growth decomposition for the region and the 14 fastest-growing countries in the region from 2000 to 2014. Factor accumulation accounts for 45 percent of growth per capita in Sub-Saharan Africa 13 We need to take this result with caution as the expansion of TFP (that is, the portion of growth that is not captured by changes in the factors of production and demographics) might be driven by greater use of natural capital (say, energy, mineral and metal commodities). In fact, recent evidence shows that natural capital has accounted for more than half of the resource abundant countries’ growth per worker from 1996 to 2017 (Calderon 2021). 14 Note that this is the threshold rate of growth per capita that defines growth accelerations in Hausman, Pritchett, and Rodrik (2005). during this period while TFP growth explains about half of that performance. Additionally, demographics also contributed positively as the share of working age population increased. Figure 14. Growth decomposition of high performers during the Africa Rising period, 2000-14 -1 -2 Ghana SSA Mali Zambia Angola Liberia Nigeria Rwanda Sierra Leone Lesotho Tanzania Mauritius Zimbabwe Ethiopia Mozambique Physical Capital Human Capital Demographics TFP Output per capita Source: Penn World Tables 10.01 (Feenstra, Inklaar, and Timmer 2015). Note: SSA=Sub-Saharan Africa. The contribution of total factor productivity (TFP) growth is positive and large for all high-growth performers in Sub-Saharan Africa. TFP contributes over half of growth in nine of the 14 fastest growing economies in Figure 14 and is positive in the remaining five. It should be noted though that the high share of TFP contribution, the residual not explained by physical and human capital growth, in resource abundant countries might result from an intensive use of natural capital. Of the six resource abundant countries, two are oil abundant (Angola and Nigeria) while the remaining four are abundant in metals and minerals (Liberia, Mali, Sierra Leone, and Uganda). More specifically, accounting for the accumulation of natural capital reduces the contribution of TFP to growth per worker by almost 1 percentage point per year (Calderon 2021). On the other hand, factor accumulation accounts for more than half of growth per capita in five of the 14 fastest growing countries in the region, namely, Ethiopia, Mauritius, Mozambique, Rwanda, and Tanzania. 4. Growth Miracles: What Could We Learn from Successful Growth Stories? This section identifies “growth miracles”— as defined by episodes of sustained growth for a prolonged period—around the globe from 1960 to 2019. Once identified, we compare the episodes in Sub-Saharan Africa with those of other world regions over time and by sector, before synthesizing lessons from the region’s best performers. We find that “growth miracles” in the Sub-Saharan African region performs as well as those in other regions; however, they are in short supply. This implies that the successful growth experiences in the region are not learnt by the rest of a continent that still struggles with low and volatile growth. While growth performance across these “miracle episodes” are similarly impressive, we will explore their differences and commonalities. Across time, rapid growth episodes in the region were typically experienced during the Africa Rising period, after countries in the region emerged from deep recessions in the 1980s and 1990s. Across sectors, industrial activity tends to grow at a faster pace during miracle episodes, although services can also play a key role. Unlike other regions, agriculture still plays an important role in Africa and could also drive miracle episodes, especially through productivity improvement, such as in Mozambique and Ethiopia. It should be noted that a balanced growth across all sectors –with interlinkages and positive externalities among them– tends to be most sustainable. In the African continent, Botswana, Mauritius, and Ethiopia are three extraordinary cases whose growth has been sustained for decades and is still continuing today. At the end of this section, we analyze these three economies, focusing on their commonalities at the onset of their growth miracles, to learn what might have triggered the period of very rapid growth. We find that all these miracle episodes happens when the country has great leadership committed to economic development, diversifies into new industries beyond a single traditional sector, expands its markets globally, and invests for the future. 4.1 Uncovering and Understanding the Growth Miracles The underperformance of economic growth in the region hides several exceptional episodes of sustained growth over long periods—which we denote here as growth miracles. More specifically, we identified these growth miracles as those episodes where real GDP per capita grew over 2 percent continuously for over a decade.15 Our methodology follows the spirit of the work from Hausmann, Pritchett and Rodrik (2005), although we focus on identifying sustained rather than accelerated growth. To avoid arbitrary thresholds, we also conducted sensitivity analysis by modifying the threshold rate of growth per capita and duration that identify these miracle episodes; namely, growth per capita above 3% for over 8 years and above 3.5% for over 7 years. All identification criteria identified around one in five countries in the world that have experienced such an event in the past six decades, suggesting these growth miracles are indeed exceptional while also attainable. Table 3 presents the list of countries for which we were able to identify a growth miracle under each criterion. Among 35 countries that are identified by all three criteria, approximately one in nine is located in Sub-Saharan Africa. Meanwhile, a total of 13 Sub-Saharan African countries are identified by at least one of the three criteria. East Asia & Pacific and Europe & Central Asia have witnessed more miracle episodes. To obtain a more comprehensive understanding of how different regions perform when it comes to “growth miracles”, we computed a series of metrics, including duration, depth, and likelihood of these episodes using our baseline criterion, i.e., above 2% growth rate for over a decade (Table 4). Results show that Sub- Saharan African countries are capable of sustaining growth, and “growth miracles” in the region perform as well as, if not better than, those in other regions. Among “miracle countries” (columns c to g), Sub- Saharan Africa actually outperforms most other regions on the duration of the miracle episode, the annual growth rate during the miracle episode, and the accumulated expansion during the episode. The average duration of growth miracles is nearly 16 years across Sub-Saharan African miracle countries, which is lower only than that of the East Asia and the Pacific region (22.6 years). On average, GDP per capita at the end of the miracle episode is on average sixfold of their initial level among the identified Sub-Saharan African countries. Where the region falls short is the likelihood of such miracle episode. Only 6.4% of country-year units in the region fall into a miracle episode, compared with 10.2% in the rest of developing countries. These results suggest that a Sub-Saharan African country could embark on a sustained growth trajectory, 15 Berg et al. (2012) and Arizala et al. (2017) define growth spells as those periods characterized by a growth surge that is followed by a prolonged period of at least 2 percent average per capita income growth. and spreading the successful lessons from these growth miracles could transform miracles of a few to the miracle of a continent. Having identified these “growth miracles”, we examine the patterns of growth before and after the miracle periods. Providing a taxonomy or classification of growth miracles could be a first step in understanding how a growth miracle starts and ends. Figure 15 classified the miracle episodes into four groups based on the growth rate of real GDP per capita in the year before and after the miracle episodes. Using zero growth rates as the delimiter, we tried to distinguish whether a miracle emerged from a crisis (i.e., negative growth rate) or stagnation (i.e., positive but lower than 2%), and whether a miracle ended into a recession (hard landing) or slows down (soft landing). Figure 15 shows that globally, countries are distributed evenly across the four groups, and a Sub-Saharan African miracle episode is more likely to emerge from a recession and growth slows down after a period of fast expansion. We acknowledge that time horizon of one year only captures temporary transition to/from a miracle episode and is far from sufficient in understanding the long-run effect of a growth miracle. Therefore, we also plotted the evolution of GDP per capita over three decades—including the decade prior to and after the first decade of miracle episode—for selected Sub- Saharan African countries in Figure 16. The growth experiences of these miracle countries suggest that while miracle episodes are similarly impressive, how they start and end varies across countries. For instance, Mauritius’s economy was fluctuating before the episode, but growth was largely sustained afterwards. Nigeria was also struggling before the episode, but growth lost momentum afterwards before reversing the course. These results motivate us to look into individual cases, especially those with growth sustained for longer periods well beyond one decade as our model performers. Besides inspecting the evolution of growth miracles over time, we also zoom into the sectoral composition of growth. In particular, we examine the sectoral contributions from agriculture, industry and services during the growth miracle episodes identified earlier. Afterwards, we calculate the average annual growth rates of each of the three sectoral components during the miracle episodes, in comparison with the overall GDP per capita growth rate. These results for Sub-Saharan African countries in comparison with some of the best performers in the world are plotted in Figure 17. We found: • The industrial sector—which includes mining and quarrying, manufacturing, construction, and public utilities—exhibits the fastest growth rate in the majority of countries. • The service sector—which includes wholesale and retail trade, hotels and restaurants, transport, government, financial, professional, and personal services such as education, health care, and real estate services—is the fastest growing sector in some economies including China, India, and Nigeria. • The agriculture sector remains an important driver of growth in SSA countries including Mozambique, the only country in this sample where agriculture is the fastest growing sector, mostly due to productivity improvement. Table 3. Countries that have experienced growth miracles since 1960 Region Country Criteria 1 Criteria 2 Criteria 3 Number of Criteria Satisfied Botswana    3 Cabo Verde   2 Côte d'Ivoire  1 Equatorial Guinea    3 Ethiopia    3 Gabon  1 Sub-Saharan Africa Mauritius   2 Mozambique    3 Nigeria   2 Rwanda   2 Tanzania  1 Uganda  1 Zambia  1 Cambodia    3 China    3 Indonesia    3 Japan    3 Korea, Rep.    3 Lao PDR    3 East Asia & Pacific Malaysia    3 Myanmar    3 Philippines  1 Singapore    3 Thailand    3 Viet Nam    3 Albania    3 Armenia    3 Austria  1 Azerbaijan    3 Belarus    3 Belgium  1 Bosnia and Herzegovina  1 Bulgaria   2 Croatia  1 Estonia  1 France   2 Georgia    3 Greece    3 Europe & Central Hungary  1 Asia Ireland    3 Kazakhstan   2 Latvia    3 Lithuania   2 Norway  1 Portugal    3 Russian Federation   2 Serbia   2 Slovenia  1 Spain   2 Tajikistan    3 Turkmenistan    3 Ukraine  1 Uzbekistan    3 Antigua and Barbuda  1 Bahamas, The   2 Latin America & Brazil  1 Caribbean Paraguay    3 Trinidad and Tobago    3 Uruguay    3 Middle East & Iran, Islamic Rep.    3 North Africa Malta    3 Bangladesh    3 Bhutan    3 South Asia India    3 Nepal  1 Sri Lanka   2 Note: Countries in the list are identified as having experienced miracle growth episode(s) by three varying criteria: 1) growth above 2% for over a decade, 2) growth above 3% for over 8 years, and 3) growth above 3.5% for over 7 years. Table 4. “Growth Miracles” in Global Context Across the Region Among 'Miracle' Countries Likelihood of 'Miracles' a. Number b. Duration c. Number d. Number e. Duration f. Annual g. Ratio of h. Share of i. Total time j. Total timek. of countries of longest of 'miracle' of 'miracle' of 'miracle' growth rate GDP pc in 'miracle' periods in periods Likelihood Region expansion countries episodes episode during end year vs countries 'miracle' with data of 'miracle' 'miracle' start year (c/a) episodes Episodes periods (i/j) East Asia & Pacific 37 9.9 11 15 22.6 6.3 5.9 29. 7% 303 1587 19.1% Europe & Central Asia 58 9.1 19 20 14.1 7.2 2.6 32. 8% 278 2291 12.1% Latin America & Caribbean 35 7.8 5 5 13.4 5.9 2.1 14. 3% 67 1982 3.4% Middle East & North Africa 21 5.9 2 3 14.5 7.4 2.4 9.5 % 40 901 4.4% South Asia 8 11.0 5 6 14.2 5.1 2.0 62.5% 82 390 21.0% Sub-Saharan Africa 48 7.3 9 10 15.9 7.7 6.0 1 .8% 157 2461 6.4% Developing countries excl. SSA 154 8.5 37 44 16.6 6.6 3. 5 2 .0% 703 6882 10.2% Note: Values in columns b, e, f, g are regional average of country level data, and the country-level data in columns e and g are based on the longest 'miracle' episode if a country has more than one 'miracle' episodes. Figure 15. Beyond the “Miracles” – A Taxonomy Notes: 1) country-episodes are placed in one of the four areas of the coordinate plane based on the growth rate of real GDP per capita before and after the miracle episode as marked in the axes with 0% as the delimiter; 2) country- episodes without data before or after the episode (e.g., Botswana, Ethiopia) are not shown in the figure. Figure 16. Similar “Miracle” Episodes with Different Beginnings and Endings 250 200 150 100 50 0 T-10 T-8 T-6 T-4 T-2 T T+2 T+4 T+6 T+8 T+10T+12T+14T+16T+18T+20 Mauritius Mozambique Nigeria Rwanda Uganda Tanzania Note: T is the starting point of the “miracle” episode when we standardized all countries’ real GPD per capita as 100. Equatorial Guinea is removed as an outlier that has over 30 times of its initial GDP per capita in two decades. Figure 17. Sectoral Drivers During “Miracle” Periods Note: countries are sorted (left to right) based on overall growth rate during miracle episodes. Korea’s growth rates are averages across two episodes that is disconnected by one year of negative growth in 1980. Figure 18. Sectoral Drivers – Examples of Three Sub-Saharan Miracle Countries Time series of sectoral value added per capita (constant 2015$) Botswana 4500 4000 3500 3000 2500 2000 1500 1000 500 0 1960 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014 2017 2020 Agriculture Industry Services Mauritius 8000 7000 6000 5000 4000 3000 2000 1000 0 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 Agriculture Industry Services Ethiopia 350 300 250 200 150 100 50 0 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 Agriculture Industry Services Note: The series divide sectoral value added by total population as a decomposition of GDP per capita. 4.2 Case Studies: Seeking Commonalities across Three Exceptional Miracle Episodes Considering the overall growth and its time and sectoral dimensions, we selected three countries in the region—Botswana, Ethiopia, and Mauritius—for a deep dive into the drivers of their miracle episodes. These three countries have experienced high and sustained growth for decades that goes beyond the “Africa Rising” period with varied sectoral contributions (Figure 18). Do the three countries share anything in common behind the seemingly different sectoral drivers? In this sub-section, we look into these individual cases and seek commonalities across these successful stories to understand what might have explained how a sustained growth episode started. A similar approach was adopted in The Growth Report led by Michael Spence in which 13 growth miracles were studied and five commonalities were identified (Commission on Growth and Development 2008).16 At first glance, Botswana, Mauritius, and Ethiopia appear vastly different. Yet they have all achieved sustained growth for a prolonged period that is on par with the world’s best performers. Botswana embarked on a sustained growth path since as early as the 1960s during its transition into independence from colonial rule. Raising cattle had been the backbone of the country’s economic activities for which climate, soil, and vegetation are particularly well suited (World Bank 1971). Mauritius, a small island nation in the Indian Ocean, experienced a growth outburst in the early 1980s, over a decade after the country became independent in 1968. Already a middle-income country, Mauritius had relied upon sugar production that was most resilient to cyclone damages endured by the island. Ethiopia, Africa’s oldest independent country, started its impressive growth episode more recently in the 2000s. Largely an agrarian economy with agriculture accounting for over 80% of employment, Ethiopia was one of the poorest countries in the continent at the onset of its miracle expansion. These diverse initial geographical, political, and economic conditions have not constrained the sustained economic expansion for decades afterwards. Moreover, these countries faced similar challenges at the outset, which should resonate with many countries in the region that are still struggling with low and volatile growth. Before their economies took off, all three countries faced physical and financial constraints. Being either landlocked or island countries, the three nations faced substantial barriers or costs to connect with external markets. Their economies largely relied upon a few, if not a single, primary industries, be it cattle raising, sugar production, or rainfed agriculture. With protracted deficits, government had to rely on foreign aid or external financing. Undaunted by the challenges, these three countries have successfully transformed their economies to a new level. Nowadays, Botswana has grown from the world’s poorest country in the 1960s to an upper-middle-income economy aspiring to become a high-income one. Mauritius has teetered on the threshold between upper-middle- income and high-income economies in recent years. At its current trajectory, Ethiopia is rapidly approaching the middle-income status as its income per capita at constant price has grown over threefold in the past two decades. The economic transformation of these countries did not happen overnight, but as a result of solid foundations built and effective economic policies implemented. By seeking common factors among the three countries at the onset of their economic takeoff, we hope to offer some insights on what most likely worked to kick off a miracle growth episode. Our case studies are synthesized from various World Bank reports and academic papers for Botswana (World Bank 1971, 1978; Acemoglu, Johnson & Robinson 2003; Hillbom, 2008; Poteete, 2009; Lewin, 2011), Mauritius (World Bank 1982, 1983, 1985; Subramanian, 2009; Sobhee, 2009; Frankel, 2016; Zafar, 2011), and Ethiopia (World Bank 2002, 2007; 16 The 13 economic miracles identified in the Growth Report are Botswana, Brazil, China, Hong Kong, China, Indonesia, Japan, the Republic of Korea, Malaysia, Malta, Oman, Singapore, Taiwan, China, and Thailand. These economies have grown an average annual rate of 7 percent or more for 25 years or longer. Easterly, 2002; Bigsten, Gebreeyesus & Soderbom, 2009; Shumuye, 2015). We focus on likely common factors driving these case studies along four different but inter-related dimensions: a) leading with conscience, commitment and compassion, b) diversifying the economic structure, c) expanding the markets, and d) investing for the future. In explaining each driver, we will tell the stories from these three countries, as well as resorting to literature for broader evidence of these drivers. Leading with Conscience, Commitment and Compassion The mindset, vision, and capabilities of leaders play a critical role in pointing the economies in the right direction and moving things forward. The African leaders who took the helm of the economies at the onset of their miracle episodes are all charismatic figures well regarded and supported by the people in their countries. Sir Seretse Khama, Botswana’s first president, led the country into independence and remained as president until his death in 1980 after being reelected three times. In particular, the president played a crucial role in safeguarding mineral exploitation and the state-tribe relations, turning the resource curse into a blessing. Sir Anerood Jugnauth served as Mauritius’ Prime Minister or President for multiple terms including from 1982 to 1995 when the economy entered periods of fast growth. Jugnauth was largely credited for initiating the diversification of the Mauritius’s economy. Similarly, long serving and well supported, Sir Meles Zenawi took a leading role in Ethiopia’s economic reforms in the modern time from 1995 until his death in 2012. In his own writings, Zenawi has advocated for a developmental state that inhibits rent-seeking activities and invests in infrastructure and market support institutions. While many important characteristics of great leaders matter (Brady and Spence Ed., 2010), all these leaders were dedicated to economic development, unlike the military juntas, corrupt or weak administrations in other countries or time periods. The fact that they were all well supported by the people suggests that these leaders put emphasis on the wellbeing of their citizens and were able to achieve consensus across multiple interest groups. It should be noted that excellent leadership is not limited to one single figure but a cohort of decision makers. For instance, Ethiopia’s health sector leaders were crucial for the success of the country’s Health Extension Program which is a government-led community health service delivery program that improves access to preventive, wellness, and basic curative services (Bilal et al., 2011). The country’s chief executive(s) are usually responsible for championing or supporting the good economic reforms or policy agendas that we discuss below. Broader evidence from cross-country analysis shows that national leaders can greatly influence the economic growth of their countries. Jones and Olken (2005) used deaths of leaders while in office as a source of exogenous variation in leadership and found robust evidence that leaders matter for growth, as well as policy outcomes. Using similar approach as Jones and Olken (2005) and applied to Sub-Saharan Africa during the period from 1960 to 2014, 19 leadership transitions due to exogenous deaths (i.e., death due to natural cause or accidents) happened, and growth increases by a statistically significant 1.87 percentage points in the aftermath of the political transition and increases further two years after the transition (Calderon, 2015). The finding implies that while leadership matters, there might be a higher incidence of weak and shortsighted leaders than good ones in Sub-Saharan Africa. Diversifying the Economic Structure All three countries have discovered and expanded into new industries beyond traditional agricultural production. It is no doubt that discoveries of diamonds in Botswana have led to a booming industry which, together with copper and nickel mines, became a new engine of growth. While diamonds constitute a natural endowment, their discoveries and mining were made possible through intensive prospecting and investment. Annual development expenditure on mineral development had ballooned from tens of thousands of rands in the 1960s to millions of rands in the early 1970s, becoming the sector receiving the most budget allocation. New industries such as refining and retailing also sprang from the mining industry. Besides mining, a development corporation was established in 1970 to promote private enterprises in manufacturing and service sectors. Figure 18 shows that the service sector in Botswana has grown more sustainably into the recent periods than agriculture and industry. After the boom and bust of the sugar industry, Mauritius embraced manufacturing and tourism as two additional pillars of their economy which grew rapidly in the 1980s. In addition to driving growth, manufacturing – in particular textiles and other light manufacturing – played a crucial role in creating jobs, especially for females, and thus, addressing the rising unemployment issue that the densely populated island was facing. Public investments were more evenly distributed across the three pillar industries and infrastructure. In Ethiopia, structural transformation progressed at a slower pace as most of the population depended on subsistence agriculture, with wide education and skill deficits before the 2000s. Considering these challenging conditions, the government adopted a growth strategy known as Agricultural Development-Led Industrialization (ADLI) with the goals of transforming subsistence farming into market production and facilitating industrialization, including fostering forward and backward linkages between the two sectors. This strategy seems to have worked well for the country and growth rates across all sectors registered impressive figures of 7 to 12 percent around the mid-2000s. Limited diversification is a common feature of low-income countries which inhibits broad-based and sustained growth. Using cross-country data and case studies, Papageorgiouon and Spatafora (2012) review and extend the evidence that diversification is a crucial aspect of the development process. They concluded that increases in diversification have been associated with both higher growth and lower volatility, especially since 1995 and in LICs with better institutions. While manifested in the structure of economic activity and exported products, diversification of “know-how”17 and their reconfiguration underpins a prosperous society, a theory championed or developed by Hausmann et al. (2014). According to this theory, these “knowhows” are embedded in brains and human networks that are accumulated through years of experience and may require structural change of an economy. Other research identified a hump-shape relationship between (export) diversification and income per capita. This pattern suggests that diversification increases alongside development but reversed after reaching certain high-income level (i.e., around PPP $25,000), both processes are mostly attributed to extensive margins rather than intensive margins (Cadot, Carrère, and Strauss-Kahn, 2011). Expanding Access to Markets Improved or new production capacity needs to be absorbed by larger markets, which in turn drives production. All three countries moved closer to open trade regimes that were built upon friendly relationships established with major markets. After independence, Botswana has largely adopted a policy of non-alignment and pursued friendly relations with other countries. It quickly joined the UN, the Southern African Customs Union, and the European Economic Community (EEC) in the 1960s/1970s, partially leading to rapid increases in export volumes. The government also relied on foreign enterprises for the mining industry, in which the government secured notable equity stake, for investment and technologies. As a small island economy, Mauritius embraced an export-oriented strategy more openly. In 1970, Mauritius was the first commonwealth country to join the European Economic Community, which became an 17 In this context, “know-how” is defined as the tacit knowledge that is embedded in the brain, or as a particular wiring of the brain, which is harder to move around, different from tools, i.e., embodied knowledge, and recipes, i.e., codified knowledge (Hausmann 2016). important market for its burgeoning manufacturing industry, notably in the Export Processing Zone (EPZ), nurtured by foreign investment, in particular from Hong Kong SAR, China, whose investors had established markets in EEC. Additionally, Mauritius also enjoyed preferential access to foreign markets under the “Multi-Fiber Agreement”. Gradually, the country has also been diversifying its markets, including for the tourism sector by adding multiple airport access points across the globe. The openness of the EPZ and the export sector, on the other hand, did not necessarily take away the protection of the import-substituting sectors in the initial stage, denoted by Rodrik (1999) as the “heterodox opening”. Ethiopia also succeeded in growing exports of textiles, clothing, and footwear from almost zero at the beginning of the 2000s. Already it had access to major economies through mechanisms such as the US Africa Growth and Opportunity Act and the EU’s Everything but Arms initiative, Ethiopia started the accession process to the WTO in 2003 and continued to expand markets to countries in East Asia and Middle East. Meanwhile, service exports had also boomed, driven by strong performance of transport and tourism, in particular Ethiopian Airlines. Today, Ethiopia has more diversified markets than most of its peers (World Bank, 2022). External markets represent additional demand for national production, and trade policies could help facilitate their access. Moreover, favorable impact from exports could go beyond its direct contribution to GDP, through other channels including improving total factor productivity that help sustain growth for a longer run. A rich body of research offers empirical evidence on the association of export expansion and economic growth. For instance, Kavoussi (1984) studied a sample of 73 developing countries for the 1960- 1978 period and found evidence in support of export expansion’s positive impact on growth and TFP among both low-income and middle-income countries. The reality is that Sub-Saharan Africa still represents a small share of global trade (around 3%), but it is growing rapidly; on the other hand, the share of trade in national income is substantial in many economies that are vulnerable to external shocks—i.e., total exports as share of GDP is 53% in Sub-Saharan Africa in 2019 (Coulibaly, Kassa, and Zeufack 2022).18 Expanding access to markets, along with an economic and export product diversification, help mitigate external shocks and sustain growth. Investing into the Future Meanwhile, the initial resources accumulated in these countries were responsibly and efficiently managed and invested for the future to sustain the growth, notably in infrastructure, human capital, and environmental sustainability. Botswana is well known for its prudent management of mineral resources. It established the Pula Fund, a sovereign wealth fund, to administer and save profits from mining activities for future generations. Important portions of these mineral revenues, one of the highest in Africa and comparable with Norway, were directed to infrastructure, education, and health care that are fundamental to sustained growth. When developing cattle production further, the Botswana government had sustainability in mind. The government prepared legislation for conservation and improvement of agricultural land to curtail overgrazing early on. Mauritius had a relatively well-educated population in the 1980s, thanks partly to the country’s universal education, propelling the country’s export-oriented strategy. In the 1980s, the government’s focus turned to improving the quality, rather than the quantity, of education. Public investment also focused on basic infrastructure, and the government encouraged investment in productive sectors such as manufacturing rather than consumption such as residential housing. The country also ventured into renewable energy by developing alternative usage of bagasse, a byproduct of its traditional sugar industry. In Ethiopia, there was a massive scale-up of primary education to the extent that the primary 18 Coulibaly et al. (2022) also identify trade agreements, global value chains and offshoring, regional integration, and trade in services as important ingredients for a successful market access strategy. gross enrollment rate rose from 20% in 1993 to 79% in 2004. Infant mortality also improved with lower rates than other countries of similar GDP per capita. Similarly, the country also invested heavily in infrastructure, and notably, the Export Processing Zones to facilitate development of light manufacturing as part of the country’s Growth and Transformation Plan, as well as hydroelectricity that builds upon the country’s relatively generous river resources as part of the country’s green development. Fostering the accumulation of physical, human, and natural capital strengthens the root of an economy so that it can branch out into diversified products and external markets. It has been argued that successful integration into the world economy requires a complementary set of policies and institutions at home where the state plays a crucial role: the evidence suggests that the countries that have grown most rapidly since the mid-1970s are those that have invested a high share of GDP and maintained macroeconomic stability (Rodrik 1999). More recent evidence on a panel of countries over time argues that public investment has a positive effect on long-term growth and labor productivity, and increases the speed of convergence of catching-up countries (Fournier 2016).19 Emerging evidence, such as from the Global Commission on the Economy and Climate, which was founded by seven countries including Ethiopia, has also demonstrated that investment in environmental sustainability is also complementary and essential for sustained economic growth (New Climate Economy 2018). 19 The paper suggests that the impact of public investment differs across sectors (i.e., more effective in areas such as health and research & development) and levels of capital stock (i.e., growth gains decline at higher levels of capital stock). 5. Conclusions The goal of this paper was to characterize the economic performance of Sub-Saharan African economies over the long term and uncover sources of growth. Our analysis was conducted along different dimensions: First, we examined the central moments of growth per capita (i.e., average, standard deviation, selected percentiles) at the regional level over the past six decades and sub-periods, benchmarking with those of the industrial and developing countries. Second, we presented a taxonomy of growth heterogeneity across countries in the region, which captures a wide array of topography, and clustered countries into three broad groups. Third, we computed the main features of episodic growth expansions and recessions after applying the Bry-Boschan algorithm that identifies turning points in real GDP per capita. Fourth, we uncovered the sources of economic growth using the Solow decomposition for the region and country groups across different time periods. Fifth, we identified ‘growth’ miracles across the globe and examined those in Sub- Saharan Africa—including the sectoral composition of those episodes. Finally, we sought commonalities at the onset of three exceptionally sustained growth episodes in the region, namely, Botswana, Ethiopia, and Mauritius. We found that the real income per capita of Sub-Saharan Africa exhibits long swings over the past six decades and has failed to converge with aspirational benchmarks. Country growth experiences are widely heterogeneous, and we clustered them into three broad groups inspired by topography—growing (uphill), long swings (mountain ranges), and stagnant (plains) if not declining. Our assessment of episodic expansions and recessions found that the region underperforms in duration, which leads to shallower expansions and deeper recessions as compared with the rest of the developing world. The growth accounting approach suggests that the contribution of TFP is small, if not negative, in the region, although it has turned positive over the recent two decades, and plays an important role in the growth of good performers. The region has also witnessed a few growth miracles that sustained growth for prolonged periods, on par with the world’s best performance. Seeking commonalities among three exceptional growth miracles during their onsets, we found that leadership, economic diversification, market expansion, and investment for the future might explain how a growth miracle starts. Our findings on the sources of growth help explain the lack of sustained growth in the region. Leadership is associated with TFP, a main driver of growth, as it boosts efficiency in the use of the factors of production. Diversifying the economy and expanding the markets both help mitigate risks and build resilience to external shocks. Investing in the future plays a counter-cyclical role and fosters inter-generational equity, besides strengthening the foundation of growth. The exceptional growth miracles of the region suggest that Sub-Saharan Africa has great potential to grow, if valuable experiences from the miracle episodes are learned and implemented by the rest of the region that still struggles with low and volatile growth. References Abramovitz, Moses. 1956. “Resource and Output Trends in the United States Since 1870.” American Economic Review 46 (2): 5–23. Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2003. “An African Success Story: Botswana.” In: Rodrik, Dani (ed.) In Search of Prosperity: Analytic Narratives on Economic Growth. Princeton: Princeton University Press, pp. 80-119 Arizala, Francisco, Jesus Gonzalez-Garcia, Charalambos Tsangarides, and Mustafa Yenice. 2017. “Growth Breaks and Growth Spells in Sub-Saharan Africa.” IMF Working Paper WP/17/202, September. Baah, Samuel K.T., Andres Castañeda Aguilar, Carolina Diaz-Bonilla, Tony Fujs, Christoph Lakner, Minh Cong Nguyen, and Martha Viveros. 2023. “March 2023 global poverty update from the World Bank: the challenge of estimating poverty in the pandemic.” World Bank, Data Blog, Washington, DC. Barro, Robert J., and Jong-Wha Lee. 2013. “A New Data Set of Educational Attainment in the World, 1950–2010.” Journal of Development Economics 104: 184–98. Basu, Anupam, Evangelos A. Calamitsis, and Dhaneshwar Ghura. 2000. “Promoting Growth in Sub- Saharan Africa: Learning What Works.” IMF Economic Issues No. 23, August. International Monetary Fund, Washington, DC. Berg, A., J. Ostry, and J. Zettelmeyer. 2012. “What Makes Growth Sustained?” Journal of Development Economics 98: 149–66. Bigsten, A., Gebreeyesus, M., & Söderbom, M. (2009). Gradual trade liberalization and firm performance in Ethiopia. Bilal, N., Herbst, C., Zhao, F., Soucat, A., & Lemiere, C. (2011). Health Extension Workers in Ethiopia: Improved Access and Coverage for the Rural Poor. In P. Chuhan-Pole, & M. Angwafo, Yes Africa Can: Success Stories from a Dynamic Continent (pp. 433-444). Washington, D.C.: The World Bank. Brady, D., and Spence, M. (Eds.). (2010). Leadership and growth. World Bank Publications. Calderon, Cesar. 2015. Africa’s Macroeconomic Vulnerabilities. Washington, DC: The World Bank, unpublished manuscript. Calderon, Cesar. 2021. Boosting Productivity in Sub-Saharan Africa: Policies and Institutions to Promote Efficiency. Washington, D.C.: World Bank. Cadot, O., Carrère, C., and Strauss-Kahn, V. (2011). Export diversification: what's behind the hump?. Review of Economics and Statistics, 93(2), 590-605. Claessens, S., M.A. Kose, and M.E. Terrones, 2009. “What happens during recessions, crunches and busts?” Economic Policy 24, 653-700 Claessens, S., M.A. Kose, and M.E. Terrones, 2011a. “Financial cycles: What? How? When?” In: Clarida, R., and F. Giavazzi (eds.) NBER International Seminar on Macroeconomics 2010, Chicago, IL: University of Chicago Press for NBER, pp. 303-343 Claessens, S., M.A. Kose, and M.E. Terrones, 2011b. “Recessions and Financial Disruptions in Emerging Markets: A Bird’s Eye View.” In: Céspedes, L.F., R. Chang, and D. Saravia (Eds.) Monetary Policy under Financial Turbulence, Santiago, Chile: Series on Central Banking, Analysis, and Economic Policies, Volume 16, pp. 59-104 Claessens, S., M.A. Kose, and M.E. Terrones, 2011c. “How Do Business and Financial Cycles Interact?” IMF Working Paper WP/11/88, April Coulibaly, S., Kassa, W., & Zeufack, A. G. (Eds.). (2022). Africa in the new trade environment: Market access in troubled times. World Bank Publications. Denison, Edward F. 1962. “Sources of Growth in the United States and the Alternatives before Us.” Supplement Paper No. 13, Committee for Economic Development, New York. Denison, Edward F., Zvi Griliches, and Dale W. Jorgenson. 1972. “The Measurement of Productivity.” Survey of Current Business 52 (5) Part II: 3–111. Easterly, W. (2002). Growth in Ethiopia: retrospect and prospect. Center for Global Development, Institute of International Studies, Washington, DC. Fosu, A.K. (2011) The African Economic Growth Record, and the Roles of Policy Syndromes and Governance. In: A. Noman, K. Botchwey, H. Stein, and J.E. Stiglitz, editors, Good Growth and Governance in Africa: Rethinking Development Strategies. Oxford University Press, pp. 175-218. Fournier, J. M. (2016). The positive effect of public investment on potential growth. Frankel, J. (2016). Mauritius: African success story. In African successes, volume IV: Sustainable growth (pp. 295-342). University of Chicago Press. Harding, D., Pagan, A., 2002a. “Dissecting the cycle: a methodological investigation.” Journal of Monetary Economics, 29: 365-381. Harding, Don. 2002. “The Australian Business Cycle: A New View.” Melbourne Institute of Applied Economic and Social Research, The University of Melbourne, manuscript. Hausmann, Ricardo. 2016. “Economic Development and the Accumulation of Know-how.” Welsh Economic Review 24: 13-16. Hausmann, R., Hidalgo, C. A., Bustos, S., Coscia, M., & Simoes, A. (2014). The atlas of economic complexity: Mapping paths to prosperity. Mit Press. Hausmann, R., Pritchett, L., & Rodrik, D. (2005). Growth accelerations. Journal of economic growth, 10, 303-329. Hillbom, E. (2008). Diamonds or development? A structural assessment of Botswana's forty years of success. The Journal of Modern African Studies, 46(2), 191-214. Jones, B. F., & Olken, B. A. (2005). Do leaders matter? National leadership and growth since World War II. The Quarterly Journal of Economics, 120(3), 835-864. Jorgenson, Dale W., and Zvi Griliches. 1967. “The Explanation of Productivity Change.” Review of Economic Studies 34 (3): 249–83. Kavoussi, R. M. (1984). Export expansion and economic growth: Further empirical evidence. Journal of development economics, 14(1), 241-250. Lewin, M. (2011). Botswana's Success: Good Governance, Good Policies, and Good Luck. In P. Chuhan- Pole, & M. Angwafo, Yes Africa Can: Success Stories from a Dynamic Continent (pp. 81-90). Washington, D.C.: The World Bank. Loayza, Norman V. and Steven Pennings. 2022. The Long-Term Growth Model: Fundamentals, Extensions, and Applications. Washington, D.C.: World Bank Group. Papageorgiou, M. C., & Spatafora, M. N. (2012). Economic diversification in LICs: Stylized facts and macroeconomic implications. International Monetary Fund. Poteete, A. R. (2009). Is development path dependent or political? A reinterpretation of mineral- dependent development in Botswana. The Journal of Development Studies, 45(4), 544-571. Pritchett, Lant. 2020. “Understanding Patterns of Economic Growth: Searching for Hills among Plateaus, Mountains, and Plains.” The World Bank Economic Review 14(2): 221-250. Ravn, Morten O., and Harald Uhlig. 2022. “On Adjusting the Hodrick-Prescott Filter for the Frequency of Observations.” The Review of Economics and Statistics 84(2): 371–380 Rodrik, D. (1999). The New Global Economy: Making Openess Work. Washington, D.C.: Overseas Development Council. Shumuye, M. Y. (2015). The role of developmental state in development: The case of Ethiopia. International Journal of African and Asian Studies, 14, 96-110. Sobhee, S. K. (2009). The economic success of Mauritius: Lessons and policy options for Africa. Journal of Economic Policy Reform, 12(1), 29-42. Solow, Robert M. 1957. “Technical Change and the Aggregate Production Function.” Review of Economics and Statistics 39 (3): 312–20. Subramanian, A. (2013). The Mauritian success story and its lessons. Achieving development success: Strategies and lessons from the developing world, 204-231. World Bank. (1971). Memorandum on Recent Economic Development and Prospects. Washington, D.C.: World Bank. World Bank. (1978). Economic Memorandum on Botswana. Washington, D.C.: World Bank. World Bank. (1982). Mauritius Economic Memorandum: Recend Developments and Prospects. Washington, D.C.: World Bank. World Bank. (1983). Mauritius Updating Economic Note: Recent Developments and Prospects. Washington, D.C.: World Bank. World Bank. (1985). Mauritius Adjustment and Growth: Country Economic Memorandum . Washington, D.C.: World Bank. World Bank. (2002). The Federal Democratic Republic of Ethiopia: Developing Exports to Promote Growth. Washington, D.C.: World Bank. World Bank. (2007). Ethiopia Accelerating Equitable Growth: Country Economic Memorandum. Washington, D.C.: World Bank. World Bank. (2022). Ethiopia's Great Transition: The Next Mile. Washington, D.C.: The World Bank. Zafar, A. (2011). Mauritius: An Economic Success Story. In P. Chuhan-Pole, & M. Angwafo, Yes Africa Can: Success Stories from a Dynamic Continent (pp. 91-106). Washington, D.C.: The World Bank. Annex I. Growth Accounting Assessing the sources of economic growth dates back to the late 1950s, with growth in real output being decomposed as the weighted average of the growth rate of labor, capital and total factor productivity (TFP). 1 Labeled as the ‘Solow residual’, TFP growth is the unexplained part of economic growth and interpreted as a measure of technological change. These calculations became more sophisticated as more general production functions and more accurate measurements of inputs and outputs were used in the 1960s and 1970s—including changes in the quantity and quality of labor and capital inputs. 2 To conduct the growth decomposition, it is assumed that technology is represented by a Cobb-Douglas production function with constant returns to scale: 3 1− ( = ℎ ) where is the gross domestic output (GDP), is the capital stock, ℎ is the effective labor used in production (or labor adjusted for human capital, h), is the total factor productivity (TFP), and α represents the labor share. Additionally, the number of workers can be decomposed into = , where is the labor participation rate (i.e., the ratio of labor force to working age population), is the working-age population to total population ratio, and is the total population. 4 Real output per capita ( ) is then defined as: 1− = = = = ℎ where is output per worker, and is capital per worker. If we define � = , then growth in real output per worker (‘labor productivity’) is: � � + ℎ ̂ + (1 − ) � = This equation suggests that growth in labor productivity is driven by capital deepening, growth in human capital per worker and total factor productivity growth. Meanwhile, growth in real output per capita is the labor productivity growth adjusted for changes in labor force participation and the share of working age population: � = � + � + � The comparison of income per capita and labor productivity levels as well as sources of growth in Sub-Saharan Africa relative to other world regions requires a dataset with ample coverage across countries and over time. To conduct this analysis at the aggregate level, this report uses PWT 10.1 data with annual information from 1950 to 2019 for a wide array of countries in the world (Feenstra, Inklaar, and Timmer 2015). Output. Computing income per capita requires the use of the expenditure approach to measure the level of economic activity (GDPe) from the PWT database. This measure captures the standard of living of the different countries in the world. On the other hand, the data on real GDP estimated from the output perspective (GDPo) 1 Abramovitz (1956); Solow (1957). 2 Denison (1962); Denison, Griliches, and Jorgenson (1972); Jorgenson and Griliches (1967). 3 Hall and Jones (1999); Caselli (2005); Loayza and Pennings (2022). 4 Loayza and Pennings (2022) assume: (a) no unemployment or underemployment, (b) no adjustment costs in capital accumulation and, (c) perfect competition in the markets of production factors. is a better approximation of the total production of the economy. Hence, it is a more adequate proxy for the output capacity of an economy. 5 Physical capital. The stock of physical capital is estimated based on the accumulation and depreciation of past investments using the perpetual inventory method (PIM). One of the novel aspects of the estimation of the aggregate capital stock since PWT 9.0 is the use of investments disaggregated by type of asset. The data on investments by asset type is obtained from either the national accounts or partly estimated using the commodity-flow method in the spirit of Caselli and Wilson (2004). The average depreciation rate varies across countries and over time because the asset composition differs across countries and the depreciation rate is not similar across assets. In addition, PWT uses information on the asset composition of the capital stock to compute the relative price of investment. Human capital per worker. The index of human capital per worker, h, is constructed using the average years of schooling (s) in the population over 25 years old (Barro and Lee 2013). Following Hall and Jones (1999), the years of schooling are converted into a measure of h through the formula ℎ = �()�, with () being a piece-wise linear function (as in Caselli 2005): 0.134 ∙ ; ≤ 4 () = � 0.134 ∙ 4 + 0.101 ∙ ( − 4); 4 < ≤ 8 0.134 ∙ 4 + 0.101 ∙ 4 + 0.068 ∙ ( − 8); > 8 If the wage-schooling relationship is log-linear, the relationship between h and s should also be log-linear. The PWT10.1 constructs the h index for 150 countries using data on schooling years from Barro and Lee (2013) for 95 countries, and from Cohen and Soto (2007) and Cohen and Leker (2014) for an additional 55 countries. International data on education-wage profiles suggest that the return to an additional year of education in Sub-Saharan Africa, the region with the lowest level of education, is about 13.4 percent. In comparison, the return to an additional year of education is 10.1 percent for the world and 6.8 percent for the OECD countries (Psacharopoulos 1994). This measure of human capital tries to reconcile the properties of a log-linear relationship between education and income at the country level with the concavity of that relationship across countries. The h index from the PWT assumes homogeneous returns across countries. Labor share of income. Finally, the PWT has estimated the labor share (or the share of labor income in economic activity) for a wide array of countries and years. There is broad availability of information on labor compensation of employees; however, a separate estimation is needed for the labor compensation of self- employed workers. The cross-country estimates of the labor share yield some stylized facts (Feenstra, Inklaar, and Timmer 2015): (a) the global average of the labor share in income is about 0.52 (significantly lower than the two-thirds typically assumed in the macroeconomic literature); (b) there is no systematic relationship between labor shares and income per capita levels; and, (c) labor shares have declined over time in most of the countries covered. 6 5 According to the PWT methodology, countries with strong terms of trade will have a higher real GDPe than GDPO. 6 See also IMF (2017). Annex II. Identifying turning points in real GDP per capita This annex outlines the methodology used to characterize business cycles for a sample of countries worldwide—including Sub-Saharan African countries. Our goal is to identify turning points (peaks and troughs) in historical series of real GDP per capita to characterize business cycles for a wide array of countries. 7 To accomplish this task, we use the annual version of the Bry-Boschan algorithm, as implemented by Harding (2002) and Harding and Pagan (2002), on GDP per capita for a wide array of countries over the past six decades. 8 The classical approach to analyzing business cycle follows the seminal work of Burns and Mitchell (1946). It identifies turning points in an aggregate series—typically, the level of real output or employment—to characterize business cycles as a sequence of expansions and contractions.9 The rule that < (>) 0 to the right (left) of a local peak (trough) provides a starting point for locating turning points in an aggregate output series. Hence, a local peak in a series occurs in period t if > for t-k < s < t and t+k > s > t, where k determines some symmetric window in time around t. A local trough can be defined in an analogous way. The choice of k relies on the time frequency of the series under consideration. In our case, k=1 for annual data analysis. A discrete version of the calculus rule described above can be summarized as follows: • A peak at time t occurs if ∆ > 0 and ∆+1 ≤ 0 • A trough at time t occurs if ∆ < 0 and ∆+1 ≥ 0 This rule requires that complete cycles run from peak to peak and have two phases—namely, contraction (peak to trough) and expansion (trough to peak), and that peaks and troughs must alternate. 10 We use annual data for a wide series of countries from 1960 to 2021 as quarterly data is unavailable for a large number of countries—and, particularly, countries in Sub-Saharan Africa. Note that although quarterly data provides a clearer picture of the business cycle than does annual data, the latter can provide useful information about major expansions and recessions for a large time series. After computing the turning points in real output, the main features of contractions (from peak to trough) and expansions (from trough to subsequent peak) in real income per capita are characterized in terms of their duration, amplitude and speed: 7 Business cycles are characterized by more than just turning points in real GDP. Having said this, the paper focuses on documenting differences in duration amplitude, output change and slope of recessions and upturns for a large group of economies using real GDP per capita, rather than describe a large number of series. 8 Predicting peaks and troughs in income per capita (or in labor productivity) in real time or comparing turning points among dating methods goes beyond the scope of this paper. 9 Alternatively, empirical business cycle research has focused on identifying “growth cycles” by calculating deviations from long-run trends —with these trends being estimated using different de-trending techniques (say, deterministic trend models, the Hodrick-Prescott filter, and the band-pass filter, among others). However, the literature argues that this methodology tends to over-estimate the frequency of turning points and under-estimate their amplitude when compared to classical cycles (Morsink, Helbling and Tokarick, 2002). Also, the dating of turning points using growth cycles rather than classical ones is sensitive to the inclusion of new data (Claessens et al. 2009; 2011a,b,c). 10 For annual data, the minimum duration of a complete cycle is of at last two years and each phase of the cycle must last at least one year. The time thresholds are equal to 5 and 2 quarters in the case of the application of the algorithm to quarterly data. • Duration. It is computed as the number of years from peak to trough during contraction episodes and from trough to the next peak in the expansion phase. • Amplitude. It is calculated as the maximum drop (increase) of GDP from peak (trough) to trough (peak) during episodes of contraction (expansion). For instance, the amplitude of a contraction, AC, measures the change in the real GDP from a peak (y0) to the trough (yK), that is, AC = yK- y0. • Speed. It is computed as the ratio of the amplitude of the peak-to-tough (trough-to next peak) phase of the cycle to its corresponding duration.