89548 3rd Ethiopia Economic Update: STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS 3RD ETHIOPIA ECONOMIC UPDATE STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS JUNE 2014 v TABLE OF CONTENTS ACKNOWLEDGEMENTS.................................................................................................................................. v LIST OF ABBREVIATIONS.............................................................................................................................. vii EXECUTIVE SUMMARY................................................................................................................................... ix RECENT ECONOMIC DEVELOPMENTS...................................................................................................... 1 The Short View...........................................................................................................................................................1 The Long View: Structural Change.............................................................................................................................9 EXPORT PERFORMANCE AND COMPETITIVENESS.............................................................................. 17 Export Performance..................................................................................................................................................18 Growth of Exports in Agriculture, Manufacturing, and Services............................................................................18 Structural Transformation through Diversification................................................................................................23 Sophistication and Quality....................................................................................................................................24 Sustainability of the Export Sector........................................................................................................................28 Export Competitiveness............................................................................................................................................32 Incentive Framework for Trade..............................................................................................................................32 Factor Inputs, Productivity, and Trade Costs.........................................................................................................38 Proactive Policies to Support Trade........................................................................................................................41 Competing on Price: Does the Real Exchange Rate Matter? .................................................................................43 SUMMARY AND POLICY RECOMMENDATIONS..................................................................................... 49 Annex 1: Ethiopia: Selected Economic Indicators (High Frequency)........................................................................52 Annex 2: Ethiopia: Selected Economic and Social Indicators (Annual Frequency)....................................................53 Annex 3: Product Space Analysis of Ethiopian Exports.............................................................................................56 Annex 4: Exporter Dynamics in Selected Products (2008–12)..................................................................................60 Annex 5: Additional details on RER calculations and theoretical considerations.......................................................79 REFERENCES.................................................................................................................................................... 83 LIST OF FIGURES Figure 1.1: Economic Activity......................................................................................................................................2 Figure 1.2: Monetary Sector.........................................................................................................................................4 Figure 1.3: Fiscal Sector................................................................................................................................................6 Figure 1.4: External Sector...........................................................................................................................................7 Figure 1.5: External sector (continued).........................................................................................................................8 vi 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS Figure 1.6: Ethiopia: Output and Employment by Sector, 1996–2011.......................................................................11 Figure 1.7: Ethiopia: Labor Productivity, 1996–2011.................................................................................................13 Figure 1.8: Ethiopia: Labor Productivity and Structural Change, 1996–2011.............................................................15 Figure 2.1: Trade Competitiveness Diagnostic Framework.........................................................................................18 Figure 2.2: Growth and Orientation of Exports..........................................................................................................22 Figure 2.3: Export Diversification and Number of Exporters......................................................................................25 Figure 2.4: Size Distribution of Ethiopian Exporters..................................................................................................26 Figure 2.5: Export Sophistication and Survival...........................................................................................................29 Figure 2.5: Export Sophistication and Survival (continued)........................................................................................30 Figure 2.6: Exporter Dynamics: Entry, Exit, Size and Survival....................................................................................33 Figure 2.7: Competitiveness and Economic Complexity ............................................................................................38 Figure 2.8: Undervaluation and Export Growth, Ethiopia and Selected Countries.....................................................45 Figure 3.1: Ethiopia’s Product Maps...........................................................................................................................57 Figure 3.2: Complexity and Density Tradeoff in Ethiopia...........................................................................................58 Figure 3.3: Exporter Dynamics: Coffee ......................................................................................................................61 Figure 3.4: Dynamics of Coffee Exporters in Main Destinations ...............................................................................62 Figure 3.5: Exporter Dynamics: Oil seeds ..................................................................................................................63 Figure 3.6: Dynamics of Oil Seeds Exporters in Main Destinations ...........................................................................64 Figure 3.7: Exporter Dynamics: Cut flowers...............................................................................................................65 Figure 3.8: Dynamics of Cut Flowers Exporters in Main Destinations .......................................................................66 Figure 3.9: Exporter Dynamics: Textile and Garments Exports ..................................................................................68 Figure 3.10: Dynamics of Textile Exporters in Main Destinations................................................................................69 Figure 3.11: Exporter Dynamics: Leather and Leather Exports ....................................................................................70 Figure 3.12: Dynamics of Leather Exporters in Main Destinations...............................................................................71 Figure 3.13: Exporter Dynamics: Live Animals.............................................................................................................72 Figure 3.14: Dynamics of Live Animal Exporters in Main Destinations ......................................................................73 Figure 3.15: Exporter Dynamics: Meat.........................................................................................................................74 Figure 3.16: Dynamics of Meat Exporters in Main Destinations .................................................................................75 Figure 3.17: Exporter Dynamics: Pulses.......................................................................................................................77 Figure 3.18: Dynamics of Pulses Exporters in Main Destinations ................................................................................78 LIST OF TABLES Table 2.1: Overall Competitiveness Rankings............................................................................................................34 Table 2.2: Doing Business Indicators for Ethiopia.....................................................................................................35 Table 2.3: Time and Cost of Exporting and Importing ............................................................................................39 Table 2.4: Major Constraints Faced by Ethiopian Companies, Exporters vs. Non-Exporters.....................................40 Table 2.5: Effects of Undervalued RERs on Export and Output Growth Ethiopian data...........................................47 Table 3.1: Evolution of Significance of Exports.........................................................................................................58 Table 3.2: Prominent Marginal Exports....................................................................................................................59 Table 3.3: Panel Estimation Effect on Exports Growth of Undervaluation First Stage Regression..............................80 Table 3.4: Undervalued RERs and Export Growth....................................................................................................81 Table 3.5: Undervalued RERs and Output Growth...................................................................................................81 vii ACKNOWLEDGEMENTS T he World Bank greatly appreciates the structural transformation, and; Swarnim Waglé on close collaboration with the Government export performance and competitiveness. Mesfin Girma of Ethiopia (the Ministry of Finance and (Economist, AFTP2) prepared the macroeco- Economic Development, in particular) in the prepa- nomic analysis with support from Ashagrie Moges ration of this report. The report was prepared by a (Research Analyst) and Eyasu Tsehaye (Economist, team led by Michael Geiger (Sr. Country Economist, AFTP2). Gelila Woodeneh (Communications AFTP2), Lars Christian Moller (Lead Economist Officer, AFRSC) reviewed the document, provided and Sector Leader, AFTP2), and Swarnim Waglé (Sr. editorial content and designed the cover page. The Economist, PRMTR). The report benefitted from report was prepared under the overall guidance of background papers prepared by: (1) Ana Margarida Pablo Fajnzylber (Sector Manager, AFTP2) and Fernandes (Sr. Economist, DECTI) and Esteban Guang Zhe Chen (Country Director, AFCE3). The Ferro (Consultant, DECTI) on exporter level analysis; peer reviewers were: Paul Brenton (Trade Practice (2) Ha Nguyen (Economist, DECMG) on exchange Leader, AFTPM) and Tom Farole (Sr. Economist, rates; (3) Pedro Martins (Consultant, ATFP2) on AFTP2). ix LIST OF ABBREVIATIONS AFDB African Development Bank MENA Middle East and North Africa BRIC Brazil, Russia, India, and China MFN Most Favored Nation CPIA Country Policy and Institutional Mill. Million Assessment MOFED Ministry of Finance and Economic CSA Central Statistics Agency Development DB Doing Business NBE National Bank of Ethiopia EEPCO Ethiopian Electric Power NCB Non-concessional Borrowing Corporation NFS Non-Factor Services ERP Effective Rate of Protection NRP Nominal Rate of Protection EXPY Measure of Export Sophistication OZ Ounce FAO Food and Agriculture Organization PPP Purchasing Power Parity FDI Foreign Direct Investment PRODY Income Content of Product FY Fiscal Year PTA Preferential Trade Area GDP Gross Domestic Product Q1 1st Quarter GOE Government of Ethiopia RCA Revealed Comparative Advantage GTP Growth and Transformation Plan REER Real Effective Exchange Rate HACCP Hazard Analysis and Critical RER Real Exchange Rate Control Points ROW Rest of the World HDI Human Development Index SEZ Special Economic Zones HICES Household Income and SITC Standard International Trade Consumption Survey Classification HS Harmonized Commodity SOE State Owned Enterprise Description and Coding Systems SSA Sub-Saharan Africa IDA International Development TRIP Trade-Related Aspects of Association Intellectual Property IEMP Index of Export Market Penetration TTRI Tariff Trade Restrictiveness IMF International Monetary Fund WB World Bank ISO International Standards WDI World Development Indicators Organization WGI World Governance Indicators LFS Labor Force Survey WTO World Trade Organization LPI Logistics Performance Index y/y Year on Year M2 Monetary Aggregate 2 xi EXECUTIVE SUMMARY Rising exports contributed to Ethiopia’s remarkable growth performance over the past decade. Buoyed by favorable external conditions, exports also helped create jobs and earn much-needed foreign exchange. The way Ethiopia created and nurtured a high-value horticulture industry and expanded its air services exports was an encouraging example of “self-discovery.” A recent drop in export prices, however, has exposed underlying vulnerabilities in export structure and highlighted the importance of strengthening competitiveness. Ethiopia is vulnerable to such price swings because unprocessed and undifferentiated agricultural products dominate its exports. While benefitting from upward price trends since 2003, the recent drop in prices of key commodities led to the worst export performance in a decade. To overcome this challenge, renewed efforts must aim to improve competitiveness, including through value addition and export diversification. More than “what” is being exported it is the “how” that is hindering potential. There is scope for improving the quality of existing commodity exports, through basic value addition, such as coffee wet processing or machine flaying of animal skins. Even in products with a revealed comparative advantage, little upgrading or branding has occurred to earn higher value per unit over time. By starting to compete on the quality of existing commodity exports (and not just on price), Ethiopia can reduce sensitivity to volatile international prices thereby supporting the gradual shift of production and exports into agro-processing and light manufacturing. Ethiopia’s export sector is currently too small to contribute to structural transformation. In East Asia, booming exports helped shift economic activity and workers away from low-productivity agriculture into higher- productivity manufacturing and sustain high rates of economic growth for decades. In Ethiopia, both exports and the manufacturing sector remain relatively small. Ethiopia has the lowest ratio of merchandise exports to GDP among populous countries in the world; it has half as many of exporting firms as Kenya (which has half the population of Ethiopia), and average exporter size is small. The business environment favors incumbent firms and deters new entrants into export businesses, and even so, no “export superstars” are emerging. The export sector lacks dynamism in terms of firm entry and exit. Rather than, increasing in scale, new entrants to the export market are already often relatively well established in other businesses such as trading. This may be due to the fact that smaller firms have limited access to credit, which would have allowed them to increase the scale and scope of their activities. This is a challenge because rising and dynamic firms often create more new jobs than established firms. Other factors such as low entrepreneurship and low regulatory quality in terms of promoting the private sector may also explain this. Despite a favorable environment for incumbents, they are yet to emerge as multi-product and multi-destination export superstars. Ethiopia lags behind its peers in Global Competitiveness rankings and trade restrictions are biased against exports. Although the country ranks better than its peers on theme-specific business regulatory measures such as enforcing contracts, its overall Doing Business performance is on the decline. Low regulatory quality also adds to the cost of doing business for exporters. Furthermore, Ethiopia has more trade restrictions in place than its peers. However, high levels of nominal and effective rates of protection provide strong evidence of the widespread existence of an anti-export bias of the tariff regimes throughout the economy. At the same time Ethiopia is under- tapping a narrow window of opportunity for diversification through full exploitation of available trade preferences. xii 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS A more competitive real exchange rate could support export promotion. Ethiopia’s real exchange rate is overvalued. Empirical evidence presented in the report suggests that a 10 percent lower real exchange rate could increase export growth in Ethiopia by more than 5 percentage points per year and increase economic growth by more than 2 percentage points. The projected positive impact for Ethiopia is based on the predominance of basic export commodities that tend to compete more on price rather than on quality. In the presence of macroeconomic trade-offs (e.g. currency depreciation would contribute to inflation), changes in exchange rate policy would need to be combined with other economic policy measures, such as a further tightening of monetary policy. Policy Recommendations The following policy recommendations emerge from the analysis: 1. Increase value-addition, quality, and branding of exports. 2. Ease binding constraints related to reliable power supply, credit, and foreign exchange. 3. Redress bottlenecks in trade logistics. 4. Establish Industrial Zones that conform to international best practice. 5. Revise burdensome business rules that obstruct firm entry, especially high start-up capital requirement and pre- registration bank deposits. 6. Improve regulatory quality, including the implementation of a pro-competition legal framework. 7. Ensure that the real exchange rate is competitive. This report is structured as follows. Chapter 1 discusses recent economic developments in the ‘short view’ and has observations on structural change in the ‘long view’. Chapter 2 looks at export performance and competitiveness utilizing key elements of the World Bank’s Trade Competitiveness Diagnostic Framework. Chapter 3 summarizes key issues and provides policy recommendations. 1 RECENT ECONOMIC DEVELOPMENTS 1 The Short View of Ethiopian Airlines and Ethio Telecom. Passenger traffic rose by 13 percent while cargo services declined Recent high economic growth is driven by services and by 4 percent (Figure 1.1.5). Similarly, Ethio Telecom agriculture on the supply side and private consumption and public investment on the demand side. Inflation remains in (unaudited) profits after tax increased from 3.0 bil- single digit territory due to tighter monetary policy and lower lion birr in 2011/12 to 4.4 billion birr in 2012/13 global commodity prices. The fiscal stance continues to be (Figure 1.1.6), and results from the first half of the expansionary in light of substantial state owned enterprise current fiscal year shows a continuation of this trend. investment. Exports are exhibiting their worst performance in more than a decade. On the other hand, electricity generation growth slowed from 27 to 21 percent and the growth rate of power sales to industries fell from 29 to 15 percent Real Sector (Figure 1.1.4). Private consumption and investment were the Ethiopia continues to register rapid economic major growth contributors on the demand side. growth driven by services and agriculture. GDP Owing to its relative size in GDP (75 percent), the growth increased by 9.7 percent in 2012/13 com- 6.4 percent annual growth rate in private consumption pared to 8.8 percent in 2011/12. Industry grew by accounts for 5.0 percentage points of the 10.4 percent 18.5 percent followed by services (9.9 percent) and (including net indirect taxes) growth rate in 2012/131 agriculture (7.1 percent). However, given the rela- (Figure 1.1.2). Investment accounts for 3.3 percentage tive size of each sector, expansion of the services points of overall GDP growth. and agriculture sector explain most of GDP growth Two opposing forces affect the growth out- (4.5 and 3.1 percentage points, respectively), while look for 2013/14: a bumper harvest and an export the contribution of industry was relatively modest contraction. Crop production is forecast to grow by (2.1 percentage points). Manufacturing, which forms 9.9 percent in 2013/14 compared to 5.8 percent in part of the industry sector, added just 0.4 percent- 2012/13 and this may add one percentage point to age points to the overall growth rate of 9.7 percent growth. Declining exports, on the other hand, rep- (Figure 1.1.1). resent a drag on economic activity. Total exports (of Positive developments in some of the major goods and services) declined by 0.6 percent in 2012/13 sub-sectors contributed to growth. Within agricul- and by 2.0 percent in the first quarter of 2013/14. If ture, crop production (accounting for 30 percent of the first quarter performance is indicative in 2013/14, GDP) was the major contributor. Crop value added then this would reduce GDP growth by about ¼ of increased from 5.0 percent in 2011/12 to 8.2 percent a percentage point. in 2012/13. Construction activity was the major driver of the non-manufacturing industry sector, with a growth contribution of 1.4 percentage points in 2012/13. Within services, transport and commu- 1 The real GDP from the demand side is computed using the GDP nications was the leading sector driven by activities deflator. 2 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 1.1: Economic Activity 1. Real GDP Growth (Supply Side) 2. Real GDP Growth (Demand Side) 14 20 12 15 10 10 8 5 6 0 4 2 –5 0 –10 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Agriculture Manufacturing Private consumption (C) Public consumption (G) Non-manuf. industry Services GDP growth Investment (I) Net exports (E-M) GDP growth (Y) 3. Crop Production Growth (%) 4. Electricity Generation and Sales Growth (%) Ethiopia: Growth in major season crop production 30% 20% 25% 15% 12.6% 9.9% 10% 7.3% 6.7% 7.4% 5.8% 20% 5.6% 5% 0% 15% –5% –10% 10% –15% 5% –20% –25% 0% 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/ 14Forc 2011 2012 2013 Commercial & street light Industrial sales Cereal Pulses Oil seeds Grains Household sales Electricity production 5. Ethiopian Airlines Activity Growth 6. Ethio Telecom: Profitability 40 14 90 35 12 80 30 70 10 25 60 8 Billin birr 20 50 15 6 40 10 30 4 5 20 0 2 10 –5 0 0 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2010/11 2011/12 2012/13 Profit after tax EBIDTA Cargo Passengers EBIDTA margin (right axis) Source: 1.1: MOFED, 1.2: MOFED, 1.3: CSA, 1.4: EEPCO, 1.5: Ethiopian Airlines, 1.6: Ethio Telecom. Note: EBIDTA is Earnings Before Interest, Depreciation and Taxes. Recent Economic Developments 3 Monetary Sector sector credit continues to be the main driver while credit to the private sector remained low. Domestic Inflation has remained in single digits for almost a credit to the public sector increased by 34 percent year. After peaking at 40.7 percent in August 2011, (y/y) in November 2013. Although the share of the headline inflation fell to a three-year low of 6.1 percent public sector in total outstanding credit increased only in April 2013 and stood at 8.8 percent in March 2014 modestly in 2012/13 (62 percent), the composition (Figure 1.2.1). Both food and non-food prices have has shifted substantially towards public enterprises in contributed to this decline. Lower international food recent years (Figure 1.2.6). prices have been helpful and it is encouraging that Maintaining single-digit inflation would non-food inflation (a proxy for core inflation) hov- require continued monetary discipline and sup- ers around single digit territory (11.8 percent). Food port from fiscal policy. The monetary authorities price inflation fell to 6.1 percent, in part due to a good would need to maintain low levels of reserve money harvest during the main agricultural season. and broad money growth. However, as discussed International factors contributed to reduced next, the fiscal policy also needs to be aligned with inflationary pressure. A decomposition of inflation the monetary policy objective to maintain low infla- into tradable and non-tradable goods reveals that tion. In the absence of this alignment, Ethiopia would internationally traded goods (imported and exported remain additionally vulnerable to renewed spikes in commodities) in Addis Ababa exhibit a much faster international food prices. decline in inflation than goods which are not inter- nationally traded (Figure 1.2.2).2 Edible oil, coffee Fiscal Sector beans, benzene, and chickpeas are examples of trad- able goods. In parallel, the FAO food price index3 has The fiscal policy stance was expansionary in been declining since August 2012. 2012/13. The estimated consolidated public sec- A tightening of monetary policy has also con- tor primary deficit reached 5.2 percent of GDP in tributed to lower inflation. Reserve money growth 2012/13 compared to 3.7 percent of GDP in 2011/12 (the nominal anchor) dropped from a peak of about (Figure 1.3.1).5 The deficit is the result of a growth 40 percent in July 2011 to –3.7 percent in December strategy that relies on substantial public investment 2013, though changes in the definition of this mea- executed through State Owned Enterprises and the sure also explain the decline.4 Broad money growth federal and regional governments. Fiscal policy sup- which remains high, but also experienced a reduction ports short term economic activity, but makes it chal- from 37 percent to 22 percent over this period may lenging to keep inflation in check. therefore be a better indicator of the current mon- etary policy stance. (Figure 1.2.3). Lower inflation, in turn, has contributed to lower real interest rates. The maximum lending rate has been positive in real 2 Owing to considerable data processing requirements, this point is il- lustrated using data for Addis Ababa only, though there is a very high terms since December 2012 while the real minimum correlation between Addis Ababa and other cities on these variables. deposit rate is –2.7 percent (Figure 1.2.4). A higher 3 The FAO Index includes international prices of a basket of meat, diary, cereals, vegetable oil, and sugar. real interest rate would help increase Ethiopia’s very 4 The reserve requirement of banks was reduced from 10 to 5 percent in low savings rate as discussed in World Bank (2013). March 2013, but banks instead had to purchase certificates of deposits. The measure thus had a relatively neutral impact on monetary policy Credit growth, mainly to State Owned (in the short term), but it contributed to a decline in growth of reserve Enterprises, is the major contributor to broad money (currency in circulation and bank deposits at the central bank). 5 An accurate overall fiscal stance for the consolidated public sector money growth. Net domestic credit growth reached (including public enterprises) is difficult to gauge and this estimate was 31 percent in December 2013 (Figure 1.2.5). Public generated from the financing side (IMF, 2013). 4 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 1.2: Monetary Sector 1. Inflation (y/y, %) 2. Inflation, Addis Ababa (y,y, %) 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% –10% 0% Jan-10 May-10 Sep-10 Jan-11 May-11 Sep-11 Jan-12 May-12 Sep-12 Jan-13 May-13 Sep-13 Jan-14 Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-12 Jan-13 Apr-13 Jul-13 Oct-13 Food Non-food Headline Traded Non-traded General 3. Broad Money (M2) and Inflation (%, y/y) 4. Real Interest Rates (%) 45% 30 40% 20 10 35% 0 30% –10 25% –20 20% –30 –40 15% –50 10% –60 Sep-07 2007/08 Mar-08 Dec07 Jun-08 Sep-08 2008/09 Dec08 Mar-09 Jun-09 Sep-09 2009/10 Dec09 Mar-10 Jun-10 Sep-10 2010/11 Dec10 Mar-11 Jun-11 Sep-11 2011/12 Dec11 Mar-12 Jun-12 Sep-12 2012/13 Dec12 Mar-13 Jun-13 Dec13 2013/14 Sep-13 5% 0% Jan-10 Mar-10 May-10 Jul-10 Sep-10 Nov-10 Jan-11 Mar-11 May-11 Jul-11 Sep-11 Nov-11 Jan-12 Mar-12 May-12 Jul-12 Sep-12 Nov-12 Jan-13 Mar-13 May-13 Jul-13 Sep-13 Nov-13 Broad money (M2) Inflation (y/y) Real minimum deposit rate Real maximum lending rate 5. Broad Money Growth (M2, y/y, %) 6. Composition of Domestic Credit Stock (%) 50% 100% 40% 90% 30% 80% 70% 20% 60% 10% 50% 0% 40% –10% 30% –20% 20% Sep-11 Oct-11 Nov-11 Dec-11 Jan-12 Feb-12 Mar-12 Apr-12 May-12 Jun-12 Jul-12 Aug-12 Sep-12 Oct-12 Nov-12 Dec-12 Jan-13 Feb-13 Mar-13 Apr-13 May-13 Jun-13 Jul-13 Aug-13 Sep-13 Oct-13 Nov-13 Dec-13 10% 0% 2008 2009 2010 2011 2012 2013 Oct-13 Net domestic credit Net foreign assets Other items (net) Broad money (M2) SOEs CG credit Private credit Source: 2.1: CSA, 2.2: staff estimate based on CSA data, 2.3 and 2.4: NBE and CSA, and 2.5–2.6: NBE. Recent Economic Developments 5 The general government fiscal deficit remains Ethiopia’s risk of external debt distress remains modest.6 The overall fiscal deficit increased from 1.2 to low despite substantial non-concessional borrow- 2.0 percent of GDP between 2011/12 and 2012/13. ing commitments in 2012/13. According to the The deficit was financed primarily from external 2013 Joint Bank-Fund Debt Sustainability Analysis, sources (2.0 percent of GDP), with domestic financing Ethiopia’s risk of external debt distress is low. However, contributing 0.2 percent of GDP.7 The government substantial contracting of non-concessional loans also financed part of the deficit through direct central of around US$5.5 billion (11.7 percent of GDP bank advances (Figure 1.3.2). or 50 percent of total external debt) in 2012/13 Government revenues improved markedly in might pose risk to long term debt dynamics (see 2012/13. This development is encouraging given Figure 1.3.6). It is imperative that the authorities Ethiopia’s low tax revenue-to-GDP ratio. Total proceed with fiscal prudence and take measures to revenues increased from 13.9 percent of GDP in improve export competitiveness in order to maintain 2011/12 to 14.6 percent of GDP in 2012/13. a low debt distress risk rating. Additional revenues were collected from indirect taxes (0.6 percent of GDP) and direct taxes (0.4 percent of External Sector GDP) (Figure 1.3.3). Non-tax revenues, on the other hand, declined by 0.3 percent of GDP as a result of Ethiopia has a chronic external current account def- relatively low dividend receipts from public enterprises icit owing to an unusually large trade and services and the National Bank. Finally, official development deficit. In 2012/13, the former reached 5.9 percent assistance (grants) declined from 1.7 to 1.5 percent of GDP (after transfers), while the latter amounted of GDP over this period. to 16.7 percent of GDP (Figure 1.4.1). External Expenditures increased more rapidly, leading borrowing (4.2 percent of GDP) and FDI (2.6 per- to a larger general government fiscal deficit. Total cent of GDP), in turn, financed the external current expenditures increased from 16.8 percent of GDP account deficit for 2012/13. The substantial reliance in 2011/12 to 18.1 percent of GDP in 2012/13. on private and public transfers (8.3 and 2.7 percent of This was due to a substantial rise in capital spend- GDP, respectively) represents a potential vulnerability. ing (0.8 percent of GDP) and a modest increase in Foreign exchange reserves are low at about 1.9 months recurrent spending (0.4 percent of GDP). Capital of imports in December 2013(Figure 1.4.2). spending increases were concentrated in agriculture, The external current account deficit improved roads, health and housing. Government spending in 2012/13 due to a declining trade and services remains tilted in favor of capital spending (10.7 per- deficit. The trade deficit improved by 0.5 percent cent of GDP) versus recurrent spending (7.4 percent of GDP as the fall in goods exports was offset by an of GDP) (Figure 1.3.4). even larger decline drop in goods imports. Similarly, An increase in the general government fis- the services balance improved by 0.8 percent of GDP cal deficit is projected for 2013/14. The approved because a drop in services exports was offset by an even general government budget (excluding SOEs) envis- larger decline in services imports. On the other hand, aged an increase in the deficit to 3.0 percent of GDP. official transfers dropped from 4.2 to 2.7 percent of During the first half of 2013/14, the general gov- GDP, counteracting somewhat the improved goods ernment (excluding SOEs) registered a surplus of and services balance. 0.9 percent of GDP against a surplus of 1.3 percent the same period a year before. A relative slowdown 6 General government includes federal and regional governments, but excludes SOEs. in surplus was the result of faster increased spending 7 Errors and omission and privatization amounted to –0.4 and 0.1 percent vis-à-vis revenue collections (Figure 1.3.5). of GDP, respectively. 6 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 1.3: Fiscal Sector 1. Primary Deficit (% of GDP) 2. Direct NBE Advances to GoE (% of GDP) General gov't and public sector primary deficit/GDP Direct advance/GDP 0 1.4 –1 –1.0 1.2 –2 –1.9 1.0 –3 –2.5 –4 –3.7 0.8 –5 0.6 –5.2 1.1 –6 1.1 0.4 –7 –6.7 0.6 –8 0.2 0.3 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 – Public sector primary deficit/GDP GG primary deficit 2009/10 2010/11 2011/12 2012/13 3. General Government Revenue (% 0f GDP) 4. General Government Spending (% 0f GDP) 16 14.6 20 14.2 13.7 13.9 14 18 12.1 12.1 2.0 2.8 2.0 2.3 16 12 9.8 9.2 10.4 10.5 9.9 10.7 2.4 14 10 3.4 4.5 4.7 4.7 4.5 12 8 10 4.8 3.6 6 8 2.8 3.1 3.2 3.8 6 4 2.1 2.2 4 9.3 8.2 8.5 8.0 7.0 7.4 2 2.9 3.0 3.9 3.9 3.9 4.3 2 0 0 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 Direct tax Domestic indirect tax Foreign trade tax Recurrent spending Capital spending Non tax revenue Total revenue 5. FY14 Budget Outturn (6 months, % of GDP) 6. Multilateral, Bilateral and Non-concessional Commitments (% of GDP) 9 8 18 7 16 14 6 12 5 10 4 8 3 6 2 4 1 2 0 0 Revenue & grant Expenditure Overall balance 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 6 mons FY13 6 mons FY14 Multilateral Bilateral Commercial (NCB) Source: 3.1: MOFED and WB/IMF, and 3.2-3.5: MOFED. 3.6 Staff Estimate based on MOFED data. Recent Economic Developments 7 FIGURE 1.4: External Sector 1. External current account balance (% of GDP) 2. Gross Official Foreign Exchange Reserves 25 4000 3.5 20 1.6 3.5 4.1 2.3 4.9 2.3 0.3 3500 3.3 4.0 4.2 15 3.1 2.8 2.4 3.0 4.9 6.5 5.9 2.5 2.6 3000 10 4.9 4.2 2.7 5 2500 2.5 9.0 8.5 9.2 8.8 7.6 8.3 0 –2.0 –0.7 2000 –1.4 –3.4 –6.5 –5.9 –5 1500 2.0 –5.1 –4.1 –5.7 –10 1000 –15.1 –16.7 1.5 –15 –18.4 500 –20 –18.1 –19.6 –21.6 0 1.0 –25 Jan-09 Mar-09 May-09 Jul-09 Sep-09 Nov-09 Jan-10 Ma-10 May-10 Jul-10 Sep-10 Nov-10 Jan-11 Ma-11 May-11 Jul-11 Sep-11 Nov-11 Jan-12 Ma-12 May-12 Jul-12 Sep-12 Nov-12 Jan-13 Ma-13 May-13 Jul-13 Sep-13 Nov-13 2008 2009 2010 2011 2012 2013 Change in reserves Loans FDI CAB/GDP Official transfers Private transfers Trade and services balance Foreign reserve (mill $) Reserve in months of import (right axis) 3. Goods Export Growth: Value, Price and Volumes 4. Services Export Growth (%) 50 40 40 35 30 30 20 25 10 20 0 15 –10 –20 10 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 5 0 Value growth Volume gowth Price growth 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 5. Composition of Goods Import (% of GDP) 6. Real Effective Exchange Rate 30 180 160 25 5.3 160 7.7 4.9 5.1 5.7 141 4.5 4.7 140 20 –42% 5.3 5.0 58% 6.5 4.0 4.5 120 51% 15 9.9 100 7.8 8.8 7.0 10 7.3 7.4 80 94 5 7.5 8.6 7.3 8.3 60 6.2 7.2 0 40 2008 2009 2010 2011 2012 2013 May-05 Sep-05 Jan-06 May-06 Sep-06 Jan-07 May-07 Sep-07 Jan-08 May-08 Sep-08 Jan-09 May-09 Sep-09 Jan-10 May-10 Sep-10 Jan-11 May-11 Sep-11 Jan-12 May-12 Sep-12 Jan-13 May-13 Sep-13 Jan-14 Consumer goods Capital goods Fuel Others Source: NBE and MOFED, except 1.4: WB. 8 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 1.5: External sector (continued) 1. Value indices of selected commodities 2. Value indices of selected commodities 1500 1500 3500 1300 1300 2800 1100 1100 900 900 2100 700 700 1400 500 500 700 300 300 100 100 0 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 Coffee Oil seeds Pulses Chat Gold (right axis) Live animals Flower (right axis) 3. Price indices of selected commodities 4. Price indices of selected commodities 400 700 100 200 350 90 190 600 80 180 300 500 70 170 250 60 160 400 200 50 150 300 40 140 150 200 30 130 100 20 120 50 100 10 110 0 0 0 100 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 Coffee Oil seeds Pulses Chat (right axis) Gold (right axis) Live animals Flower (right axis) 5. Volume indices of selected commodities 6. Volume indices of selected commodities 850 290 9000 2100 750 7500 240 1700 650 550 6000 190 1300 450 4500 350 140 900 3000 250 90 1500 500 150 50 40 0 100 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 Coffee Oil seeds Pulses Chat (right axis) Gold (right axis) Live animals Flower (right axis) Source: NBE and IMF. Recent Economic Developments 9 Goods exports are exhibiting their worst per- The real effective exchange rate has appreciated formance in a decade. Export of goods declined by by more than 50 percent over the past 3.5 years 2.5 percent in 2012/13 and by a further 7.6 per- (Figure 1.4.5). As discussed in Chapter 2, main- cent in the first half of 2013/14, exceeding the drop taining a competitive exchange rate is an important observed in 2008/09 during the global financial crisis component of maintaining external competitiveness, (Figure 1.4.3). Over the previous three years, export especially since Ethiopia’s export basket consists of growth had averaged 30 percent per year (and about primary products that compete more on price than 20 percent annually since 2004). quality. Declining international prices are the main culprit for the recent poor performance of goods The Long View: Structural Change9 exports. A decomposition of export value growth into its price and quantity effects illustrates this point. The analysis reveals that the structure of Ethiopia’s output Export prices dropped by a staggering 15 percent has changed considerably—predominantly from agriculture to services—while similar changes in the composition of in 2012/13—much more than what was observed employment have lagged behind. Labor productivity growth in 2008/09. The price drop was so severe that even has been strong across most sectors, albeit mainly driven a healthy export volume growth of 15 percent in by productivity improvements within each economic sector rather than through labor reallocation. Nonetheless, the 2012/13 could not prevent export values from declin- pace of structural change is accelerating and its relative ing. Declining international prices for coffee and gold, contribution to output growth is increasing. The services which account for close to half of total goods exports, sector has been the major driver of structural change over past 15 years, while manufacturing remains small. were particularly pronounced (see product analysis in Figures 1.5.1–1.5.6). Services exports, which are similar in size to Structural change is vital for sustaining eco- goods exports, also exhibited poor performance nomic and social development. In simple terms, in 2012/13. Exports of services increased by only structural change can be defined as the reallocation of 1.5 percent compared to an average annual growth rate labor from low-productivity sectors to more dynamic of 18 percent in the previous nine years (Figure 1.4.4). (higher-productivity) economic activities.10 For most Transportation services (Ethiopian Airlines), which developing countries, this would usually require shift- account for two thirds of services exports, increased ing labor from subsistence agriculture to commercial by 12 percent in 2012/13 compared to 28 percent agriculture, manufacturing, and modern services. “The the previous year. On the other hand, there was a speed with which this structural change takes place is 25 percent decline in travel (tourism).8 the key factor that differentiates successful countries Goods and services imports moderated in from unsuccessful ones” (McMillan and Rodrik, 2011). 2012/13. Imports of goods increased by 3.7 percent in Enhancing the tradable sector and promoting 2012/13 compared to 34 percent growth the previous exports is a viable option to facilitate structural year. Figure 1.4.5 illustrates the composition of goods imports as a share of GDP. It emerges that consumer goods and fuel declined in 2012/13, while capital 8 Data weaknesses may explain part of this decline 9 This section is based on a background paper prepared by Martins goods imports increased. There was also a substantial (2014). In addition to this Economic Update there will be an upcoming decline in “other imports.” On the services import side, Growth Study currently being prepared by the World Bank team that will look in-depth into the issue of structural change in the Ethiopian a decline of 14 percent was observed in 2012/13 com- economy. pared to a growth rate of 44 percent the previous year. 10 Structural change can also refer to the changing composition of output. However, since shifts in production tend to precede shifts in employment, Travel services imports, i.e. payments to other airlines it can be argued that this transformative process is only under way once that provided services to Ethiopia dropped 33 percent. labor starts to relocate. 10 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS change through productivity increases. Lewis employment doubled from 0.5 million in 1996 to (1954) first described the mechanisms behind the 1.0 million in 2011, it remains relatively small. The link between exports and productivity increases: as the services sector accounted for about a quarter of total export sector (mostly manufacturing exports) expands, employment growth (2.9 million people) over the it attracts labor from the agricultural sector in rural 15-year period of analysis (Figure 1.6.4). areas. Since labor moves from agriculture—a relatively less productive sector—to manufacturing—a more Labor Productivity12 productive sector, the economy’s aggregate productiv- ity rises, and so does output. Improvements in living standards in Ethiopia are largely explained by rising labor productivity. Output and Employment Figure 1.7.1 presents a decomposition of output (value added) per person, which can improve either as a result Ethiopia’s output more than tripled in real terms of demography (if the relative share of the working over the past fifteen years—driven primarily by age population rises), increases in the employment services and agriculture. Real gross value added rate (if relatively more people are working) and rising increased from 160.2 billion birr in 1996 to 510 bil- labor productivity (if each worker produces more). lion birr in 2011, as illustrated in Figure 1.6.1. Services Between 1996 and 2011, real value added per person (“trade,” “transport and communications,” and “other in Ethiopia doubled from 3,040 to 6,174 thousand services”11) contributed to half of output growth since birr. More than 90 percent of this increase was due 1996, while agriculture contributed by one third to higher output per worker (labor productivity). (Figure 1.6.3). The contribution of manufacturing to The increase in the employment rate from 74.4 per- economic growth was 4 percentage points. cent to 79.2 percent also played a role, explaining The structure of output shifted considerably about 10 percent of per capita growth. The impact of from agriculture towards services while the corre- demographic change, meanwhile, was negligible. The sponding employment shift was modest. The output results confirm the relevance of a detailed analysis of share of agriculture declined from 62 percent in 1996 labor productivity as key to understanding why living to 45 percent in 2011 (Figure 1.6.5). Services output, standards have improved in Ethiopia in recent years. meanwhile, increased from 30 to 44 percent of the Labor productivity growth in Ethiopia has total over this period, while industry increased from been rapid, especially since 2005. Between 1996 8 to 11 percent. Agriculture continued to dominate and 2011, value added per employed worker increased employment, however, as illustrated in Figure 1.6.6. by 4.5 percent per year. This average, however, masks Its employment share fell only marginally from 81 to considerable differences depending on the period of 78 percent in 1996–2011. Services jobs, meanwhile, analysis, as illustrated in Figure 1.7.2. Labor pro- increased from 14 to 17 percent of the total with indus- ductivity growth was particularly rapid in 2005–11, try staying constant at 5 percent. Within the industry reaching 9.3 percent per year. category, manufacturing employment rose marginally from 2.3 to 3.0 percent between 1996 and 2011. Total employment increased by 11.6 million people since 1996 reaching 34.2 million in 2011 11 The category “other services” consists of the following sub-categories: “real estate, renting, and business activities,,” “public administration compared to 22.6 million in 1996 (Figure 1.6.2). and defense,,” “health and social work,,” “other community, social and Agriculture accounted for the majority of new jobs personal services,,” “education,,” and “private households with employed persons..” created with almost three-quarters of employment 12 Labor productivity estimates may be imprecise owing to national ac- growth (8.4 million people). Although manufacturing counts data limitations mentioned in Footnote 1. Recent Economic Developments 11 FIGURE 1.6: Ethiopia: Output and Employment by Sector, 1996–2011 1. Value added (billion birr, 2010/11 prices) 2. Employment (million people) 500 35 450 30 400 350 25 300 20 250 200 15 150 100 10 50 5 0 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 1996 1999 2005 2011 3. Value Added Growth Contribution, 1996–2012 (%) 4. Employment Growth Contribution, 1996–2012 (%) 5. Value Added Shares (%) 6. Employment Shares (%) 100 100% 90 90% 80 80% 70 70% 60 60% 50 50% 40 40% 30 30% 20 20% 10 10% 0 0% 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 1996 1999 2005 2011 Agriculture Mining and quarrying Construction Manufacturing Trade Electricity and water Transport and communication Other services Source: National Accounts Directorate, MoFED. Central Statistical Agency (CSA): HICES 1995/96, LFS 1999, LFS 2005, HICES 2010/11. See Martins (2014) for details. 12 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS Productivity gains in trade and “other services” transformation—manufacturing—is vastly under- were several times faster than in agriculture and developed. Not only are developing countries not manufacturing. Labor productivity increased by industrializing fast enough, many have begun to de- 6.1 percent per year for “other services” and by 5.4 per- industrialize before attaining higher shares of employ- cent per year for trade in 1995–2011 (Figure 1.7.3). ment in manufacturing (Rodrik 2013a). The trodden By comparison, agriculture and manufacturing labor path of economic development, first taken by Western productivity growth ranged at around 2.5 percent per countries and replicated in recent decades by East Asia, year while construction productivity grew 4.0 percent is one where farmers move into higher-productivity per year. The mining sector experienced double-digit manufacturing or agro-processing; economies diversify labor productivity growth, though the employment and begin to export more sophisticated goods. The share of mining is very small. share of the labor force employed in manufacturing Labor productivity levels are highest in sectors peaked at 25 to 45 percent in countries like the UK, such as electricity/water, mining, and transport/ U.S. and Sweden before these countries de-industri- communications, and is lowest in agriculture. alized. Even Korea, where the manufacturing employ- Output per worker reaches up to 108.9 thousand ment share was in the single-digit range in the 1950s, birr (2010/11 prices) in the electricity/water sector. In peaked at nearly 30 percent before decreasing in the agriculture, at the other extreme, the value added per 1980s. In Africa, fewer than 10 percent of workers find worker is only 8.0 thousand (Figure 1.7.4). It is useful, jobs in manufacturing; in Ethiopia only 3 percent do. however, to put these figures into perspective, since Manufacturing remains a desirable path their ultimate impact on the economy greatly depends towards modernization because it absorbs a large on the relative weight of each sector. Figure 1.7.5 com- section of the labor force on higher productivity bines information on labor productivity and sectoral tasks. Further, labor productivity in manufacturing shares of employment for two points in time (1996 has been shown to converge to the global frontier: and 2011). The figure illustrates the productivity lev- manufacturing industries that are less productive tend els and growth of the three major economic sectors: to grow even faster (Rodrik 2013b). However, there agriculture, trade, and other services. is a caveat. In today’s era of fragmented production, Labor productivity growth was accompanied manufacturing has become more capital intensive by employment growth. Figure 1.7.6 shows that all and disconnected from domestic economies. In such sectors enjoyed the attractive combination of positive cases, it has been pointed out that certain services sec- labor productivity growth and employment growth. tors—such as food and clothing retail services—look Labor productivity growth has a dual character: antiquated, because they absorb technologies and while it is an important source of economic growth employ semi-skilled workers in large numbers. While and dynamism, it can also have a labor-saving effect, Ethiopia may not replicate East Asia’s industrializa- potentially reducing employment demand. Hence, tion model ( which occurred in the second half of productivity growth needs to be accompanied by the twentieth century in a less globalized world with employment growth, which was indeed the case for minimal tradability of services) Ethiopia’s path forward Ethiopia in 1996–2011. could be one where its surplus labor is absorbed in manufacturing tied much more closely with emerging Structural Change services that can potentialy absorb workers at higher productivity levels than in the primary sector. An emerging policy concern in Sub-Saharan Labor productivity growth in Ethiopia is mostly Africa is that the continent is growing rapidly explained by improvements within each economic but transforming slowly. The sector that speeds up sector as opposed to structural change. Figure 1.8.1 Recent Economic Developments 13 FIGURE 1.7: Ethiopia: Labor Productivity, 1996–2011 1. Decomposing Output per Person (1996–2011) 2. Labor Productivity Growth 10 9.3 Demographic change 8 6 Employment rate 4 3.0 2 0 Output per worker 2 –1.7 4 –500 0 500 1000 1500 2000 2500 3000 1996–99 1999–05 2005–11 3. Labor Productivity Growth, 1996–2011 (%) 4. Labor Productivity, 1996 & 2011 (1,000 birr) Agriculture 2.6 Agriculture Mining & quarrying 10.4 Mining & quarrying Manufacturing Manufacturing 2.4 Electricity & water Electricity & water 6.8 Construction Construction 4.0 Trade Trade 5.4 Transport & comms Transport & comms 3.3 Other services Other services 6.1 Total Total 4.5 0 20 40 60 80 100 120 0 2 4 6 8 10 12 1995/96 2010/11 5. Labor Productivity Level and Labor Shares 6. Employment Growth, 1996–2011 120 Agriculture 2.6 100 Mining & quarrying 0.4 Labor Productivity (1,000 birr) 80 Manufacturing 4.6 60 Electricity & water 2.1 40 Construction 7.3 20 Trade 4.2 0 Transport & comms 6.7 1996 2011 Other services 4.2 Agriculture Mining & quarrying Construction Total 2.8 Manufacturing Trade Transport & comms Electricity & water Other services 0 1 2 3 4 5 6 7 8 Source: National Accounts Directorate, MoFED. Central Statistical Agency (CSA): HICES 1995/96, LFS 1999, LFS 2005, HICES 2010/11. WB/WDI. See Martins (2014) for details. 14 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS divides the labor productivity effect into two separate improvements in the sector are indispensable for components: “within” and “between” sector contribu- Ethiopia’s prospects for continued improvements in tions. The “within sector” effect refers to improved total labor productivity, and living standards. output per person in sectors such as agriculture or Ethiopia faces considerable challenges in terms manufacturing, i.e. that workers within each sector of achieving development through manufacturing have become more productive. The “between sectors” or “industrialization.” First, the sector remains rela- effect is what has been previously described as struc- tively small both in terms of output and employment. tural change. It reflects the move of workers from low Second, the manufacturing share of output remained productivity sectors (e.g. agriculture) to high productiv- constant over the past 15 years, while the employment ity sectors (e.g. services). The results show that about share declined. Third, the sector has the second lowest 80 percent of the improvements in output per person labor productivity level amongst major sectors, only over the past 15 years were due to “within sector” pro- twice as high as agriculture. Fourth, manufacturing ductivity gains. Structural change, or “between sector” labor productivity growth was the lowest among all productivity gains, explain 11 percent, while a higher sectors since 1995. Fifth, the GDP per capita share of employment rate account for the remaining 10 percent. manufacturing and industry is among the ten lowest Ethiopia’s degree of structural change com- in Sub-Saharan Africa, as illustrated in Figure 1.8.2 pares well in an international perspective. A study (the size of the bubbles reflects GDP per capita). by McMillan and Rodrik (2011) concludes that Finally, the contribution of manufacturing to struc- the Africa region is experiencing negative structural tural change was quite small (Figure 1.8.6). change implying a shift of workers from high pro- On the other hand, the services sector has dem- ductivity to low productivity sectors.13 This process onstrated considerable potential for Ethiopia over obviously undermines sustained economic growth the past 15 years, including through its contribu- and economic development. tion to structural change. The services sector is the Encouragingly, this study ranks Ethiopia as the largest in terms of economic output and is the second second highest performer in the sample of 38 devel- largest employer. It accounts for most of the structural oped and developing countries in terms of positive shifts in output and labor away from agriculture since structural change for the 1990–2005 period (at 1995. Levels of labor productivity are relatively high 11.2 percent), although this may be at the higher end and labor productivity growth has been substantial. owing to data issues.14 When compared to Korea’s This implies that the structural change has been golden period (1970–90), it is noted that Ethiopia’s dynamic in Ethiopia. Finally, the services sector has pace of structural change is much slower. The con- been the major driver of structural change in Ethiopia tribution of structural change to output per capita over the past 15 years (Figures 1.8.5 and 1.8.6). observed in Korea was 28.3 percent (Figures 1.8.3 In sum, Ethiopia needs to move forward across and 1.8.4). all sectors in terms of boosting labor produc- Given the large size of the agriculture sector, tivity. Agriculture productivity improvements are it is imperative that continued efforts are made to make the sector more productive. The agriculture 13 However, a recent revision of the estimates for sub-Saharan Africa sector is, by far, the biggest employer in Ethiopia and does suggests a more positive picture for the period between 2000 and the second largest in terms of output. The sector also 2005—see AfDB et al (2013). 14 McMillan and Rodrik (2011) use the 1994 Census data for Ethiopia, accounted for most of the net employment growth which estimates the share of agricultural employment at 90 percent over the period of analysis. Although some labor compared to only 81 percent in HICES 1995/96 and 80 percent in the 1999 Labor Force Survey. Another regional study by de Vries et al. (2013) shifted out of agriculture, substantial shifts are likely to does not use the latest available data for Ethiopia. Readers are referred take a long time. As a result, further labor productivity to Martins (2014) instead. Recent Economic Developments 15 FIGURE 1.8: Ethiopia: Labor Productivity and Structural Change, 1996–2011 1. Sources of Output Growth per Person (1996–2011) 2. Industry and Manufacturing in SSA 45 40 Manufacturing (percent of GDP) 35 30 25 20 15 ETH 10 Between sector productivity gains 5 Within sector productivity gains 0 0 20 40 60 80 100 Changes in employment Industry (percent of GDP) 3. Labor Productivity Growth Decomposition, Ethiopia 1996–2011 4. Labor Productivity Growth Decomposition, Korea 1970–1990 Inter-sectoral shift Inter-sectoral shift L+M+N+O -Public administration; Education; health an social work; other community, social and personal services Other Services J+K - Financial intermediation; Real estate, renting and business activities I - Transport, storage and Transport and Comms communications G +H - Wholesale and retail trade; repair of motor vehicles, Trade motorcycles and personal and household goods; Hotels and Restaurants F - Construction Construction E - Electricity, gas and water supply Electricity and Water D - Manufacturing Manufacturing C - Mining and quarrying Mining and Quarrying A+B - Agriculture, hunting and forestry, and fishing Agriculture 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 0 500 1,000 1,500 2,000 2,500 2005 Korean Won 5. Contribution to Structural Change 6. Contribution to Structural Change (Inter-Sectoral Labor Productivity) by Sector (%) (Inter-Sectoral Labor Productivity) by Sector (%) Other Services Agriculture Transport and Comms 20% Trade Industry Construction (excluding Services 66% Manufacturing) Electricity and Water 10% Manufacturing Mining and Quarrying Manufacturing Agriculture 4% –1 0 1 2 3 Source: National Accounts Directorate, MoFED. Central Statistical Agency (CSA): HICES 1995/96, LFS 1999, LFS 2005, HICES 2010/11. See Martins (2014) for details. 16 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS indispensable, as the majority of the labor force will will not result in large shifts of labor away from agricul- continue to work in the sector. Manufacturing growth ture. The services sector is also of high importance given is essential for structural transformation, but given the its potential for structural change and positive proper- size of the sector, even very high growth in this sector ties in terms of labor productivity levels and growth. 17 EXPORT PERFORMANCE AND COMPETITIVENESS 2 Ethiopia’s development model is partly inspired On the other hand, competitively priced imports used by the East Asian experience that realized high as inputs to the production process are an import part economic growth through the development of new of overall competitiveness of companies. export sectors and government-led development Exports indeed appeared as a driver for eco- investments. No doubt, exports played a major role nomic development in Ethiopia over the past in East Asia, and developing a larger export base in a decade—but the export engine is sputtering. High market-based system provides a unique opportunity export growth was one of many factors contributing for Ethiopia. However, the country’s exports mea- to Ethiopia’s economic takeoff since 2004. Although sured in percent of GDP falls short of reaching the causation is always difficult to prove empirically, it heights seen in Korea, China, or Vietnam during their is noteworthy that Allaro (2012) finds that exports development periods (World Bank 2012a). It remains “Granger cause”15 growth in Ethiopia.16 However, unclear whether current export levels in Ethiopia in as highlighted in Chapter 1, Ethiopian exports are fact are sufficient to support the course of large-scale exhibiting their worst performance in a decade. Even productivity increases and structural change men- if outside factors (e.g. declining prices) are partly to tioned in Chapter 1. blame, it is important to introduce policies now that Expansion of exports is often behind spurts can improve competitiveness and boost future export in economic growth. A thriving export sector helps and growth performance. align the domestic economic incentive structure with The challenge of development has become areas in which a country has comparative advantage. more complex since the rapid growth experience This is desirable from the perspective of resource of the East Asian economies. While East Asia relied allocation. Furthermore, successful exports create on manufactured exports for its growth, this course dynamic efficiency gains by exploiting economies of alone will not suffice for Ethiopia during an era of scale, adopting best practice foreign technologies and fast-changing modes of trade and production in the business processes, and by being subject to higher world economy. Growth and competitiveness today is international competition. Export sectors are also increasingly linked to a tight complementary potential associated with productivity gains leading to wage of exports and imports, as well as capital inflows, out- premiums and job creation. flows, and domestic investment to enhance productiv- There is also a foreign exchange element of ity in agro-based as well as classical manufacturing that exports that is important for sustainable growth increasingly draws on modern, competitive services as of an economy. Exports help finance imports, espe- intermediate inputs. cially of capital goods, and enable countries to main- tain a more favorable balance of payment situation. Ultimately this means that countries are in a better position to repay their external loans. Better availabil- 15 Granger causality means that one indicator (time series data) is able to forecast another. ity of foreign exchange in an economy will also ease 16 This finding is tempered by research conducted by Papageorgiou and the overall financing burden for companies to trade. Spatafora (2012). 18 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS This chapter analyzes Ethiopia’s export per- Export Performance17 formance according to four different dimen- sions, each modified analysis from the four Growth of Exports in Agriculture, key issues presented in the World Bank’s Trade Manufacturing, and Services and Competitiveness Diagnostic Framework (Figure 2.1): (1) the intensive margin (Growth of Rising exports contributed to Ethiopia’s remarkable growth performance over the past decade. Buoyed by favorable exports in agriculture, manufacturing, and services); external conditions, exports also helped create jobs and (2) the extensive margin (Structural transforma- earn much-needed foreign exchange. The way Ethiopia tion through diversification); (3) the quality margin created and nurtured a high-value horticulture industry and expanded its air services exports was an encouraging (Sophistication and quality); and (4) the sustainability example of “self-discovery.” A recent drop in export prices, margin (Sustainability of the export sector). however, has exposed underlying vulnerabilities in the export In a second step, the chapter will look at export structure and highlighted the importance of strengthening competitiveness issues underlying the performance competitiveness. Ethiopia is vulnerable to such price swings because unprocessed and undifferentiated agricultural pattern. To do so, the analysis will utilize the remain- products dominate its exports. While upward price trends ing elements of the Trade Competitiveness Diagnostics since 2003 have had a positive impact on exports, the Framework and focus on three dimensions: (1) the recent drop in prices of key commodities led to the worst performance in a decade. To overcome this challenge, incentive framework for trade; (2) factor inputs, renewed efforts must aim to improve competitiveness, productivity, and trade costs, and (3) proactive poli- including through value addition and export diversification. cies to promote trade (see lower part of Figure 2.1). A complementary section will look at the role of 17 This section is based on background papers prepared by Wagle (2014) and by Ferro and Fernandes (2013). A product level analysis based on the exchange rate as a determining factor of export exporter level customs data shows dynamics within selected key products competitiveness. in Annex 4. FIGURE 2.1: Trade Competitiveness Diagnostic Framework TRADE OUTCOMES ANALYSIS Intensive margin: Extensive margin: Quality margin: Sustainability margin: Orientation & growth Diversification Sophistication & quality Export survival Channels Entry & exit Factor and transactions costs Technology and efficiency DIAGNOSTIC Incentive framework for trade Trade, tax & competition Business regulatory External trade policy policy environment & governance environment Factor inputs, productivity and trade costs Intermediates and Labor markets, skills & Transport & trade backbone services technical efficiency facilitation Proactive policies to promote trade Export and investment Economic zones, clusters, Innovation Standards & certification promotion and industrial policy Source: Farole, Reis and Wagle (2010); and Farole and Reis (2012). Export Performance and Competitiveness 19 Benefiting from a global commodities price upcoming companies. For instance, incumbent (estab- windfall in the 2000s, Ethiopian exports grew at one lished) coffee exporters constitute around 70 percent of the highest rates in Sub-Saharan Africa. At an aver- of the total number of coffee exporters. The same is age annual growth rate of about 20 percent, Ethiopia’s increasingly true in oil seeds and cut flowers. In the lat- goods exports have doubled (in nominal terms) every ter, the importance of entrants into the cut flower busi- four years since 2001/02. This puts the country in the ness as a share of total cut flower exports has declined top decile of developing countries in terms of high from 15 percent in 2009 to less than 1 percent in 2012 growth in non-mineral exports not only in Africa but (for more product level analysis see Annex 4). With also globally. Export growth benefited immensely from the exception of the cut flower business, where foreign the surge in the price of Ethiopia’s main exports like participation is allowed, the fact that trading coffee and coffee and gold, as well as oilseeds, pulses, and spices. cereals is reserved for domestic investors may explain Figure 2.2.1 decomposes export growth into the price some of the low level of new entrants into these areas. and quantity effects. The price effect is clearly domi- At the same time, reliance on commodity exports nant, yet there was also a positive quantity response is also a source of vulnerability to international price to higher prices between 2003 and 2010. In the non- fluctuations. Food and beverages account for 77 per- commodities sector, export growth driven by prices cent of total goods exports, while agricultural raw mate- could be a reflection of rewards to improved quality. rials and metals account for 8 and 6 percent, respectively However, in the case of export baskets dominated by (Figure 2.2.3). Manufactured exports account for less commodities, as in Ethiopia, export growth can be than 10 percent. As was seen in section 1 (Figure 1.4.3), attributed more to worldwide movements in prices declining international prices are the main reasons than intrinsic improvements in domestic productivity. behind the recent poor performance of goods exports. Most of Ethiopia’s top exports are products Looking at price and quantity effects behind export for which world demand is increasing. Figure 2.2.2 growth rates show that export prices dropped so strongly compares Ethiopia’s product orientation with respect (by 15 percent in 2012/13) that it could not be balanced to the average world growth rate of specific imports by equally strong export volume growth. and individual markets. And indeed, most of its top Ethiopia’s trade activity is not only determined exports are those for which world demand is increasing, by exports, but also by a disproportionate increase including coffee, sesame seeds, soya beans, and foot- of imports—another source of vulnerability to the wear (left hand side of Figure 2.2.2). But how a coun- economy. Goods imports have always outstripped try’s industry is faring in international competition can exports in Ethiopia, but the gap has been widening also be gauged by its share in strategic markets, such since around 2000. Over the past decade, the trade as those that are expanding (and importing) rapidly deficit as a share of GDP has consistently hovered (right hand side of Figure 2.2.2): With the exception between 16 to 22 percent. Driven by the need to of China, Turkey, Egypt, and India, Ethiopia’s major purchase capital equipment for construction and export markets still lie in rich countries that are gen- industry, a higher share of imports is to be expected erally slow-growing (bubble size represents GDP of for low-income countries during a phase of rapid import destinations). Historically, most developing modernization. However, Ethiopia’s imports, and countries have grown by first exporting to rich coun- consequently its trade deficit, are much higher than tries, but it is also preferable to maintain a foothold in what has been seen in the past in other countries.18 fast-growing emerging economies (in which case the regression line would ideally have been more positive). 18 For instance, Vietnam’s average trade deficit from 1991 to 2001 was 6.4 percent of GDP, in contrast to Ethiopia’s 18.2 percent during Yet, the commodity export business is not 2002–2012. Since opening up, China has always run a trade surplus particularly dynamic and promising for new and (except in 1993). 20 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS The unusually high trade deficit makes Ethiopia vul- as agricultural raw materials, can be expected to share nerable to external shocks because private transfers more product characteristics with the processed foods (remittances) and official aid from abroad finance category, which includes fresh fruits and vegetables, much of the imports. poultry, fish, and dairy products. These are known to Agriculture proved to be a driving force in fetch higher value in world markets than unprocessed output growth over the past 15 years, second only agricultural commodities. They require some form of to services (see the “Long View” of Chapter 1 for technological processing before being exported and details). At the same time agriculture is the sector in are then typically subjected to stringent food safety which most new employment was generated, with standards. Processed foods are, therefore, distinct from a share of more than 72 percent of the 11.6 million traditional beverages (such as tea and coffee) and cereal jobs created over the past 15 years. Yet, labor pro- grains (such as wheat, maize, or rice) which are gener- ductivity advances are relatively low in the sector ally exported in bulk. with growth rates at around 2.5 percent compared to There are three reasons why processed food, 6.1 percent for the services sector and 5.4 percent in and by the same logic, horticulture, are important the trading sector. Given the large size of the agricul- for export growth: First, income and price elasticity ture sector, it is imperative that continued efforts are of demand for processed food are higher than most made to make the sector more productive. Indeed, traditional primary agricultural products. Therefore, relatively higher past productivity growth rates in diversification of the export mix into this commod- the trading sector indicate the potential of advanc- ity category can nudge export growth combined ing agriculture trade. with terms of trade gains. Second, the final stages of Ethiopia’s expansion of horticulture marks a food processing are labor-intensive and help create spectacular export success of the past decade. The jobs. Finally, processed food products typically have cut flowers industry grew from one single firm in the greater domestic input content and value-addition year 2000 to about 100 firms today, contributing to (Athukorala and Waglé 2011). export earnings to the tune of US$200 million (Dinh Ethiopia has a revealed comparative advan- et al. 2013). Estimates are that the sector employs tage in about 80 export products. Comparing the more than 50,000 individuals who, in turn, support relative share of Ethiopian export sectors with corre- the livelihoods of about 250,000 people. While over sponding shares for the world computes the Revealed 80 percent of the flowers are destined for the Dutch Comparative Advantage (RCA). If the value of RCA auctions, there have been recent efforts to seek new exceeds one for a sector, the country is said to have markets. New routes opened by Ethiopian Airlines, “revealed” comparative advantage in that sector. In such as South Korea and Singapore often determine terms of major export sectors, Ethiopia has a revealed the direction of this search. Indeed, a decisive factor comparative advantage in two of eight, namely in the exponential growth of the flower industry is food and beverages (RCA: 9.8) and agricultural raw the expansion of Ethiopian Airlines’ cargo capacity materials (RCA: 4.3).19 Somewhat surprisingly, given and passenger flights. With a functioning air cargo recent foreign direct investment, the results suggest system now in place, the experience of the flower that Ethiopia does not have a revealed compara- industry could be relevant to developing new (diversi- tive advantage in apparel and footwear (RCA: 0.5). fied) export opportunities, which are in close “proxim- On the other hand, Ethiopia has augmented its ity” to flowers (see Annex 3). The discovery of this new export activity is 19 The other major sectors include (RCA in brackets): Fuels, ores, metals welcome against the backdrop of a nascent manu- (0.2), chemicals (0), material-based manufactures (0.5), machinery & facturing industry. Cut flowers, which are classified equipment (0), other manufactures (0). Export Performance and Competitiveness 21 performance in this category compared to a decade Turkey have heavily invested in the apparel sector, as ago where RCA was 0.1.20 In three products, Ethiopia have the Chinese in footwear. ranks among the top eight exporters in the world: Services exports are booming largely due fourth in cut flowers, second in sesame seeds and to Ethiopian Airlines. Ethiopia is among the few eighth in coffee beans. developing countries where services exports are as Ethiopia’s nascent manufacturing industries important as goods. Between 2005 and 2012, the are beginning to grow rapidly from a low base. services-to-goods export ratio hovered around one, Sectors such as leather and footwear are attracting implying that services exports were as large as goods FDI. The Huajian Chinese shoe company came exports (Figure 2.2.4). Services exports are dominated to Ethiopia in 2011 and started exporting to well- by transport (63 percent), followed by construction known brands in the United States. Another sig- (15 percent), other business (10 percent), travel nificant investor in this sector, George Shoes from (5 percent) and insurance (4 percent). The majority Taiwan, is also expanding its export-related capacity. of services export is attributed to Ethiopian Airlines, Figure 3.1 in Annex 3 illustrates the location of sev- which is Ethiopia’s biggest export earner—three times eral leading and emerging exports from Ethiopia on as big as coffee. a network that connects pairs of tradable merchan- Ethiopia’s services exports are higher than dise goods that are co-exported by a large number those of countries at its own level of development. of countries. This product space analysis identifies Figure 2.2.5 illustrates the results of a regression of several exports that could potentially be scaled up the services exports-to-GDP ratio on the GDP per in Ethiopia given the capabilities they share with capita in a cross-country sample. Ethiopia is located products already exported, or other desirable prop- above the regression line, as are several of its peers. erties they have such as inherent complexity or high As countries develop, the share of services (relative to demand in world commerce. value-added as well as exports) is expected to increase. Manufacturing exports have been positively While the share of services value added has increased affected by changes in the global trade regime and over the past decade (see Chapter 1), this was not the the introduction of new trade preferences. Within case for services exports as illustrated in Figure 2.2.4. manufactures, those products more closely related Other developing countries with large trade expansion to the primary sector (such as leather and wood) are often owe their success to growth in goods exports, the most significant. In leather-related industries which has also not been the case for Ethiopia. In sum, in particular (such as shoes and gloves), Ethiopia is while Ethiopia’s performance in services exports is beginning to nurture capabilities for higher domestic noteworthy, it also accentuates the converse: the degree value addition. Compared to a decade ago, apparel to which Ethiopia under-exports goods. Indeed, when and textiles have also become more prominent. This the dominant export of transport services is excluded, comes as no surprise because the period coincides with the share of services exports to GDP falls from 7.6 the end-2004 termination of the WTO Agreement percent to 2.8 percent, and Ethiopia’s performance on Textiles and Clothing, which ended quotas that is no longer an outlier for a country at its stage of governed global trade in garments for decades. The development.21 end of this distortionary regime led to a significant re-orientation of production locations benefiting low-income African economies like Ethiopia, which 20 This confirms an important caveat about the RCA Index that it is only additionally benefit from generous trade preferences a ‘static’ measure of competitiveness. It says little about the future, and cannot tell which sectors are less promising over time. in the EU and the United States with a near-universal 21 When export of transport services is excluded, Ethiopia’s performance coverage of goods. In recent years, investors from is in line with that of Kenya and Rwanda, but below Uganda. 22 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 2.2: Growth and Orientation of Exports 1. Exports: Price and Quantity, 2000 to 2001 2. Orientation of Exports 2002–2012 600 Products Destinations Export growth (index for year 2000=100) 4 4 National import growth 2002−12 (log) World import growth 2002−12 (log) 500 3 3 400 2 2 300 1 1 200 0 0 100 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 −6 −4 −2 0 2 −6 −4 −2 0 2 4 Product’s share in exports Destination’s share in exports Value Volume 2011−12 (log) 2011−12 (log) 3. Ethiopia’s Export Structure, 2012/13 4. Relative Growth of Services Exports Ethiopia Vietnam Coffee 13% 4 120 Services exports (US$ billion) Transportation (EAL) Gold 32% 10% 3 2012 90 Oil seeds Travel 7% 2 60 9% Chat Other services 4% 7% Pulses 1 2005 30 Textiles 4% 2% Other goods 2012 2005 Leather 4% Live animals Flowers 0 0 2% 3% 3% 0 1 2 3 0 30 60 90 120 Goods exports (US$ billion) 5. Services Export (% of GDP) in 2011 6. Export-to-GDP 2002 2012 30 100 100 VNM Export of goods and services 75 75 relative to GDP (%) 20 Percent of GDP KOR VNM 50 50 ZMB 10 RWA ZMB KOR TZA ETH VNM KEN CHN UGA KOR 25 CHN 25 KEN KEN ZMB TZA UGA CHN UGA ETH ETH 0 TZA 0 0 6 8.5 11 6 8.5 11 6 8.5 11 Log of GDP per capita (PPP) Log of GDP per capita (PPP) Source: (1) and (2): WDI and UN Comtrade. (3) WB Services Trade Database. (4) UN Comtrade. (5) CEPII, WDI and UN Comtrade. (6) WDI and UN Comtrade. Export Performance and Competitiveness 23 Structural Transformation through the “Long View” section of Chapter 1 and illustrated Diversification in Figure 2.3.1. Export diversification matters for growth. While Ethiopia’s export sector is currently too small to contribute Ethiopia’s recent success in horticulture and progress to structural transformation. In East Asia, booming in labor-intensive light manufacturing augurs well the (manufacturing) exports helped shift economic activity and workers away from low-productivity agriculture into eventual modernization of the economy, there is much higher-productivity manufacturing and sustain high rates of room for transformation towards higher-productivity economic growth for decades. In Ethiopia, both exports and manufacturing, as well as services. It matters if coun- the manufacturing sector remain relatively small: it has the lowest ratio of merchandise exports to GDP among populous tries earn high export dollars from a domestic produc- countries in the world. The country has 1,800 exporting firms tion base that is well diversified (rather than a narrow compared to 4,600 in Kenya (with half the population). basket of sectors) because the former can expect a more Average exporter size is small (US$1.2 million versus US$4.1 million in Zambia). The Ethiopian manufacturing sector sustainable growth pattern. McKinsey (2010) argues accounts for only 4 percent of GDP . that as countries develop and increase real export per capita, they tend to meet both the objectives of earning Ethiopia has a relatively closed economy, foreign exchange to finance capital imports needed for though trade openness has increased substan- investment, and developing a diverse source of growth tially over the past decade. Exports and imports away from natural resources and agriculture.22 of goods and services as a share of GDP increased The challenge for Ethiopia is to further diver- from 37.5 percent in 2001/02 to 48.7 percent in sify the economy and boost exports. This point is 2011/12. Ethiopia’s degree of international integra- illustrated in Figure 2.3.1 (right panel), which plots tion lags behind countries such as Kenya and Tanzania countries in terms of their shares of service and manu- (74.9 and 79.8 percent of GDP, respectively) while facturing sectors and exports per capita. The diagram it exceeds that of Rwanda (45.2 percent). In fact, is then split into four quadrants, using the median Ethiopia has the lowest goods export-to-GDP ratio values of the two indicators. The long-term implica- (7 percent) among populous developing countries. tion for development policy is to nudge countries like The country appears to under-export goods and Ethiopia in the northeast direction towards Korea, services by over 10 percentage points of GDP. This China, Vietnam, and Kenya. insight is derived from Figure 2.2.6, which illustrates Ethiopia has experienced modest export a cross-country regression model of the exports-to- diversification over the past decade. Figure 2.3.2, GDP ratio on GDP per capita controlling for popu- illustrates the Hirschman–Herfindahl (HH) index, lation size and the cost of exporting. For a country which is the sum of squares of the shares of export of of Ethiopia’s size and location, there is substantial a country in its total exports. A country with a per- potential to increase exports further. fectly diversified export portfolio will have an index Ethiopia’s small manufacturing sector implies close to zero. Ethiopia has a degree of export concen- that the domestic economy is not yet sufficiently tration in line with what is expected given its level of well diversified to wean exports away from agricul- development. Encouragingly, aggregate exports have ture. Ethiopia has one of the highest shares of agricul- diversified over the past decade in terms of product ture in GDP in the world (47 percent in 2011–12). composition, unlike in Tanzania or Zambia. While the share of services in domestic value-added is moderately high for a developing country, the contri- 22 See also Lederman and Maloney (2012) who demonstrate the im- bution of manufacturing is negligible. The combined portance of export diversification, even for countries such as Costa Rica (in the case of Intel semi-conductors). Rodrik (2013b) also stresses the share of manufacturing and services has increased importance of diversification for economic growth in low—and middle- only modestly over the past decade, as discussed in income countries. 24 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS However, in terms of products exported per developed a country is the more entrepreneurs a coun- firm,23 the measure of export diversification is try has (Figure 2.3.5). In a cross-country comparison lower than in peer countries. Ethiopian firms with 100 countries, however, Ethiopia has lower-than- export on average less than 4 products which makes expected entrepreneurs, lagging behind countries such them the least diversified among firms in East Africa as Uganda, Kenya, Rwanda, and Zambia. (Figure 2.4.1) such as Kenya, which exports on aver- age 8 products. The diversification per exporter also Sophistication and Quality appears to have declined over time in Ethiopia, from 4.4 in 2009 to under 3 in 2012 (Figure 2.4.2). In terms More than “what” is being exported it is the “how” that of the average number of markets served by each firm, is hindering potential. There is scope for improving the Ethiopia’s ratio of 2.4 compares well with countries at quality of existing commodity exports, through basic value addition, such as coffee wet processing or machine flaying of similar levels of development, and has increased over animal skins. Even in products with a revealed comparative the past few years. advantage, little upgrading or branding has occurred to earn Five percent of firms in Ethiopia account for higher value per unit over time. By starting to compete on the quality of existing commodity exports (and not just on about 65 percent of the country’s total exports. price), Ethiopia can reduce sensitivity to volatile international This ratio is lower than in Zambia (97 percent), prices thereby supporting the gradual shift of production Kenya (79 percent) and Uganda (73 percent). The and exports into agro-processing and light manufacturing. top 5 percent also have annual exports above US$5 million; however, the majority of exporters have More than “what” is being exported, it is the total exports lower than US$100,000 (Figure 2.4.3). “how” that is hindering potential. Semi- or un- Figure 2.4.4 shows that large-sized firms make up processed agricultural goods dominate Ethiopian most of total exports in Ethiopia. What is notewor- exports. In the short run, there could be a greater thy though is that although the big firms dominate focus to augmenting value in existing exports. Take exports, in terms of overall distribution, Ethiopia’s coffee: Ethiopia exports green beans, over two-thirds exporter firms are less concentrated by size compared of which are sun-dried but not wet-processed (washed to Zambia or Tanzania where more than 90 percent of or semi-washed), yet wet-processed beans are cleaner exports are accounted for by firms with annual exports and earn a significant mark-up. There is almost no above US$5 million. roasting in-country, which the European importers The number of exporters has been on a declin- do instead. Roasting increases the value of beans by ing trend since 2010. Overall, Ethiopia’s number of about 200 percent.24 In cut flowers, use of higher qual- exporters is not significantly different from that of its ity packaging dramatically increases prices, but value peers after controlling for income per capita, size and can be unlocked with better management of freight time trends. Ethiopia has on average fewer exporters and the cold chain. Instead of exporting live animals, than Tanzania and Kenya but more exporters than processed meat is more lucrative (while retaining hides countries like Zambia, Botswana, or Uganda. In fact, and skins for the leather industry), yet HACCP and Ethiopia experienced an important expansion in the number of exporters from 1,475 in 2008 to 2,033 in 23 Measured as the average number of exported HS 6-digit products per exporter. 2010 (an increase of 27 percent). But between 2010 24 Coffee exporters in Addis Ababa say that roasting involves specialized and 2012 the number of exporters declined again by knowledge about demand for specific tastes and blends by location, which the European importers have mastered; they source a diverse range of about 10 percent (Figures 2.3.3 and 2.3.4). coffee (including robusta) and use their knowledge to blend coffees in Not limited to export businesses, Ethiopia has profit maximizing ways. Roasting is also a machine-dominated activity, and low wages alone would not be sufficient to lure them to locate in mainly low level entrepreneurs to support a grow- Ethiopia. Once roasted, coffees lose flavor quickly, so efficient trade ing economy and supply jobs. Generally, the more logistics becomes critical. Export Performance and Competitiveness 25 FIGURE 2.3: Export Diversification and Number of Exporters 1. Exports and Economics Structure 2. Exports Concentration 100 20 All Goods (2-digit) All Goods (4-digit) CHN CHN Log of real export per capita 2010−12 Services & manufacturing in GDP (%) 90 Korea, Rep. 18 ETH ETH 80 16 Korea, Rep. KEN KEN China 2000−2002 70 RWA RWA Zambia Kenya 14 China 60 Vietnam Vietnam TZA TZA Uganda Rwanda Zambia UGA UGA Tanzania 50 12 Kenya Tanzania VNM VNM Ethiopia Uganda Rwanda 40 ZMB ZMB 10 Ethiopia 30 0 0.2 0.4 0.6 0 0.1 0.2 0.3 0.4 0.5 30 40 50 60 70 80 90 100 30 40 50 60 70 80 90 100 Hirschman−Herfindahl Index Services & manufacturing in GDP (%) 2010−2012 2001−02 2011−12 3. Number of Exporters, Selected Countries 4. Number of Exporters in Ethiopia, 2008–2012 Kenya 4,610 2000 Ethiopia 1,825 1900 Number of Exporters Tanzania 1,441 1800 Zambia 1,796 Bostwana 1700 1,391 Uganda 910 1600 Cambodia 597 1500 0 1,000 2,000 3,000 4,000 5,000 2008 2009 2010 2011 2012 Number of Exporters Year 5. Business Registration Density, 2010 9 8 # new firms per 1,000 working 7 age population (Density) 6 5 4 3 Brazil 2 Malaysia Zambia Korea, Rep. 1 Uganda Kenya Rwanda Thailand 0 Ethiopia –1 2.5 3.0 3.5 4.0 4.5 5.0 5.5 log (GDP per capita, PPP, constant 2011 international $) Sources: Authors’ calculations, based on data from UN Comtrade, and the Exporter Dynamics Database, Entrepreneurship Snapshots (2010), and World Development Indicators (2010). Note: Figures 5 and 6 show averages for the time periods: 2008–2012 for Ethiopia, 2007–2011 for Tanzania and Zambia, 2007–2010 for Bo- tswana and Uganda, 2006–2009 for Kenya, and 2007–2009 for Cambodia. 26 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 2.4: Size Distribution of Ethiopian Exporters 1. Average Number of Products Per Exporter, 2. Average Number of Products Per Exporter in Selected Economies Ethiopia, 2008–2012 4.5 Avg. number of products per exporter Cambodia 8.1 Bostwana 8.1 Kenya 7.8 4.0 Zambia 5.8 Tanzania 4.4 3.5 Uganda 4.1 Ethiopia 3.7 3 0 2 4 6 8 2008 2009 2010 2011 2012 Average number of HS 6-digit products per exporter Year 3. Distribution of Exporters Across Firm Sizes: 4. Distribution of Exporters Across Firm Sizes: Share of Total Number of Exports Share of Total Merchandise Exports 0.8 1.0 0.8 0.6 Share of total exports Share of exporters 0.6 0.4 0.4 0.2 0.2 0 0 500K–1M 1M–5M >5M 500K–1M 1M–5M >5M 500K–1M 1M–5M >5M 500K–1M 1M–5M >5M 500K–1M 1M–5M >5M 500K–1M 1M–5M >5M 500K–1M 1M–5M >5M 500K–1M 1M–5M >5M 500K–1M 1M–5M >5M 500K–1M 1M–5M >5M 100K 100–500K 100K 100–500K 100K 100–500K 100K 100–500K 100K 100–500K 100K 100–500K 100K 100–500K 100K 100–500K 100K 100–500K 100K 100–500K Bostwana Ethiopia Tanzania Uganda Zambia Bostwana Ethiopia Tanzania Uganda Zambia Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graphs 1, 3 and 4 show the following averages: 2008–2012 (Ethiopia), 2007–2011 (Tanzania and Zambia), 2007–2010 (Botswana and Uganda), 2006–2009 (Kenya), and 2007–2009 (Cambodia). Graphs 5 and 6 are based on data for the year 2010 for all countries. ISO standards are generally not met. Further, most a view to encouraging domestic value-addition. animal skins are hand-flayed, whereas machine-flayed However, because the quality of leather treatment is skins generally earn 20–30 percent more in value not yet of high standard, value of leather is lost. (Sutton and Kellow 2010).56 Processing of the tons Almost none of the exports in which Ethiopia of oilseeds, pulses and spices is also minimal, with has a Revealed Comparative Advantage can be most products largely exported raw. These are not considered complex. Product complexity is based new observations: value-addition has been a mantra for years, yet progress has been slow. Sometimes good 25 It is important to note that machine flaying of animal skins requires the operation of modern meat processing plants or use of abattoirs towards intentions have backfired: the government introduced improving supply to both meet processing and tannery industries. These a 150 percent tax on the export of crust leather with are not readily available in Ethiopia. Export Performance and Competitiveness 27 on a function of two variables: diversity (how many future, possibly by building upon existing know-how products does a country make?) and ubiquity (how and adopting better technologies. many countries a product is made by?). The empiri- The income content of Ethiopia’s export basket cal observation is that a product made by only a few is slightly less sophisticated than expected given countries (that also have the capability to produce its levels of development. In Figure 2.5.2, Ethiopia’s many other products) tend to be complex. Similarly, export basket is shown to be slightly poorer than products made by many countries that produce few what its average per capita income suggests. This is other products tend to be less complex (Hausmann based on one of the measures of export sophistication et al. 2011). Only 3 of Ethiopia’s 71 export products (EXPY), which assesses the export baskets of coun- in which it has a revealed comparative advantage tries by the incomes of countries that produce similar have above-average complexity. They are all material- products, weighted by the share of those exports in based manufactures: yarn of regenerated fibers (SITC: the national basket.26 The sophistication of Ethiopia’s 6517), continuous regenerated woven fabrics (6535) and exports is comparable to that of peers like Rwanda and polishing stones (6631). Conversely, all of Ethiopia’s Tanzania and ahead of Zambia, but behind Uganda, top 20 exports (accounting for over 90 percent of Vietnam, and China. This reflects, again, the domina- export earnings) are in the bottom half of the Product tion of Ethiopian exports by agricultural products not Complexity Index, with the most important products exported by many rich countries—the lower income (coffee, sesame seeds, fresh/chilled vegetables, legumes, content of such products drags down the average and flora) deemed least complex globally. sophistication of the export portfolio. Ethiopia has a high potential to expand or This is of importance because export sophistica- upgrade export sectors where the core competen- tion is a good predictor of future economic growth. cies (land, labor, capital, and institutions) are simi- As shown by Hausmann and Klinger (2007), countries lar to those already acquired. Figure 2.5.1 plots all like Vietnam and China that have more sophisticated of Ethiopia’s goods exports along two dimensions: export baskets than expected for their level of income how complex (y-axis) they are, and how proximate grow faster and “become what they export.” That is (x-axis) they are in terms of productive knowledge they specialize in activities that are more typical of rich to the basket of goods with revealed comparative countries today. Such countries are located above the advantage (colored in blue). Product proximity is the line in Figure 2.5.2. The implication for countries like degree to which each of Ethiopia’s export products Ethiopia is that growth would require the discovery of shares a common productive knowledge with the new, more sophisticated export activities. This could basket of goods that are comparatively competitive mean going beyond enhancing productivity in existing (RCA>1). The closer a product is to the portfolio of activities, and indeed putting a disproportionate focus existing goods that are competitive, the easier it would on discovering new export activities (Klinger 2010). be for a country to acquire the capabilities to scale up Encouragingly, Ethiopia has done this before: the exports. In the case of Ethiopia, there are more than discovery and nurturing of the horticulture industry 400 exports (below the average proximity index of is a textbook example thereof. 15) that are “near” existing exports with a revealed Goods with low human and physical capital comparative advantage. In other words, hundreds dominate the factor content of Ethiopian exports. of exports share close and complementary produc- tive knowledge with goods that are already competi- 26 EXPY is preceded by the calculation of the “income content” of prod- tive. This suggests that, with the right incentives and ucts (PRODY). It is calculated by using the method of Hausmann et al. (2007). EXPY is the weighted sum of the PRODY of all the products a business environment, many of these exports have country exports. Higher EXPY means the country has a larger share of the potential to grow and become competitive in the more sophisticated (high PRODY) products. 28 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS The two diagrams in the top of Figure 2.5.3 plot the the 12 products with a market share of over 0.1 per- physical and human capital content of Ethiopia’s cent, Ethiopia increased both the market share and export basket in 2002 and 2012. The size of the bub- relative quality in 6 of them; gained market share but bles reflects the relative importance of each product lost relative quality in 3; and lost on both aspects in 3 in total goods exports. The dashed reference lines are (including sesame seeds). One reason that there is not the median values of revealed human and physical a more noticeable increase in relative quality of exports capital. In 2002, most of Ethiopia’s big export earners is that most of Ethiopia’s exports are homogeneous utilized substantially less physical and human capi- goods rather than reference-priced or differentiated tal than the median. By 2012, the median values of goods (Rauch 2006). The first two groups include capital content had increased modestly. While some commodities that are traded in organized exchanges major exports with low-to-medium capital content or whose reference prices can be obtained from trade have disappeared between 2002 and 2012 (bottom publications (without even knowing the name of the left diagram), there is some dynamism on the “new manufacturer). The examples of products in this cat- export” front (bottom right diagram). In other words, egory are agricultural commodities, metals, and chem- new exports are numerous, and a few major ones are icals. The price of differentiated goods, on the other moderately intensive in physical and human capital. hand, varies with the brand of the manufacturer. As Ethiopian exports generally do not embody countries become richer, they tend to specialize more any modern technology. The dominance of primary/ in differentiated products that present room for prod- agricultural exports is manifest in yet another indica- uct upgrading and earning of higher value per unit. tor: the technology content of final exports. Ethiopia’s absence of foothold in medium-to-high technology Sustainability of the Export Sector intensive manufacturing ranks it alongside countries like Zambia that are also highly reliant on primary The business environment favors incumbent firms and deters commodities (Figure 2.5.4). Compared to China and new entrants into export businesses, and even so, no “export superstars” are emerging. The export sector lacks dynamism Vietnam, or East African countries, Ethiopia’s share of in terms of firm entry and exit. Rather than increasing in exports with a medium to high degree of technology scale, growing from small to large, new entrants to the content is conspicuously low. The caveat here is that export market are already often relatively well established in other businesses such as trading. This may be due to the fact in an era of global production sharing, technological that smaller firms have limited access to credit, which would groupings of high-tech and medium-tech products have allowed them to increase the scale and scope of their may be misleading, as a country may export sup- activities. This is a challenge because rising and dynamic firms often generate more new jobs than established posedly high-tech goods (like computers), but its firms. Other factors such as low entrepreneurship and low role may simply be in the final stages of low-value- regulatory quality in terms of promoting the private sector adding assembling operations. Similarly, Ethiopia’s may also explain this. Despite a favorable environment for incumbents, they are yet to emerge as multi-product and success in the export of skill-intensive floriculture multi-destination export superstars. or processed foods is under-stated by this indicator given the assumption that agricultural exports are less The business environment favors incum- technology-intensive. bent firms and deters new entrants into export Despite the dominance of homogenous exports businesses. The average size of exporters is much with little room for quality differentiation, Ethiopia lower in Ethiopia (US$1.2 million) than coun- has made inroads into the European market while tries like Botswana (US$3 million) and Zambia sustaining relative quality. Figure 2.5.5 plots the (US$4.1 million), but comparable to Kenya and change in the relative quality of Ethiopian exports Uganda (Figure 2.6.1). The median size of exporters, and the change in their market share in the EU. Of however, is much higher than all comparator countries Export Performance and Competitiveness 29 FIGURE 2.5: Export Sophistication and Survival 1. RCA in Complex Products 2. Export Sophistication 2 20 1 CHN Product complexity index 15 EXPY (in US$ ’000) 0 VNM 10 –1 UGA 5 ETH RWA –2 TZA ZMB –3 0 5 10 15 20 25 30 35 40 45 50 6 7 8 9 Product proximity Log of GDP per capita 2012 No RCA RCA 3. Revealed Factor Intensity of Exports RFI 2012 RFI 2002 12 12 8 Human capital Human capital 8 4 4 0 0 0 60000 120000 180000 0 60000 120000 180000 Physical capital Physical capital 12 12 Human capital 8 8 Human capital 4 4 0 0 Dead exports by 2012 New exports in 2012 (continued on next page) 30 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 2.5: Export Sophistication and Survival (continued) 4. Technological Classification of Exports 5. Change in Quality and Market Share (in EU) 6 80 70 Diff. in relative quality 2005−2010 60 Share in total exports (%) 2 50 40 30 –2 20 10 –6 CHN ETH KEN KOR RWA TZA UGA VNM ZMB –8 –4 0 4 8 HighTech MediumTech LowTech Primary Resource Diff. in market share in EU, 2002−2011 Agri Manu 6. Survival of Aggregate Exports To Neighbors By PTA 0.99 0.99 0.66 0.66 Probability 0.33 0.33 0.00 0.00 0 1 2 3 4 5 6 7 8 9 10 11 0 1 2 3 4 5 6 7 8 9 10 11 Year Year Non−Neighbors Neighbors Non−PTA partners PTA partners 0.99 0.99 0.66 0.66 Probability 0.33 0.33 0.00 0.00 By initial export value By export type Low High Non−manu Manu Source: (1) The Observatory of Economic Complicity and Comtrade. (2) UN Comtrade. (3) UNCTAD. (4) WDI. (5) CEPII and UN Comtrade. (6) UN Comtrade. Export Performance and Competitiveness 31 in Africa (Figure 2.6.2). This difference between the too, supports the view that entry costs are difficult for mean and the median suggests a skewed distribution small firms to meet. of exporters in terms of size in Ethiopia—but one that There is some variation in export survival is not as dramatic as in most other countries. The ratio rates depending on export types and size of initial of the average and median exports is 500 in Botswana, exports. In Figure 2.5.6 the bottom left diagram 56 in Uganda and 39 in Kenya, compared to just 11 shows that export spells that start big (with orders val- in Ethiopia. In other words, the degree of dissimilar- ued at least US$50,000) have higher rates of survival. ity between the median exporter and the larger firms This is consistent with the trade literature that associ- towards the top of the distribution is not as wide as ates survival patterns with search costs, and finds that elsewhere in the region. higher initial export value conveys a degree of trust, The export sector lacks dynamism in terms and an investment already made in the supplier. The of firm entry and exit, where exporter turnover fourth graph (right, bottom panel) shows that export is relatively low (Figures 2.6.3–2.6.4). On average, survival does not depend on broad product types: in a given year, 36 percent of Ethiopian firms that manufactured exports do not fare better in terms of export did not do so in the previous year whereas survival than agricultural products. Given the nascent 32 percent of Ethiopian firms that were exporting in stage of Ethiopian manufacturing this is unsurprising. the previous year stopped exporting. This indicates Survival rates of new exporters are most promis- that there a relatively low level of “churning of export- ing in SSA and least promising in the United States. ers” (renewal of exporters) in Ethiopia. In fact, entry The one-year survival rates (at the firm-product-desti- rates by Ethiopian firms into export markets have nation level) range between 40 and 50 percent while declined since 2008 while exit rates have been rising. the two-year survival rates are much lower ranging Low dynamism is also manifested through the from 20 to 35 percent (Figures 2.6.5–2.6.6). Despite fact that established and rather large companies the fact that the entrants’ share of Ethiopia’s exports to engaged in exports from Ethiopia exhibit high SSA in 2012 is the lowest among all destinations, those one-year and two-year exporter survival rates. entrants perform comparatively well as their average High survival rates are often seen in environments one-year and two-year survival rates are the highest dominated by high fixed costs to enter the exporting among all destinations. This is not too surprising given business. High fixed costs are driven by a mix of fac- that product requirements are lower in SSA than in tors in Ethiopia ranging from infrastructure and trade more sophisticated markets. Geographical proximity logistics (where individual companies may need their and cultural similarities also help. The EU seems to own fleet to transport their inputs and outputs) to low be a difficult market over the medium run, with the regulator quality to constraints on the credit market two-year survival rate of just 22 percent. (where new market entrants of smaller size have a The reason that larger or established companies low chance of receiving finance from private banks). are dominating the export business may reflect Smaller firms suffer most in a high fixed cost environ- the fact that smaller firms have limited access lack ment, and the few firms that are able to overcome high access to credit, which would have allowed them to costs of entry are then those that are large enough to increase the scale and scope of their activities pre- stay in business for a longer period. It is no surprise venting them from growing bigger. A forthcoming that this produces an environment where incum- World Bank study on SME finance in Ethiopia shows bent firms have relatively high survival probabilities beyond the second year.27 The average size of entrants 27 This reasoning draws on recent trade models with heterogeneity across in Ethiopia is also big, with the median entrant nearly firms. See Bernard, Jensen, Redding, and Schott (2011) on the empirical five times larger than in comparator countries. This, implications of such trade models. 32 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS the existence of a “missing middle” phenomenon, that cover legal enforcement, competition advocacy, whereby small enterprises are more credit constrained and institutional effectiveness (Gill et al. 2014). than either micro or medium/large enterprises. This Despite a favorable environment for incumbents, may be explained on one hand by the lack of adequate they are yet to emerge as multi-product and multi- business models to serve SMEs from financial institu- destination export superstars. Large firms often define tions, which in turn reflects the lack of an SME finance exports from one country; well-known examples culture (i.e. no harmonized SME definition nor SME include Nokia in Finland, Samsung in Korea, and specific strategy) and on the other hand by the exces- Intel in Costa Rica (Freund and Pierola 2012). But sively high collateral requirements that often discour- Ethiopian exporters are poorly diversified both in terms age SMEs from even applying for loans; the latter is of products and destinations. They exhibit significantly being aggravated by SMEs having low levels of asset lower numbers of HS 6-digit products exported per accumulation to collateralize and the added require- firm and significantly lower number of destination ment for domestic borrowers to provide personal markets per firm than comparable countries. This unlimited guarantees. This represents a key challenge poor diversification performance is also explained by because typically young firms are a great source of job the absence of a few highly diversified multi-product creation but this trend is not seen in Ethiopia, where multi-destination export superstars dominating more established firms dominate the net job creation, exports, a phenomenon commonly observed in other suggesting that there is a lack of competitiveness and countries. In Ethiopia, exporters selling 4 or more innovation in the private sector. products and serving 4 or more destinations account Other factors such as low entrepreneurship for 5 percent of the total number of exporters but and low regulatory quality in terms of promoting for just 20 percent of total exports—a share which is the private sector may also provide an explanation. substantially lower than those in comparator countries Figure 2.3.5 showed earlier that Ethiopia has a very (the next lowest share is Uganda’s 41 percent). low level of entrepreneurs. At the same the dominance of state-owned enterprises in crucial economic sectors Export Competitiveness28 in Ethiopia is well known. This influences the inten- sity of local competition and the rigor and fairness Incentive Framework for Trade with which anti-monopoly policy is applied including against state-owned firms that obstruct competition. Ethiopia lags behind its peers in Global Competitiveness These are issues in which Ethiopia ranks relatively low Rankings and its performance is on a declining trend. Although the country ranks better than its peers on theme- in the Global Competitiveness Report 2013–2014: specific business regulatory measures such as enforcing its overall rank is 127 out of 148 economies, but on contracts, its overall Doing Business performance is on aspects of goods market efficiency such as “intensity of the decline. Low regulatory quality also adds to the cost of doing business for exporters. Furthermore, Ethiopia has local competition,” “extent of market dominance,” and more trade restrictions in place than peers — its average “effectiveness of anti-monopoly policy,” the country nominal rate of protection is 25.3 percent. But high levels ranks below average, at 133rd, 144th and 131st, respec- of nominal and effective rates of protection provide strong evidence of the widespread existence of an anti-export tively (Figure 2.7.2). There is empirical evidence on bias of the tariff regimes throughout the economy. At the the link between effective competition policies and the same time Ethiopia is under-tapping a narrow window of expansion of an efficient private sector. Indeed, the opportunity for diversification through full exploitation of existing trade preferences. best practice is not just implementing a legal frame- work for competition that targets market dominance, monopolistic collusion, unfair competition, and anti- 28 This section is based on background papers prepared by Wagle (2014) trust investigations, but a broader set of regulations and Nguyen (2014). Export Performance and Competitiveness 33 FIGURE 2.6: Exporter Dynamics: Entry, Exit, Size and Survival 1. Average Exporter Size, Selected Economies 2. Median Exporter Size, Selected Economies Cambodia 5.2 Cambodia 0.644 Zambia 4.1 Ethiopia 0.109 Bostwana 3.0 Tanzania 0.026 Tanzania 1.6 Uganda 0.023 Uganda 1.3 Kenya 0.023 Ethiopia 1.2 Zambia 0.016 Kenya 0.9 Bostwana 0.006 0 1 2 3 4 5 0 0.2 0.4 0.6 Average exporter size in millions of USD Median exporter size in millions of USD 3. Entry Rates, Selected Economies 4. Exit Rates, Selected Economies Tanzania 0.47 Tanzania 0.43 Uganda 0.42 Zambia 0.40 Zambia 0.42 Kenya 0.37 Ethiopia 0.36 Uganda 0.34 Kenya 0.35 Ethiopia 0.32 Bostwana 0.32 Bostwana 0.30 Cambodia 0.31 Cambodia 0.28 0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 Exporter entry rate Exporter exit rate 5. One-Year Survial Rates by Destination 6.2 Two-Year Survival Rates by Destination SSA 0.51 SSA 0.34 ROW 0.49 ROW 0.26 EU27 0.44 CHN 0.26 CHN 0.43 USA 0.24 USA 0.42 EU27 0.22 0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5 1−Yr. survival rate 2−Yr. survival rate Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graphs 1, 2, 3 and 4 show the following averages: 2008–2012 for Ethiopia, 2007–2011 for Tanzania and Zambia, 2007–2010 for Bo- tswana and Uganda, 2006–2009 for Kenya, and 2007–2009 for Cambodia. 34 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS Ethiopia lags behind peers in Global alignment between pay and productivity (125th) add to Competitiveness Rankings and its performance the deteriorating labor market environment. Ethiopia is on a declining trend. According to the 2013/14 also requires significant improvements in the areas of World Economic Forum Global Competitiveness infrastructure (124th), higher education and training Report, Ethiopia ranks 127th out of 148 countries (137th), and technological readiness (139th). On the in the world. This is behind most of its peers in Sub- other hand, primary education with a net enrollment Saharan Africa such as Rwanda (66th), Zambia (93rd), rate of 87 percent is comparatively good (although Kenya (96th), and Tanzania (125th), but slightly ahead the quality of primary education is very low), and of Uganda (129th).29 Ethiopia’s ranking dropped by women account for a high percentage of the country’s 6 places over the past year (Table 2.1). Figure 2.7.1 labor force. juxtaposes rankings in overall competitiveness with Although the country ranks better than its economy-wide complexity. Ethiopia is conspicuous for peers on theme-specific business regulatory mea- its low rankings on both dimensions. While Uganda is sures such as “enforcing contracts,” overall Doing less competitive, it has a more complex economy; and Business performance is on the decline. In terms of while Zambia is less complex, it ranks much higher the narrower range of regulatory measures affecting in overall competitiveness. the life cycle of small businesses, as measured by the Ethiopia is facing challenges across all pillars World Bank Doing Business 2014 indicators, Ethiopia of the Global Competitiveness Report. The coun- ranks higher than Kenya (129th), Uganda (132nd), and try ranks among the top 100 only for its market size Tanzania (145th) out of 189 economies. Ethiopia’s best (67th), security (55th), and the quality of its institutions scores are for enforcing contracts and dealing with (95th), although the assessment of the latter has been construction permits. falling over recent years. Furthermore, the country’s Of concern for its business and trade perfor- goods (136th) and labor markets (108th) are deterio- mance, Ethiopia lags considerably behind in three rating in the index; the lower performance in these areas: “starting a business,” “protecting investors,” areas is driven by an increase in procedures and time and “trading across borders.” In fact Ethiopia even required to start a business. Increasing concerns about shows a slight deterioration in these three indicators, the quality of labor-employer relations (121st), and the from 162 to 166 over the 2013 to 2014 period in starting a business, from 156 to 157 in protecting investors, and 165 to 166 in trading across borders. TABLE 2.1: Overall Competitiveness Rankings As shown in Table 2.2, of the 10 indicators covered 2013–2014 2012–2013 by Doing Business 2014, there was deterioration in Rank (out of 148) Rank (out of 144) ranking on six, no change in two, and an improvement Ethiopia ↓ 127 121 in two indicators (getting electricity and resolving Kenya 96 106 insolvency). Overall, Ethiopia dropped from 124 to ↑ 125 in the Doing Business Indicators between 2013 Tanzania ↓ 125 120 and 2014. Zambia ↑ 93 102 “Starting a business” measures the number of Rwanda ↓ 66 63 steps an entrepreneur can expect to go through to Uganda ↓ 129 123 Vietnam ↑ 70 75 29 The index takes into account the following 12 components: institu- China ↑ 12 13 tions, infrastructure, macro-economic stability, health and primary education, higher education and training, goods market efficiency, labor Rep. of Korea ↓ 25 19 market efficiency, financial market sophistication, technological readiness, Source: Global Competitiveness Report (GCR). market size, business sophistication, and innovation. Export Performance and Competitiveness 35 launch a business, the time it takes on average, and TABLE 2.2: Doing Business Indicators for the associated cost and minimum capital required Ethiopia to do so. Ethiopia is doing well on the time dimen- 2014 2013 sion, where it takes 15 days to start a business com- Rank (out of 189) Rank (out of 185) pared to 30 days for the Sub-Saharan average. The Starting a Business 166 162 number of procedures is also only slightly higher than Dealing with 55 55 in the SSA average with 9 vs. 8 procedures. But costs Construction Permits are much higher in Ethiopia and drive down the over- Getting Electricity 91 98 all ranking. The cost in percent of income per capita Registering Property 113 107 are 100 percent in Ethiopia vs. 67 percent in SSA; Getting Credit 109 105 likewise, the paid-in minimum capital is 184 percent Protecting Investors 157 156 vs. 126 percent. “Protecting investors” measures three dimen- Paying Taxes 109 103 sions to arrive at an index for the “strength of inves- Trading Across 166 165 Borders tor protection”: “transparency of transactions,” “liability for self-dealing,” and “shareholders’ Enforcing Contracts 44 44 abilities to suit management against misconduct.” Resolving Insolvency 75 77 Ethiopia’s overall index value is 3.3 vs 4.5 in SSA (the Overall Rank (Ease 125 124 of Doing Business) higher the better). The lowest values are in the “trans- Source: Doing Business 2014. parency of transactions” and “shareholders’ abilities to Note: DB 2014 features two new economies: Libya, Myanmar, San sue management against misconduct” where Ethiopia’s Marino, and South Sudan. index shows 3 while the SSA average is at 5. “Trading across borders” looks at import- ing and exporting a container and the number of of governance, including regulatory quality (14 per- documents it takes, the time required and its asso- cent) with the lowest rank among all peer countries. ciated cost. (US$ per container). Ethiopia fares best This rating reflects the relative inability to formulate in the number of required documents dimension and and implement sound policies and regulations that worst in the time dimension. There are 7 documents permit and promote private sector development. required to export (8 in SSA) but a shipment requires More specifically, regulations could be affecting the 44 days to export compared to 31 in the SSA average. dynamism of the private sector by restraining entry Likewise, there are 10 documents required to import and exit of firms subject to undistorted incentives a shipment (9 in SSA), but it takes 44 days compared in the marketplace. This hypothesis is lent credence to 38 days in SSA. In overall cost, Ethiopia is in line by firm-level evidence from Ethiopia, which finds a with the SSA average container cost at US$2,180 per low level of churning in export markets. Compared container to export (US$2,108 in SSA) and US$2,760 to peer countries in Africa, new entrants tend to be per container to import (US$2,793 in SSA). larger and have a higher rate of survival (at least for Low regulatory quality adds to the cost of the first two years). This suggests that the business doing business for exporters. In the World Bank environment is characterized by high cost of entry. World Governance Indicators (WGI) 2012, Ethiopia Encouragingly, Ethiopia scores better than Kenya, scores best on government effectiveness (40 per- Tanzania, Zambia, and Uganda on the “governance cent) and control of corruption (32 percent), where effectiveness” dimension of the WGI. 100 percent performance implies best practice Ethiopia has more trade restrictions in place (Figure 2.7.3). It performs poorly on other aspects than peers—its average Nominal Rate of Protection 36 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS is 25.3 percent. According to the World Trade producers from cheap imports. Ethiopia also contin- Indicators, Ethiopia scores low on the latest MFN ues to rely excessively on tariffs for revenue: among Tariff Trade Restrictiveness Index (TTRI)—109th out large developing countries, it has the highest share of 125 economies. The country’s TTRI is currently of customs and related import duties in tax revenue 13 percent, above average for Sub-Saharan Africa (11.3 (45 percent in 2011 according to World Customs percent) or peers such as Kenya (8.2 percent), Tanzania Organization). (7.8 percent) and Zambia (9 percent). Agricultural In fact, high rates of protection on outputs imports into the country face higher barriers combined with high transport costs change the (16 percent) compared to non-agricultural exports profit incentives for producers by influencing which (12.3 percent), where restrictions are low worldwide. sector to invest in and which markets to serve. The Also of note are textiles and clothing—almost two- Effective Rate of Protection (ERP) quantifies the com- thirds of tariff lines are protected by the maximum bined effects of tariffs (or price distortions) on both tariff of 35 percent plus a surtax of 10 percent. The inputs and outputs. The ERP measures the propor- country’s trade-weighted average MFN applied tariff tion by which an activity’s value-added at domestic is 17.5 percent. (protected) prices differs from that which would be The maximum MFN applied tariff (excl. alcohol the case in a non-protected (or the world) market. and tobacco) is 35 percent, a significant reduction Protecting the domestic market effectively creates a since the peaks of over 200 percent in 1991.30 There bias in favor of domestic producers and discriminates are only six tariff bands. At present, commodities that against exporters through creating an anti-export bias. enter duty-free are fertilizers, railway locomotives, and High levels of Nominal and Effective Rates of aircraft. Most machinery is charged between 5 per- Protection provide strong evidence of the wide- cent and 10 percent duty, although those imported spread existence of an anti-export bias of the tariff for investment purposes can have their duties waived regimes throughout the economy. An average NRP as per the Investment Proclamation. To estimate the on products of 25.3 percent and tariffs of 15 percent price raising impact of import taxation one method (although in many cases they are rebated which will used is to calculate the Nominal Rate of Protection further increase the rate of effective protection) will (NRP), which is the combined rate of ex post duty and result in significant anti-export bias across the likely surtax.31 The average NRP for Ethiopia is 25.3 per- range of technical coefficients (0.3–0.7) from 1.64 cent. The NRPs in agriculture, foodstuffs, textiles and to 2.29 respectively.32 The technical coefficient refers clothing and footwear are all greater than 30 percent. to the relationship between physical inputs (such as A generally high level of inward trade restric- raw materials and intermediate products) and the tion is not helpful to export competitiveness if it physical output-the difference represents value added restricts access to imported inputs at world prices. The model of “enlightened mercantilism” that guided 30 See this link for details: http://info.worldbank.org/etools/wti/docs/ trade negotiations for decades, with exports seen as Ethiopia_taag.pdf 31 In 2007 Ethiopia introduced a surtax of 10 percent to be levied on a desirable and imports as bad, needs now to be dis- large range of imports. Fertilizer, fuel, lubricants, commercial vehicles carded to foster globally competitive private firms. were exempted along with raw materials on which there is zero duty. The surtax aimed to raise revenue to reduce the government deficit. It is Keen to develop its manufacturing sector, Ethiopia levied on all imports on an MFN basis. seeks to maintain relatively high tariffs for finished 32 Sector level or firm level data on the technical coefficients (the value of inputs as a proportion of the value of outputs) is required to derive imports. As part of its accession negotiations at the the Effective Rate of Protection. While this is not available for Ethiopia, WTO, it has sought to bind tariffs at about 1.5 times evidence from the 2011 South African supply table indicates that in labor intensive manufacturing the technical coefficient may range between 0.6 the current applied rate, seeing the wedge between and 0.75 (that is inputs account for 60–75 percent of the value of total bound and applied rates necessary to protect domestic production), and from 0.3 mining to 0.5 for agriculture. Export Performance and Competitiveness 37 (returns to factors of production-generally assumed Uganda (6.3 percent). Indeed, this has helped Ethiopia to be wages and profits. With a technical coefficient attract export-platform FDI to take advantage of the of 0.3 it is 1.30 times more profitable to produce for country’s market access both in the European Union the domestic market than sell all the production in and the United States as a beneficiary of preferences the world market, and when the coefficient is 0.7 the under the African Growth and Opportunity Act anti-export bias increases to 2.29. (AGOA). Almost all of Ethiopia’s goods exports are Recent moves towards seeking membership of eligible for duty-free treatment in the EU and the U.S. COMESA are encouraging. As mentioned earlier, Ethiopia’s rate of utilization of trade prefer- Ethiopia under-trades with several major economies ences is not low, but it is under-tapping a nar- in Sub-Saharan Africa. While intra-regional non-tariff row window of opportunity for diversification. barriers on the part of partners, such as Nigeria, often According to Davies and Nilsson (2013), in 2010, deter trade, Ethiopia has not negotiated preferential Ethiopia utilized about 94 percent of preferences in trade arrangements with regional trading partners, the EU, a rate lower than Kenya and Ghana. In the thereby facing tariffs in exports to nearby countries U.S., Ethiopia’s preference utilization rate was 84 like Kenya, Uganda, and Tanzania. Indeed, in Africa, percent, and much lower than that of Bangladesh, it has a free trade agreement (with zero duties on all Kenya, Mozambique and Uganda. Ethiopia also goods) only with Sudan. It is one of the six countries enjoys tariff preferences under the Generalized System (out of 19) that are not yet a Preferential Trade Area of Preferences (GSP) in Australia, Canada, Japan, New (PTA) member within COMESA, and the country Zealand, Norway, and Switzerland. It also has pref- has doubts about the application of the rules of ori- erential access in China, India, Russia, South Korea, gin. While there have been some problems within and Turkey, but not Brazil or South Africa. Ethiopia COMESA on rules of origin,33 they are still deemed to is, therefore, under-performing not so much in the be much simpler (a 35 percent value-added rule) than sense of current utilization rates of trade preferences, product and process-specific rules under the Southern but in terms of the potential volume of exports it could Africa Development Community (SADC). Weak be selling duty-free. Ethiopia is exporting less than regional integration of countries like Ethiopia limits US$3 billion worth of merchandise when an identi- the country’s potentials for growth and diversifica- cally populous Vietnam is exporting US$120 billion tion because there exist large trading opportunities while facing higher tariffs both in the EU and U.S. within Africa in food products, basic manufactures, As has been argued by Collier and Venables (2007), and services (Brenton and Isik 2012). Furthermore, trade preferences can be a catalyst for diversified the cross-border production networks of the kind that manufactured exports and FDI.  However, this will have emerged in East Asia are largely absent in East likely only be available for a narrow window of time Africa. Regional markets tend to thrive particularly because preferences erode over time. when they become investment platforms for indus- tries that can participate vigorously in world markets. 33 As detailed in World Bank (2011), not all countries have adopted the simple rules of origin in COMESA. Egypt unilaterally imposes a 45 Ethiopia enjoys generous market access to the percent local content rule. Until recently, Zambia, Uganda, and Malawi world’s largest economies, which is an important did the same. Rules in two sectors have also proven to be particularly contentious under COMESA. For wheat flour, the 35 percent value competitiveness advantage. In terms of access to for- added rule has generated difficulties for exporters in Egypt and Mauritius eign markets, based on its Market Access TTRI includ- that do not produce wheat grain, but import the raw material from the world market. In periods of high wheat prices, such as those experienced ing preferences of 1.8 percent, the country enjoys recently, this meant that these countries were unable to meet the value more favorable access to international markets than added requirement. With palm oil, there have been disputes over refined oils (e.g. Zambia-Kenya) because of difficulties assigning value added. its SSA peers (average: 3.9 percent), and, in particular These have arisen because a number of products can be produced from Tanzania (5.2 percent), Rwanda (8.3 percent), and the raw material such as cooking oil, soap and margarine. 38 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 2.7: Competitiveness and Economic Complexity 1. Competitiveness and Economic Complexity 2. Indicators of Market Efficiency (country ranking) 50 110 Kenya 60 KOR 35 37 90 CHN Zambia 39 48 41 70 Uganda 125 67 VNM 79 50 KEN Botswana 97 93 UGA 30 89 TZA Tanzania 114 ZMB 127 ETH 10 131 Ethiopia 144 133 10 30 50 70 90 110 0 20 40 60 80 100 120 140 160 Rank in global competitiveness 2013–2014 Anti-monopoly policy Market dominance Local competition 3. Governance Indicators (percentiles) 46 Zambia 36 38 18 Uganda 44 33 22 Tanzania 37 28 12 Kenya 42 35 32 Ethiopia 14 40 79 Botswana 74 67 0 20 40 60 80 100 Control of corruption Regulatory quality Government effectiveness Source: (1) UN Contrade. (2) Global Competitiveness Report (2013/14). (3) World Governance Indicators (2012). Factor Inputs, Productivity, and Trade Costs The pace of regulatory reforms aimed at stream- lining procedures, and lowering time and cost of engaging in trade is slow. Among the ten World Bank Time and cost for companies to engage in trade is long and high in Ethiopia, yet, the pace of regulatory reforms aimed Doing Business indicators, Ethiopia fares the worst at streamlining procedures is slow. Being landlocked cannot in Trading Across Borders, ranking 166 out of 189 be the sole cause for bad logistics performance, as Rwanda economies (Table 2.2). Ethiopia not only ranks behind has demonstrated, and Ethiopia is currently undertaking some crucial investments to improve trade logistics in the Kenya and Rwanda but also Uganda, Zambia, and medium-term. But more generally, access to finance, land, Tanzania in the procedural aspects of trade as judged and electricity are some of the most binding constraints that by mid-sized Ethiopian exporters. Vietnam, Ethiopia’s urgently need to be addressed. With this background it is no surprise that while wages are low productivity is low, too. peer country in East Asia, in fact ranks 100 slots ahead of Ethiopia. There has been no major improvement Export Performance and Competitiveness 39 TABLE 2.3: Time and Cost of Exporting and Importing   Year Ease of Trading No. of Number Cost to No. of Number Cost to Doing Across documents of days to export documents of days to import Business Borders to export export (US$ per to import import (US$ per (Overall) Ranking container) container) Ranking Ethiopia 2010 ... ... 7 50 2,230 10 44 2,660 Ethiopia 2011 ... ... 7 45 2,180 10 44 2,660 Ethiopia 2012 ... ... 7 44 2,180 10 44 2,660 Ethiopia 2013 124 165 7 44  2,180 10 44 2,660 Ethiopia 2014 125 166 7 44 2,180 10 44 2,760 Rwanda 2010 ... ... 8 38 3,275 10 35 4,990 Rwanda 2011 ... ... 7 35 3,275 9 34 4,990 Rwanda 2012 ... ... 7 29 3,275 9 31 4,990 Rwanda 2013 54 160 7 29 3,245 9 31 4,990 Rwanda 2014 32 162 7 26 3,245 9 30 4,990 Source: Doing Business (DB). Note: Methodology revised in 2014, which has been applied to the rankings in 2013 & 2014. in this trade-related indicator over the past five years 50 to 44 even though it can be argued that it is easier in Ethiopia (Table 2.3).34 to reduce the time and costs when they are highly Being landlocked cannot be the sole cause for inefficient to begin with. bad logistics performance, as Rwanda has demon- Ethiopia is currently undertaking some cru- strated. In 2010, Rwanda’s ranking on the Trading cial investments to improve trade logistics in the Across Borders indicator was lower than Ethiopia’s, medium-term. With several new public investments and its shipping costs even today are significantly in roads, a rehabilitated rail link between Addis larger than those faced by Ethiopia. Over the past five Ababa and the rapidly modernizing container port years, Rwanda has made substantial improvements of Djibouti, the expansion of the dry port in Modjo, in areas in which it has direct policy control: i) it expanded coverage of the multi-modal transport sys- improved trading times with administrative changes tem and coordinated reforms between customs and such as increased operating hours and enhanced shipping-related agencies, trading is expected to be cooperation at the border; ii) it reduced the number of trade documents required and enhanced its joint border management procedures with Uganda and 34 Indeed, Ethiopia’s relative ranking has fallen from 157th in 2011 (out other neighbors, and iii) it introduced an electronic of 183 economies) to 166th in 2014 (out of 189 economies), even though the rankings cannot be directly compared to previous years (except in single-window system at the border. The impact of 2013 when it ranked 165th) because the methodology for calculating these reforms has been to reduce the number of days the Trading Across Borders indicators was changed in 2014. “Trading Across Borders” indicator was revised in 2014 to reflect the fact that it takes to export from 38 to 26 (an improvement of documents (such as the certificate of origin) required to claim preferential 32 percent). In Ethiopia, the average number of days tariffs are no longer counted. This, however, does not alter the number of documents to process an average export shipment out of Ethiopia, to export dropped by only 6 days (12 percent) from which remains unchanged at 7 for the past 5 years. 40 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS TABLE 2.4: Major Constraints Faced by Ethiopian Companies, Exporters vs. Non-Exporters   All firms Non-exporters Exporters Broad constraints Percent of firms identifying access to finance as a major constraint 31 32 21 Percent of firms identifying electricity as a major constraint 23 22 38 Percent of firms identifying an inadequately educated workforce as a major constraint 4 4 2 Infrastructure constraints Losses due to electrical outages (% of annual sales) 2.6 2.6 3.7 Number of water insufficiencies in a typical month 2.8 2.9 2.5 Percent of firms identifying transportation as a major constraint 10.2 10.1 12.5 Business processes and innovation Percent of firms with an internationally-recognized quality certification 14 12 48 Percent of firms using technology licensed from foreign companies 43 40 67 Percent of firms having their own website 43 42 69 Percent of firms using e-mail to interact with clients/suppliers 80 79 99 Percent of firms with an annual financial statement reviewed by external auditors 72 72 80 Source: Enterprise survey 2011–12 simplified with costs and dwell time reduced. This But there is also a noticeable discrepancy in the will ultimately help increase farm-gate prices relative manner in which exporters35 and non-exporters to exports, which at present is only about 60 percent feel burdened. Among exporters, a substantially (IMF 2013b). larger share (38 percent), compared to 23 percent of More generally, access to finance, land, and all firms, find that electricity is a major constraint. electricity are some of the most binding constraints Losses due to electrical outages are also a full percent- faced by Ethiopian firms. While these issues are also age point higher for exporters than non-exporters. common in peer countries, in the 2011–12 Enterprise However, exporters appear to have a privileged access Survey, between one-third and two-fifths of Ethiopian to finance, with only one-fifth of exporting firms firms cited access to electricity, land, and finance as reporting that to be a constraint, in contrast to nearly major hurdles to growing their businesses. In con- one-third of all firms. On transport, nearly 90 percent trast, less than 5 percent of the firms surveyed men- of the firms do not view it as a major constraint. This tioned lack of skilled workers as a major constraint ratio is higher than in any other peer country except (Table 2.4). Perceptions vary across firm size, however. China. However, a slightly higher share of Ethiopian The largest firms with more than 100 employees also exporters cites transportation to be a constraint than cite customs and trade-related regulations as one of the non-exporters (Table 2.4). top constraints. Also for the largest firms, electricity is Ethiopia is investing to produce a skilled, yet the biggest hurdle, even though the average loss due to affordable labor force. There is significant expan- electrical outages (in percent of annual sales) is less in sion occurring in university-trained personnel. The Ethiopia than in Kenya, Tanzania, and Uganda. Water shortages (number of outages per month faced by manufacturing firms) are less severe than in Tanzania 35 Exporters are defined in the Enterprise surveys as firms whose direct and Kenya, but still high compared to Vietnam. exports are 10 percent or more of sales. Export Performance and Competitiveness 41 number of universities has increased from just two Proactive Policies to Support Trade about 15 years ago to 32 at present. There is some alignment between the future industrial aspirations of Product differentiation and branding offer substantial the country and the kind of graduates it is producing; potential in Ethiopia. New unexplored instruments for product differentiation, such as Geographical Indications, for example, horticulture degrees are being taught in have been a potent source of premium income and quality six universities. Ethiopian Airlines has its own Aviation enhancement not only in Europe (think Champagne or Academy with plans to increase the number of trainees Parma ham), but also countries like India (Basmati rice and Darjeeling tea) and Vietnam (Phu Quoc fish sauce). from about 200 per year to 1000 (and more). A key Ethiopia is a latecomer in utilizing another instrument to effort of the government in this regard is the estab- promote trade, i.e. the creation of Industrial Zones (IZs). lishment of a Technical and Vocational Education The country’s experience so far with the development of zones has not been promising but it is committed to learning and Training (TVETs) system. The TVET system from global good practices. In fact, the government seeks in Ethiopia follows the German model and empha- to expand its existing IZ program to support its target for sizes ‘apprenticeship’. TVET colleges are required to growth and job creation. identify potential employees for their students; 70 percent of the entire program duration is to be spent Ethiopian regimes for standards and certification apprenticing with an employer. Ethiopia’s TVET sys- are inadequate, but in line with that of its peers. tem is transitioning to a competency based approach. According to the World Bank Enterprise Surveys, To graduate, TVET trainees are now required to pass almost 14 percent of firms have earned quality cer- assessment tests which are based on Occupational tification that is recognized by global bodies like Standards (OS). the International Standards Organization (ISO), a Ethiopian wages are relatively competitive proportion higher than in Kenya (10 percent) and not only against Chinese or Vietnamese, but Rwanda (12 percent). In terms of auditing/financial also Kenyan averages. A recent study found that standards, Ethiopia does even better. The scale of Ethiopia has comparable labor cost and productivity adoption of modern production and business pro- to Bangladesh and that Ethiopia is one of the very few cesses by Ethiopian firms also appear to be better African countries where its low income level is likely than what the per capita figures suggest. Forty-three to translate into a comparative advantage in low-wage percent of firms in Ethiopia have their own website basic manufacturing (Gelb et al. 2013). According to and use technology licensed from a foreign company, Chinese investors in the country, the average wage at the highest proportion among its peers. In fact, split- a plant outside the capital is about one-sixth of that ting the responses by exporting and non-exporting in coastal China (500 RMB in Ethiopia versus 3000 firms, the former appear to be far ahead of the latter in RMB in China, on average). This will clearly be one adopting modern business processes: the share of firms of the main investment attractors from countries with internationally-recognized quality certification is like China and Turkey where wages are rising. But four times larger among exporters than non-exporters; export-oriented firms look at many other factors similarly almost all exporting firms use email to inter- beyond wage. They prefer business-friendly regula- act with clients, and more than two-thirds of exporters tions, in terms of costs and procedures, to govern use some kind of technology licensed from foreign hiring and redundancy practices. While there is a fine companies (Table 2.4). line between protecting workers’ rights and ensuring Stakeholder interviews in major export sectors firm efficiency, onerous regulations hamper efforts to (pulses, flowers, coffee) revealed industry specific restructure the firm, hurting overall opportunities for challenges. The current challenge in the area of stan- employment in the formal sector and movement of dards is primarily the weak capacity of regulatory workers across jobs. bodies to: i) help exporters maintain and upgrade 42 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS quality in general (e.g., coffee, leather) and ii) redress infrastructure than those prevailing in the rest of deficiencies where inadequate quality standards are the economy. They are essentially “second-best” directly hurting export potentials. For instance on institutions designed to relax constraints when the reason that an overwhelming share of pulses (such as ideal solution of undertaking wider reforms nation- groundnuts) enter emerging markets, but not markets ally is not possible for political or financial reasons. like the EU or the U.S.is the inability of exporters to Emerging manufacturing firms, in particular, could fulfill the high sanitary requirements of those markets. benefit from duty-free inputs imported with less The EU only permits aflatoxin contamination of two hassle, lax labor laws, and predictable access to qual- parts per billion (ppb) for edible groundnuts. The ity infrastructure. The performance of most zones, flower industry, too, has worked hard to implement however, is uneven. While some have played a trans- a Code of Practice for social responsibility grading formative role, especially in East Asia in the 1980s firms into categories of bronze and silver based on and the 1990s, many fail to live up to initial promises certification levels. It is moving towards implement- and some end up being wasteful misadventures. ing Integrated Pest Management in lieu of chemicals. The government seeks to expand its existing Beyond sanitary issues, the origins of produce are IZ program to support its target for growth and not easily traceable. About 50 percent of coffee, for job creation. The Government has accordingly iden- example, is domestically sold, sometimes at prices tified 5 new potential IZ sites that have a strategic higher than in international markets. With incentives interest for the country.37 The Ethiopian Industrial distorted by tight controls, intermediaries often resort Development Zone Corporation (EIDZC) has to hoarding, misleading classification on the quality recently been established which, among other things, and grade of coffee, and illicit trading, all of which shall develop and manage Industrial Zones, lease weaken a transparent system of quality control based developed land, and outsource through manage- on verification and traceability. ment contracts administration of industrial zones. Product differentiation and branding offer In addition, the One Stop Shop (OSS) regulations substantial potential. New unexplored instruments have been approved by the Council of Ministers for product differentiation, such as Geographical and are now operationalized within the Ethiopian Indications,36 have been a potent source of premium Investment Agency (EIA), which will facilitate 28 income and quality enhancement not only in Europe out of the 29 procedures investors may need to go (think Champagne or Parma ham), but also coun- through. Numerous development partners, especially tries like India (Basmati rice and Darjeeling tea) and the World Bank Group, are involved directly or indi- Vietnam (Phu Quoc fish sauce). Ethiopia ought to pay rectly in the IZ program. In addition to government much bigger attention to establishing a functioning initiatives, private players are also floating big ideas. Geographical Indications regime to brand its quality Huajian, the successful shoe factory is now planning a products differently, such as coffee from Yirgachefe, Harar, and Sidamo; or Humera sesame seeds. The present system of inadequate grading and sorting, and 36 The WTO Agreement on Trade-Related Aspects of Intellectual Property (TRIPS) defines Geographical Indications as indications (words, phrases, a tendency to export in bulk undervalues Ethiopia’s symbols, images), which identify a good as originating in the territory of most important exports. a member, or region or locality in that territory, where a given quality, reputation, or other characteristic of the good is essentially attributable to Ethiopia can learn from the lessons of its geographical origin. See Waglé (2007) for details on the topic of GIs. Industrial Zones (IZs) or Special Economic Zones 37 They are as follows: Bole Lemi (Addis Ababa Charter city), Kilinto (Addis Ababa Charter city), Dire Dawa IZ (Dire Dawa Charter city), (SEZs) in the rest of Africa. In principle, SEZs offer Kombolcha IZ (Amhara Region) and Hawassa IZ (Southern Nations, different trade policies, regulations, and quality of Nationalities and Peoples Region) Export Performance and Competitiveness 43 major expansion in the country having acquired 317 growth goes up by 0.88 percentage points a year. In the hectares of land to establish the “Ethio-China Light case of Ethiopia, a 10 percent undervaluation would Manufacturing Special Economic Zone,” which is potentially boost exports by 5.2 percentage points and planned to employ 100,000 people. Signaling that it economic growth by 2.2 percentage points. is in for the long haul, the company has taken more Empirical results are used to explore the ques- than two hundred Ethiopians to train in Chinese tion whether an undervalued exchange rate can shoe factories.38 Greater emphasis by the govern- help boost export and therefore also output growth. ment to locate Ethiopian SMEs in these zones would The issue needs to be addressed from an empirical per- augment competitiveness in the long run through spective since the theoretical relationship between the intra-industry knowledge spillovers and formation of real exchange rate and exports/outputs is not clear-cut clusters. SEZs could reduce start-up costs and risks (Annex 5). The analysis in this report utilizes a cross- for SMEs by taking advantage of larger facilities at a country RER misalignment index. A country’s RER phase in their development when they are unable to is defined as the relative price of the domestic con- obtain bank loans (Dinh et al. 2012). sumption basket and the foreign consumption basket. The domestic consumption basket includes domestic Competing on Price: Does the Real non-tradable goods, domestic tradable goods, and Exchange Rate Matter? some foreign tradable goods; the foreign consumption basket includes foreign non-tradable goods, foreign A more competitive real exchange rate could support tradable goods, and some domestic tradable goods. export promotion. Ethiopia’s real exchange rate is The analysis uses a simple theory-based overvalued. Empirical evidence presented here suggests approach first developed by Rodrik (2008) to cal- that a 10 percent lower real exchange rate could increase export growth in Ethiopia by more than 5 percentage culate the RER misalignment. In his work, Rodrik points per year and increase economic growth by more has shown that undervalued real exchange rates are than 2 percentage points. The potentially strong impact for associated with higher output growth. Since the origi- Ethiopia is a reflection of the predominance of basic export commodities that tend to compete more on price rather nal work does not include export growth this report than on quality. Given a number of macro-economic trade- will slightly modify Rodrik’s approach to measure offs — e.g. devaluation has a tendency to increase import RER misalignment and will present evidence about prices and thus contribute to inflation — any changes in the exchange rate may need to be accompanied by adjustments the relationship between undervaluation and export in the macroeconomic policy mix. growth. Annex 5 provides the details on how the study measures the RER misalignment index. The key This section utilizes a simple theory-based real is to establish the RER misalignment index through exchange rate (RER) Misalignment Index for controlling for the Balassa-Samuelson effect. Balassa- countries around the world from 1950–2011, and Samuelson captures the effect of an economy’s produc- shows that Ethiopia’s RER has been overvalued. tivity on its non-tradable goods’ prices. The empirical Overvaluation was about 31 percent in 2010 and results show that the Balassa-Samuelson effect is highly 2011. The section shows that an undervalued RER significant with a negative sign. is associated with higher real export and output growth, especially for developing countries today and 38 Although perhaps not quite on the same scale, this is reminiscent of another initiative from history that launched the apparel sector in developed countries in the earlier decades. Across all Bangladesh: in 1978, the Desh Company signed a five-year agreement countries and time, on average, for each additional with Daewoo, a Korean multinational, which trained Desh employees in production and marketing in Korea. Within a year, 115 of the 130 10 percent RER undervaluation, the country’s export trainees had left Desh to start their own garment export firms (Kabeer growth goes up by 0.6 percentage points and its output and Mahmud 2004). 44 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS There are other approaches to calculating RERs coefficients are significant for high-income countries and to determine any undervaluation. Currently, the in the earlier decades (1950–1980) and for low- most popular one is to regress a country’s real exchange income countries in the latter decades (1981–2011). rate against a large set of the country’s fundamentals This implies that countries need to be at a certain to establish a real exchange rate norm.39 The gap level of income to be able to take advantage of the between a country’s actual real exchange rate and its undervalued RER strategy. This is likely to be when norm (i.e. the residual in the regression) is considered a country is at the stage of exporting relatively simple, the “misaligned” part. The most well-known research light manufacturing products. In turn, this means using this approach is from the IMF (Lee et al. 2006), that if a country’s economy is dominated by agricul- which forms the basis for the IMF’s work on assess- ture production and/or only exports commodities, ing countries’ RER misalignment in its Article IV an undervalued RER may not help the exporting papers. Tensay (2006) also uses this approach to study firms. On the other hand, when a country is already Ethiopia’s RER misalignment. This report does not rich enough and its firms already operate at the tech- rely on the methodology given the complexity of the nological frontier, an undervalued RER might not approach and the difficulty to identify “fundamentals.” matter anymore for the country’s export. Rather, at Instead, Rodrik’s amended methodology is found to the frontier the country’s firms need to rely on better be more intuitive. technology and innovation. Ethiopia’s RER has been consistently over- On average, across all countries and the full valued for the past two decades, which is in stark time period considered (1950–2011), for each addi- contrast to Asian experiences but in line with Latin tional 10 percent RER undervaluation a country’s American experiences. Plotting the RER misalign- export growth will rise by 0.6 percentage points per ment index against export growth for selected coun- year. Among high-income countries between 1950 tries shows that Ethiopia’s RER was highly overvalued and 1980, an additional 10 percent undervaluation for the past two decades (Figure 2.8.1). This is in stark boosted export growth by 1.26 percentage points. contrast to the experiences in Asian countries. Figures Among low-income countries between 1981 and 2.8.2, 2.8.3 and 2.8.4 show China, India, and Korea. 2011, that figure is 0.7 percent. A similar exercise All of them had undervalued exchange rates during for countries’ real output growth shows a significant their catching-up periods: China between 1981 and positive impact of an undervalued exchanged rate on 2009, India since 1999 until now, and South Korea real output growth. The positive impact holds for both during much of the 1960s and 70s. On the other hand, high income and low income countries: on average Latin American countries adopted a rather overvalued across all countries and all time, if a country’s RER is strategy—like Ethiopia—where RERs were overvalued 10 percent undervalued its real output growth goes during most of the past decades (see Figures 2.8.5 and up by 0.88 percentage points per year. 2.8.6 for Argentina and Brazil). A regression analysis more definitely confirms 39 In the so-called “kitchen sink” approach, researchers throw believed-to- be fundamental variables to the right hand side of the regression — often the graphical observations of the undervaluation without a clear theoretical rationale to why they are fundamental — and index and export growth plotted in Figure 2.8. hope to find some significance. There are two problems with this. First, there may be neglected fundamentals that also affect the real exchange Annex 5 provides details about the regression and its rate, but are not included. Negligence may come from the fact that it is results, which are very intuitive. Overall, undervalu- virtually impossible to come up with an exhaustive list of factors affecting productivity and consumption and saving decisions. Second, variables ation is associated with higher export growth; this considered “fundamentals” might actually contain elements that distort phenomenon is true for both high-income and low- the real exchange rate. For example, government consumption is con- sidered a “fundamental.” However, government consumption could be income countries (defined as having an annual GDP directly affected by an incentive to lower the real exchange rate. Eden and per capita below $6,000 in 2000). Interestingly, the Nguyen (2012) offer more detailed criticism of the current approaches. –1.5 –1.0 –0.5 0 0.5 1.0 –0.7 –0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 0.3 –1.2 –1.0 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 1951 1951 1951 1953 1953 1953 1955 1955 1955 1957 1957 1957 1959 1959 1959 1961 1961 1961 1963 1963 1963 1965 1965 1965 1967 1967 1967 1969 1969 1969 1971 1971 1971 1973 1973 1973 1975 1975 1975 1977 1977 1977 1979 1979 1979 Note: Undervaluation zone is above 0. 1981 1981 1981 Ethiopia 1983 1983 1983 India 1985 1985 1985 Argentine 1987 1987 1987 1989 1989 1989 1991 1991 1991 1993 1993 1993 1995 1995 1995 1997 1997 1997 1999 1999 1999 2001 2001 2001 2003 2003 2003 2005 2005 2005 2007 2007 2007 Undervaluation 2009 2009 2009 2011 2011 2011 –1.2 –1.0 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 –1.0 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 0.8 –0.8 –0.6 –0.4 –0.2 0 0.2 0.4 0.6 1951 1953 1953 Source: World Bank staff own calculations, based on data from PENN World Tables. 1953 1955 1955 1955 1957 1957 1957 1959 1959 1959 1961 1961 1961 1963 1963 1963 1965 1965 Export growth 1965 1967 1967 1967 1969 1969 1969 1971 1971 1971 1973 1973 1973 1975 1975 1975 1977 1977 1977 1979 1979 1979 1981 1981 1981 Brazil China 1983 1983 1983 1985 1985 South Korea 1985 1987 1987 1987 1989 1989 1989 1991 1991 FIGURE 2.8: Undervaluation and Export Growth, Ethiopia and Selected Countries 1991 1993 1993 1993 1995 1995 1995 1997 1997 1997 1999 1999 1999 2001 2001 2001 2003 2003 2003 2005 2005 2005 2007 2007 2007 Export Performance and Competitiveness 2009 2009 2009 2011 2011 2011 45 46 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS Patterns identified in the undervaluation index export and output growth is strongly associated with are consistent with other estimates. For instance, the and highly sensitive to the change of the country’s IMF Article IV in 2012 estimated that the Ethiopia RER misalignment. real exchange rate was about 11 to 23 percent over- The high negative impact of RER appreciation valued in May 2012 (IMF 2012). In 2013, the IMF implies that Ethiopia’s export comprises mainly estimated that the Real Effective Exchange Rate low-value products that compete on prices and (REER) was at least 10 percent overvalued. The differ- less of high quality products. In theory, the appre- ences in results arise because of differences in method- ciation of the Ethiopian Birr hurts exports because ology (“kitchen sink” vs. Balassa Samuelson effect) and they are now less competitively priced than identical real exchange rate concepts. The IMF uses the REER, products from other countries, all else being equal. which is the Real Exchange Rate of a country relative In practice, the relationship between competitive- to a set of major countries.40 This report uses the Real ness and movements in the real exchange rate is not Exchange Rate relative to the United States Dollar. straightforward. The latter can appreciate as a result In the case of Ethiopia, each additional of an improvement in competitiveness when there are 10 percent RER undervaluation is associated with gains in productivity of tradable goods relative to that higher export growth of up to 5.2 percentage points of non-tradable goods. Competitiveness is lost when per year and higher real GDP growth of up to there is a misalignment from the equilibrium RER. 2.2 percentage points (Table 2.5). In terms of abso- In particular, the agricultural commodities that are lute impact in 2010 and 2011—where the RER was the mainstay of Ethiopian exports tend to be affected earlier estimated to be 31 percent overvalued—this by real appreciation because import-content in these finding would indicate (upper-bound) lower export sectors is generally lower than in manufacturing. growth to the order of 16 percent (31*0.52) and Potential changes in the exchange rate would 6.8 percent lower real output growth, compared to an need to be considered in the context of the overall equilibrium situation. The large magnitudes of these macroeconomic policy mix, which may also need upper-bound estimated impacts suggest that Ethiopia’s adjustment. There are macro-economic trade-offs to consider. Nominal currency depreciation increases import prices and thus contributes to inflation. Higher TABLE 2.5: Effects of Undervalued RERs on cost of imports of capital equipment may make public Export and Output Growth Ethiopian data investment more expensive. There are also balance   Δln(real exports) Δln(real GDP) sheet effects through a rise in external public debt (1) (2) when expressed in local currency. Some of these trade- ln(Initial Real Exports) –0.203** offs can be addressed by adjusting other policies. A (0.076) tighter monetary and/or fiscal policy can help contain ln(Initial Real GDP per –0.104** capita) (0.044) the inflationary impact, for instance. The analysis is silent on the equilibrium effect Undervaluation 0.523** 0.226*** (0.233) (0.064) and notes that, from an international perspective, Constant 1.674*** 0.807** competitive devaluations from many countries are (0.577) (0.307) not optimal. If every country devalues to take advantage R-squared 0.151 0.340 Adjusted R-squared 0.122 0.317 40 REER formally is the weighted average of a country’s currency relative S.E. of regression 61 61 to an index or basket of other major currencies adjusted for the effects Note: Newey-West HAC Standard Errors in parenthesis. of inflation. The weights are determined by comparing the relative trade *** p<0.01, ** p<0.05, * p<0.1 balances with each other country within the index. Export Performance and Competitiveness 47 of the lower RER, countries might undercut each other’s view of the developing world as a whole, competitive export (i.e., a race to the bottom), and the end result devaluation (recently dubbed as currency wars) from might be that no exporting country will benefit from many developing countries may not be optimal. An their exchange rate devaluations. In other words, from international exchange rate coordination system could the point of view of a specific country such as Ethiopia, play a role in this regard, but this is beyond the scope maintaining an undervalued exchange rate is beneficial of this paper. In the long run what really matters for for its export and growth. However from the point of exports to strive is productivity and product quality. 49 SUMMARY AND POLICY RECOMMENDATIONS 3 To support economic growth over the past decade, largely unprocessed and undifferentiated bulk exports Ethiopia has successfully leveraged agriculture in existing agro-exports towards value-addition on a exports to advanced countries, but the challenge commercial scale. At present all of Ethiopia’s leading now is to better link it to processes of quality-addi- exports rank in the bottom half of product complexity tion and ultimately industrial scale value-addition. and sophistication, yet with the removal of supply-side The pace of economic growth over the past decade, hurdles and basic additional steps of processing, there together with its population of over 90 million, give is potential to increase the quantity and earn more per a hint about its future potential. It has leveraged agri- unit. The second avenue is to ramp up capabilities in culture exports to advanced countries, has a booming manufacturing in new ways particularly by making air travel export, and is generating significant inter- greater use of modern, competitive services as inter- est from emerging economies like China, India, and mediate inputs. Turkey. Despite being landlocked, its preferential But Ethiopia is not utilizing its “advantage of market access in the world’s largest economies and rela- backwardness” for export growth to the full extent, tive proximity to Europe and the Middle East provide losing out on productivity increases and struc- opportunities for economic diversification. To sustain tural change. For a developing country of its size, its transformation, Ethiopia needs to build on its agri- Ethiopia’s goods exports are the lowest in the world cultural foundation by adding quality to commodity and its manufacturing value-added in the economy is exports and eventually industrial scale value-addition. almost three times less than the average for SSA. Yet, It will also need new tradable activities in manufactur- Ethiopia’s potential is vast. The country has the land ing and services that have the ability to absorb large for pastures and cultivation—both to support one of numbers of young semi-skilled workers associated with the world’s largest counts of livestock and to grow cash a shift away from agriculture. Redressing supply-side crops, from coffee and cotton to flowers and oilseeds. constraints aggravated by indifferent business and It will have a growing middle class at home, and has investment incentives can facilitate this process. preferential market access in rich countries abroad. Ethiopia’s process of structural transformation With the expansion of higher education, it will have has the “advantage of backwardness.” Starting from an educated workforce that is trainable at affordable a low base of per capita exports and investment, it wages. It will also soon have abundant electrical power, can avoid mistakes made by peers in the past in some and more efficient access to a container port, through areas and leapfrog in others, rapidly adapting to the a rehabilitated railway and improved trade facilitation. changing modes of trade and production in the 21st In addition, it sits poised to attract labor-intensive century. It will most likely not duplicate the develop- investments from emerging countries that are losing ment path that East Asia took, but achieving the kind cost-competitiveness. of export success that comparably populous countries This report showed that rising exports contrib- like Thailand and Vietnam have had is possible. Like uted to Ethiopia’s remarkable growth performance them, Ethiopia has two clear channels for expanding over the past decade but that a recent drop in export merchandise exports. The first is to move away from prices has exposed underlying vulnerabilities in 50 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS export structure. This highlighted the importance of an improved certification regime could be dovetailed strengthening competitiveness, key in which it is to to establish a functioning regime of registering and realize that more than “what” is being exported it is enforcing Geographical Indications, a form of intel- the “how” that is hindering potential. In fact Ethiopia’s lectual property, which could help brand the country export sector is currently too small to contribute to and facilitate its unique exports to earn premium. structural transformation. The business environment Second, ease binding constraints related to reli- favors incumbent firms and deters new entrants into able power supply, credit, and foreign exchange. export business, and even so, no “export superstars” Among exporters, 38 percent find that electricity are emerging. is a major constraint compared to 23 percent of all To unleash the potential a policy and insti- firms. Losses due to electrical outages of 3.7 percent tutional framework is needed that is constantly of annual sales are also a full percentage point higher adjusting to be able to provide the right incen- for exporters than non-exporters. Even though export- tives to entrepreneurs at home and investors from ers have better access to credit than non-exporters, abroad. There is a strong Ethiopian developmental up to one fifth of them still report this as a major state that can help the transformation if its policy constraint. Finally, firms frequently mention access leverage is utilized wisely. For instance, most recently, to foreign exchange as a major constraint in doing it played a transformative role in creating a success- business in Ethiopia. ful horticulture export sector by providing incentives Third, redress bottlenecks in trade logistics. A in land, fuel, and freight to early investors. However, key to competitiveness is shipping containers quickly drawing on the experience of countries that have and inexpensively. Rwanda, which faces more crip- graduated from low-income status, Ethiopia’s future pling shipping costs, performs better than Ethiopia growth can only be sustained by a dynamic private in overall trade-related operations because of its sector that increasingly takes up space released by the reforms in operating hours, joint border manage- public sector. ment procedures with neighbors, and introduction The following policy recommendations aim to of an electronic single-window system. Effective inform policy makers on how to develop a more implementation will require significant coordination competitive trade and business environment in across government departments to avoid having a re- Ethiopia. The analysis reveals seven areas of policy engineering effort simply lead to the accumulation focus that would support the objective of reducing of inefficiencies in one place. Consultations with the vulnerability of exports to price fluctuations, scale-up private sector are also crucial and could be entertained the size of exports, support structural transformation by a permanent information campaign (including a through higher-productivity exports, and promote a website) to transparently inform the private sector more dynamic export business environment. Actions on what is necessary for trading or establishing busi- to be considered could be the following: nesses, the exact documentation requirements, and First, increase value-addition, quality, and the steps involved. branding of exports. Existing exports have great Fourth, establish Industrial Zones that con- potential for augmenting their value per unit with bet- form to international best practice. This is one area ter processing, packaging, testing, and general applica- where Ethiopia’s laggardly status gives it an advantage. tion of international standards. Fostering adherence In designing new policy, regulatory and institutional to international product standards and certification framework for managing industrial zones, which are regimes would help enforce quality control based on at a very early stage, Ethiopia could avoid the mis- verification, and traceability on the origins of pro- takes of many SEZs in the rest of Africa. There could duce would help brand Ethiopian exports. Indeed, be exceptions, but in general, this implies adopting a Summary and Policy Recommendations 51 modern best practice that emphasizes quality infra- Sixth, improve regulatory quality, including the structure and sound business environment, rather implementation of a pro-competition legal frame- than fiscal incentives (e.g., tax holidays). The World work. Intensifying local competition and reducing Bank is actively supporting the Government in this market domination by individual companies could regard with a forthcoming large-scale IDA-financed be an effective means to improving the business envi- project that will bring good practices from other ronment. Effective competition policies are associated countries. with private sector growth. Indeed, the best practice Fifth, revise burdensome business rules that is not just to engage in anti-trust investigations, but obstruct firm entry, especially high start-up capital also to cover legal enforcement, competition advocacy requirement and pre-registration bank deposits. and invest in institutional effectiveness. Ethiopia’s worst ranking in the 2014 Doing Business Seventh, ensure that the real exchange rate is indicators is for “starting a business.” But, within this competitive. Empirical analysis of country experi- theme, it is the high paid-in capital requirement as ences suggests that a competitive exchange rate is well as the cost of pre-registration bank deposits that associated with rapid economic growth, often led stand out. Reducing start-up capital for enterprises by exports. In Ethiopia, the real exchange rate is may have an immediate impact on facilitating greater overvalued, which favors imports but hurts exports. firm entry into the formal sector. Similarly, Ethiopia Therefore, monetary and fiscal policy should aim to has weak protection of minority shareholders; rules keep inflation low and the exchange rate policy should and practices around the coverage, scope, and acces- support a nominal exchange rate that is competitive. sibility of credit information are poor; and the rate The macro policy mix should take into account that of profit taxes is high relative to regional averages. A a faster pace of nominal currency depreciation would growing body of empirical research shows that simpler potentially induce inflation. Moreover, there are trade- processes of business start-up is associated with higher offs vis-à-vis a potentially higher cost of imports of levels of entrepreneurship and higher productivity capital equipment and balance sheet effects of external among existing firms. public debt to consider. 52 Annex 1: Ethiopia: Selected Economic Indicators (High Frequency) Jan-13 Feb-13 Mar-13 Apr-13 May-13 Jun-13 Jul-13 Aug-13 Sep-13 Oct-13 Nov-13 Dec-13 Inflation (Year-on-Year): % 12.7 11.0 7.7 6.1 6.3 7.4 8.4 7.0 6.9 8.5 8.0 7.7 Food 12.7 9.3 5.2 1.6 3.4 3.7 5.8 3.8 4.3 7.8 6.8 5.8 Non-Food 12.7 12.9 10.6 11.6 9.9 11.9 11.5 10.8 10.0 9.3 9.3 9.8 Inflation in AA (Year-on-Year):% 12.0 10.8 7.9 7.0 8.3 7.3 8.5 7.0 9.4 9.0 8.0 7.5 Traded Goods 4.3 2.8 1.2 2.6 3.4 4.6 3.7 3.4 3.7 4.0 4.0 4.4 Non-Traded 14.4 13.4 10.1 8.5 9.8 8.1 9.9 8.0 11.0 10.3 9.0 8.2 Monetary Growth (Year-on-Year):% M2 29.1 28.7 31.5 29.6 28.8 24.2 27.6 26.3 23.5 22.2 22.2 22.3 Domestic credit 27.4 29.3 26.2 24.3 23.9 23.4 24.6 29.4 28.4 28.5 28.5 30.6 Net Foreigh Assets –15.1 –15.4 2.6 11.4 16.9 14.7 5.7 4.9 –2.5 –5.4 –7.3 –15.6 Reserve Money* 16.9 23.8 13.2 18.4 18.5 13.6 12.1 9.6 7.7 1.5 –3.9 –3.7 Gross reserves (Mill. $) 2461 2725 2736 2599 2424 2368 2654 2561 2469 2393 2366 2598 In months of import 2.0 2.2 2.2 2.1 1.9 1.9 1.9 1.8 1.8 1.7 1.7 1.9 Exchange rate Exchange rate (Birr/$), pa 18.2 18.3 18.4 18.5 18.5 18.6 18.7 18.7 18.8 18.9 18.9 19.1 Real Effective Exchange index 134.7 134.1 134.4 133.6 133.7 133.4 137.0 136.9 138.5 138.6 141.1 142.1 annual growth, % 3.8 3.7 1.5 0.0 –0.9 –1.7 –0.1 –0.1 1.5 2.9 3.0 5.6 Black market premium (%) 7.4 8.7 9.2 10.3 9.2 7.5 6.4 5.9 4.0 3.1 2.6 Trade Deficit, goods, billion US$ –0.8 –0.7 –0.8 –0.7 –0.7 –0.6 –0.7 –0.8 –0.8 –1.0 –0.8 –1.0 Export, (billion US$) 0.2 0.2 0.2 0.2 0.3 0.3 0.2 0.2 0.2 0.2 0.2 0.2 Import, (billion US$) 1.0 0.9 1.1 1.0 1.0 0.9 0.9 1.0 1.0 1.2 1.0 1.3 International Prices Crude oil, average ($/bbl) 105.1 107.6 102.5 98.9 99.4 99.7 105.3 108.2 108.8 105.4 102.6 105.5 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS Coffee, arabica (cents/kg) 346.8 329.5 330.2 330.3 324.5 304.8 305.1 298.0 291.6 283.7 269.0 277.7 Gold ($/troy oz) 1671.8 1627.6 1593.1 1487.9 1414.0 1343.4 1285.5 1351.7 1348.6 1316.6 1275.9 1221.5 World Growth (quarterly: y-o-y) % Q1 Q2 Q3 Q4 China 7.7 7.5 7.8 7.7 Euro area –1.2 –0.6 –0.3 0.5 US 1.3 1.6 2.0 2.7 OECD-Total 0.6 1.0 1.5 … Sources: CSA; NBE, Customs, WB, OCED-National Accounts. * the growth may indicate the true picture since the change in reserve requirement from 10 to 5 percent resulted the balance converted in centificate of deposits which is kept outside reseve money. Annex 2: Ethiopia: Selected Economic and Social Indicators (Annual Frequency) Fiscal year ending July 7 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Average 2004–2013 Income and Economic Growth GDP growth at factor cost (annual %) 11.7 12.6 11.5 11.8 11.2 10.0 10.6 11.4 8.8 9.7 10.9 GDP growth at market price (annual %) 13.6 11.8 10.8 11.5 10.8 8.8 12.6 11.2 8.7 10.4 11.0 GDP per capita growth (annual %) 10.4 8.7 7.8 8.5 7.9 6.0 9.6 8.3 6.0 7.6 8.1 GDP per capita (US$) 134 160 192 241 322 375 337 351 467 498 308 GDP per capita, PPP (current international $) 575 646 718 799 879 938 1041 1150 1240 1354 934 Atlas GNI per capita, US$ 130 160 180 220 270 330 370 390 410 470 293 Private Consumption, nominal (annual %) 7.2 33.8 25.9 26.9 51.8 35.7 15.3 28.6 44.6 11.3 28.1 Gross Fixed Investment ( % of GDP) 29.8 26.5 27.9 24.5 24.7 25.6 27.4 27.9 33.1 33.0 28.0 Gross Fixed Investment - Public ( % of GDP) 14.9 13.9 16.0 14.1 14.7 12.3 15.5 18.7 14.3 12.2 14.7 Gross Fixed Investment - Private ( % of GDP) 14.9 12.6 11.9 10.4 10.0 13.3 11.9 9.2 18.7 20.8 13.4 Money and Prices Inflation, consumer prices (annual %, end of year) 1.7 13.0 11.6 15.1 55.3 2.7 7.3 38.1 20.8 7.4 18.4 Inflation, consumer prices (annual %, period average) 8.6 6.8 12.3 15.8 25.3 36.4 2.8 18.1 33.4 12.6 17.7 Treasury bill rate (91-days maturity, annual average) 0.5 0.1 0.0 0.8 0.6 0.9 0.9 1.3 1.9 2.2 0.9 Nominal Exchange Rate (End of period) 8.6 8.7 8.7 9.0 9.6 11.3 13.5 16.9 17.8 18.6 12.3 Real Exchange Rate Index (1990=100) 87.1 84.7 83.2 76.8 66.9 61.5 53.3 39.2 40.2 38.2 63.1 Fiscal Revenue (% of GDP) 16.2 14.8 15.0 12.8 12.1 12.1 14.2 13.7 13.9 14.6 13.9 Expenditure (% of GDP) 23.9 23.5 22.5 20.9 19.1 17.4 18.8 18.6 16.8 18.1 20.0 Current (% of GDP) 14.0 12.6 11.7 10.1 9.3 8.2 8.5 8.0 7.0 7.4 9.7 Capital (% of GDP) 9.6 10.8 10.8 10.8 9.8 9.2 10.4 10.5 9.9 10.7 10.3 Fiscal balance including grant (% of GDP) –3.0 –4.4 –3.9 –3.1 –2.9 –0.9 –1.3 –1.6 –1.2 –2.0 –2.4 Fiscal balance excluding grant (% of GDP) –7.7 –8.7 –7.5 –8.1 –7.0 –5.3 –4.6 –4.9 –2.9 –3.5 –6.0 Primary fiscal balance including grants (% of GDP)* –1.8 –3.5 –3.1 –2.4 –2.5 –0.6 –0.9 –1.2 –0.9 –1.6 –1.8 Consolidated public sector primary balance (% of GDP)* … … … –4.6 –6.7 –1.9 –1.0 –2.5 –3.7 –5.2 –3.7 ANNEXES Total public debt (% of GDP) 105.6 78.9 66.8 43.9 38.5 35.5 39.4 37.8 32.8 36.1 51.5 External public debt (% of GDP) 73.3 48.9 37.3 11.8 10.4 14.8 18.3 22.2 17.9 18.7 27.4 53 (continued on next page) 54 Annex 2: Ethiopia: Selected Economic and Social Indicators (Annual Frequency) (continued) Fiscal year ending July 7 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Average 2004–2013 External Accounts Export growth (%, yoy) 24.4 41.1 18.1 18.7 23.1 –1.0 38.3 37.1 14.8 –2.5 21.2 Import growth (%, yoy) 39.3 40.4 26.4 11.6 32.8 13.4 7.7 –0.2 34.0 3.7 20.9 Merchandise exports (current US$ billions) 0.6 0.8 1.0 1.2 1.5 1.4 2.0 2.7 3.2 3.1 1.8 of which coffee exports (current US$ billions) 0.2 0.3 0.4 0.4 0.5 0.4 0.5 0.8 0.8 0.7 0.5 Merchandise imports (current US$ billions) 2.6 3.6 4.6 5.1 6.8 7.7 8.3 8.3 11.1 11.5 7.0 Services, net (current US$ billion) 0.3 0.3 0.1 0.2 0.1 0.4 0.5 0.8 0.2 0.5 0.3 Service exports (current US$ billion) 0.9 1.0 1.1 1.3 1.6 1.9 2.0 2.6 2.8 2.9 1.8 of which travel 0.2 0.1 0.2 0.2 0.3 0.4 0.4 0.7 0.7 0.5 0.4 Service imports (current US$ billion) 0.6 0.7 1.0 1.1 1.5 1.5 1.5 1.8 2.6 2.3 1.5 Private transfers, net (BoP, current US$ billions) 0.8 1.0 1.2 1.7 2.4 2.7 2.7 3.2 3.2 3.9 2.3 Current account balance before grant (BoP, current US$ billions) –1.0 –1.5 –2.3 –2.1 –2.8 –3.2 –3.2 –2.1 –4.6 –4.0 –2.7 Current account balance after grant (BoP, current US$ billions) –0.4 –0.7 –1.4 –0.9 –1.5 –1.6 –1.3 –0.2 –2.8 –2.5 –1.3 Foreign Direct Investment (current US$ bilions) 0.2 0.2 0.4 0.5 0.8 0.9 1.0 1.2 1.1 1.2 0.7 External debt, total (Current US$, billion) 7.4 6.0 5.7 2.3 2.8 4.4 5.6 7.8 8.9 11.1 6.2 Merchandise exports (% of GDP) 6.0 7.0 6.7 6.1 5.5 4.5 6.8 8.8 7.4 6.6 6.5 of which coffee exports (% of GDP) 2.2 2.8 2.4 2.2 2.0 1.2 1.8 2.7 1.9 1.6 2.1 Merchandise imports (% of GDP) 26.0 29.8 30.6 26.5 25.6 24.3 28.1 26.3 25.8 24.5 26.8 Services, net (% of GDP) 3.1 2.3 1.0 0.8 0.5 1.3 1.7 2.4 0.4 1.0 1.5 Service exports (% of GDP) 9.0 8.3 7.4 6.7 6.0 6.1 7.0 8.2 6.6 6.1 7.1 of which travel (% of GDP) 1.6 1.1 1.0 0.8 1.1 1.1 1.2 2.3 1.6 1.1 1.3 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS Service imports (% of GDP) 5.9 6.0 6.4 5.9 5.5 4.8 5.2 5.8 6.2 4.9 5.7 Private transfers, net (BoP,% of GDP) 7.8 8.4 8.2 8.8 9.0 8.5 9.2 10.1 7.6 8.3 8.6 Current account balance before grant (BoP, % of GDP) –9.8 –12.5 –15.0 –10.7 –10.6 –10.0 –10.9 –6.6 –10.7 –8.6 –10.5 Current account balance after grant (BoP, % of GDP) –4.1 –6.1 –9.2 –4.5 –5.7 –5.1 –4.4 –0.7 –6.5 –5.4 –5.2 Foreign Direct Investment (% of GDP) 1.5 1.2 2.4 2.5 3.1 2.8 3.3 4.0 2.5 2.6 2.6 External debt, total (% of GDP) 73.3 48.9 37.3 11.8 10.4 14.8 20.1 26.1 21.3 24.3 28.8 Multilateral debt (% of total external debt) 63.5 82.7 81.1 51.6 55.7 46.7 48.6 46.0 45.4 45.0 56.6 Debt service ratio (% of goods and NFS) 10.7 8.9 8.0 7.3 2.9 2.3 2.7 4.5 6.9 9.3 6.4 (continued on next page) Annex 2: Ethiopia: Selected Economic and Social Indicators (Annual Frequency) (continued) Fiscal year ending July 7 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Average 2004–2013 Population, Employment and Poverty Population, total (millions), UN 74.1 76.2 78.3 80.4 82.6 84.8 87.1 89.4 91.7 94.1 83.9 Unemployment Rate (urban) 17.0 20.4 18.9 18.0 17.5 16.5 18.1 Poverty headcount ratio at national poverty line (% of population) 38.7 29.6 34.2 Poverty headcount ratio at $1.25 a day (PPP) (% of population) 39.0 30.7 34.8 Poverty headcount ratio at $2 a day (PPP) (% of population) 77.6 66.0 71.8 Inequality — Income Gini 29.8 29.8 29.8 Population Growh (annual %) 2.8 2.8 2.8 2.7 2.7 2.7 2.7 2.6 2.6 2.6 2.7 Life Expectancy 55.6 56.6 57.6 58.7 59.7 60.6 61.5 62.3 63.0 59.5 Others: GDP (current LCU, billions) 86 105 130 170 246 332 379 506 739 853 354 Nominal GDP (current US$, billions) 10 12 15 19 27 32 29 31 43 47 27 Doing Business (rank) a 101 97 102 116 107 104 111 124 108 b Human Development index ranking 170 170 169 169 171 157 174 172 173 169 CPIA (overall rating) 3.4 3.4 3.4 3.4 3.4 3.4 3.5 3.4 Economic management 3.7 3.5 3.5 3.3 3.7 3.7 3.7 3.5 Structural policies 3.2 3.2 3.2 3.2 3.2 3.2 3.2 3.2 Policies for Social Inclusion and Equity 3.6 3.6 3.7 3.6 3.6 3.6 3.7 3.7 Public Sector Management and Institutions 3.1 3.3 3.3 3.3 3.2 3.2 3.3 3.4 Note: a This indicator is ranked out of 175 countries in 2007, 178 in 2008, 181 in 2009 and 183 in 2010 and 2011. b The HDI ranking in 2001 is in relation to 175 countries; from 2005 to 2008, to 177; in 2009, to 181; in 2010, to 169 countries; and, from 2011–2013 to 187 countries. *Consolidated public sector primary balance includes SOE which is derived from the financing side while the primary fiscal balance includes federal and regional government but exclude SOEs. ANNEXES 55 56 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS Annex 3: Product Space Analysis of wheat, and bulbs, cuttings, plants. Figure 3.1 Map 4 Ethiopian Exports41 shows “stagnating” or declining exports as those that had RCA>1 in 2001–02 but not in 2011–2012. There Over the past decade, Ethiopia has undergone were 26 such export products. The most prominent some transformation in the production of goods. include raw materials of vegetable origin, sugars/beets, This transformation is reflected in the increase in the sugar cane, sheep, lamb skin (without the wool), and number of products in which it has demonstrated raw hides of goats, and calf skins. comparative advantage, from about 56 to 71. These There were hundreds of “marginal” exports, are products in which Ethiopia has performed bet- but only eight had a share of more than 0.1 per- ter in world markets compared to its overall export cent in national exports in the most recent years. performance. The change in industrial structures is Figure 3.1 Map 5 shows “marginal” exports as those slightly more dramatic when comparing 2012 with that had RCA<1 in both 2001–02 and 2011–12. They 1992 (not shown). In Figure 3.1 Map 1 (left panel), include footwear leather, meat of bovine animals, 46 of the 56 black dots (products with RCA>1 in bread/pastry, and motor vehicles. Analyzing why these 2001–02) belong to the primary sector. An additional marginal products continue to be exported, but not in six belong to manufactures derived from materials a competitive manner could shed light on the country’s (with origins in the primary sector, such as leather). export promotion efforts. Indeed, although Ethiopia’s Only two products belong to sectors that are deemed emerging exports include the highly promising horti- sophisticated, namely organic chemicals (5983) and culture items, several “marginals” are located close to power generating machines (7188). A decade later, 43 the more “desirable” clusters of product space. They are out of the 71 products belonged to the primary sector. desirable because a denser network between products Fifteen products had an RCA in material-based manu- suggests that those products share a high degree of facturing. About 10 products belonging to apparel knowledge about production techniques, facilitating (84) also had a Revealed Comparative Advantage. a speedier pace of structural transformation. The maps illustrate that Ethiopia has begun to have a Ethiopia could pay closer attention to the trade- foothold in light, labor-intensive manufacturing such off between density of exports and complexity/ as leather-based industries as well as apparel. sophistication. Figure 3.2 shows the tradeoff between There were 30 products that had a revealed com- parative advantage (RCA>1) in both 2001–02 and 41 According to the pioneers of the product space analytical tool, Haus- 2011–12. Nineteen of these contributed more than 1 mann et al. (2007), every product requires capabilities and knowledge percent each of the national export value in 2011–12. that are specific to that activity, from labor-training and physical assets to regulatory requirements, property rights, and infrastructure. The ease Table 3.1 filters products, at a more disaggregated with which an economy can move to producing new exports depends (SITC 4-digit) level, through the lens of significance on what its installed capability looks like already. The main hypothesis of the product space methodology is, therefore, that firms or nations in each of the four categories described next. In this that build up competence in producing a certain good can redeploy and report “significant” exports are those that had RCAI>1 adapt their human, physical, and institutional capital more easily if they seek to produce goods that are “nearby” those that they are producing in both 2001–02 and 2011–12. The most important already. Nearby goods share similar productive knowledge: a country are coffee, sesame, and vegetables, which together that makes one good is likely to also have the capabilities to produce others that are adjacent on a network of tradable merchandise goods account for more than 50 percent of total goods exports that maps the distance between pairs of products that are co-exported in 2011–12. Figure 3.1 Map 2 shows some of such sig- by a large number of countries world-wide. The greater the proximity, the easier it is for goods to be newly produced or scaled up, helping nificant exports in Ethiopia. Figure 3.1 Map 3 shows identify future opportunities in trade, production, and innovation. The “emerging” exports: those that had RCA<1 in 2001–02 process of structural transformation through reallocation of production (and employment) from low-productivity to high productivity sectors but RCA>1 in 2011–12. There were at least 33 of these, requires market failures to be redressed so that firms can move longer including a prominent category of cut flowers/foliage, distances in the product space. ANNEXES 57 FIGURE 3.1: Ethiopia’s Product Maps Map 1: 2001–2002 Map 1: 2011–2012 Map2 : Classic Exports Map3: Emerging Exports Bed linen, Sheep and Sesame seeds Other fresh or Beans, peas, table linen, goats, live chilled vegetables [2225] toilet & kitchen [12] lentils & other [6584] Other wheat [545] [542] (including spelt) Leather of other Under garments, [412] Animals hides or skins knitted of cotton [9410] [6116] [8462] Coffee Cotton yarn Sheep and [6513] [711] lamb skin leather Bulbs, tubers & [6115] rhizomes of flower Gold [2926] (Non-monetary) [9710] Animals of the Cut flowers bovine species and foliage Spices (except pepper [11] [2927] and pimento) Meat of sheep Other precious [752] and goats, fresh, chilled & semi-precious stones [112] [6673] Map 4: Declining Exports Map 5: Marginal Exports Leather of other Bakery products Goat & kid skins, (e.g., bread biscuit) raw (fresh, salted) bovine cattle [6114] Footwear [484] Vegetables, frozen [2114] Sheep & lamb [8510] Perfumery, cosmetics [546] skins without the wool and toilet prep Sugars, beet [2117] [5530] and cane, raw, solid [611] Passenger motor cars, for transport Calf skins, raw [7810] (fresh, salted, dried) Other outer [2112] garments & clothing Motor vehicles [8459] Bovine & equine hides for transport of goods [2111] [7821] Other materials Motor vehicles for Parts of heading of vegetable origin transport of goods 792, excl. tyres [2929] [7821] [7929] Source: Author’s own calculations utilizing the Product Space Methodology. 58 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 3.2: Complexity and Density Tradeoff proximity—the inverse of density42—and the com- in Ethiopia plexity of 102 marginal exports with a share of at least 0.005 percent in 2012. The relationship is upward- 2 sloping, i.e., the more sophisticated products are also harder to reach. In addition, the graph shows that Product complexity index 1 exports that are not yet significant require capabili- ties already acquired by successful (other) exports and 0 which are relatively sophisticated. Among those are −1 footwear, oilcake, knitted undergarments of synthetic fibers, knitted clothing accessories of textile fabrics, −2 luggage and handbags, clothing accessories of textile 5 10 15 20 25 30 35 40 45 50 fabrics, and women’s coats and jackets. This finding Product proximity is consistent with what was highlighted as Ethiopia’s Primary exports Chemicals Machinery emerging advantages in apparel and leather manu- Material−based manufactures Efficiency frontier facturing. Furthermore, an expanded list of 14 large “marginal” exports indicates that some fairly sophis- Source: Product Complexity Index (av. for 2007–09) obtained from ticated manufacturing products (such as cars and The Observatory of Economic Complexity; Density calculated by authors from COMTRADE data. aircraft parts) embody capabilities that are proximate Note 1: The efficiency frontier is a regression fit of PCI on density. to existing export competencies (Table 3.2). Note 2: Brown dashed lines indicate the mean (15) and the threshold of 1 s.d. below it (8.7). Note 3: Product density measures the proximity of a product to an 42 Density and proximity are related concepts. The higher the density of existing basket of exports with RCA. a product, the closer it is to existing exports with RCA. TABLE 3.1: Evolution of Significance of Exports Product Share Share RCA RCA Complexity Status SITC Product 2002 2012 2002 2012 (Rank) Significant 711 Green & roasted coffee 37.16 31.33 338.2 160.3 744 Significant 2225 Sesame seeds 6.8 14.11 960.6 1124.8 768 Significant 545 Other fresh or chilled vegetables 0.74 9.21 4.2 60.3 667 Significant 542 Dried or shelled legumes 6.25 6.1 142.2 128.4 724 Significant 9710 Gold, non-monetary 0.98 5.4 2.9 4 756 Significant 11 Live bovines 0.1 4.21 1.5 79.7 444 Significant 6115 Sheep & lamb leather 6.75 2.66 282.4 334.3 734 Significant 112 Sheep & goat meat 0.38 2.48 9.4 67.3 638 Significant 9410 Live animals, N.E.S. (zoo animals, pets, insects, etc) 0.01 1.74 2.2 302.7 684 Significant 752 Spices other than pepper 0.82 1.03 31 41.6 708 Significant 6116 Leather of other hides or skins 4.56 1.02 187.1 57.8 733 Marginal 8510 Footwear 0.01 0.32 0 0.5 541 Marginal 7810 Cars 0 0.27 0 0.1 140 Marginal 111 Bovine meat 0 0.16 0 0.7 531 Marginal 7929 Aircraft equipment parts N.E.S. 0 0.15 0 0.4 320 Marginal 484 Bakery 0.03 0.15 0.2 0.9 455 ANNEXES 59 TABLE 3.1: Evolution of Significance of Exports Product Share Share RCA RCA Complexity Status SITC Product 2002 2012 2002 2012 (Rank) Marginal 8459 Other knitted outerwear 0.03 0.14 0.1 0.4 659 Marginal 5530 Perfumery & cosmetics 0.01 0.13 0 0.3 379 Marginal 7821 Trucks & vans 0 0.13 0 0.2 303 Emerging 2927 Flora 0.02 6.03 0.3 101.7 715 Emerging 412 Other wheat & meslin, unmilled 0 1.31 0 4.9 521 Emerging 2926 Live plants 0 0.82 0 12.7 602 Emerging 12 Live sheep & goat 0.01 0.76 0.6 89.2 675 Emerging 6513 Cotton yarn 0 0.36 0 4.5 729 Emerging 6673 Not mounted precious stones 0 0.31 0 14.9 566 Emerging 6584 Linens & furnishing textile articles 0.06 0.28 0.4 1.9 622 Emerging 8462 Knitted undergarments of cotton 0.01 0.28 0 1.4 669 Declining 2929 Vegetable origin materials 11.78 0 302 0 583 Declining 2117 Raw sheep skin with wool 1.52 0 275.2 0 758 Declining 611 Raw sugar beet & cane 2.6 0 37.4 0 754 Declining 2112 Raw calf skins 1.38 0 115.7 0 625 Source: Authors computation from UN Comtrade data. Note 1: Rank out of 786; 1 is top.  TABLE 3.2: Prominent Marginal Exports SITC Product Density RCA in 2012 8459* Other knitted outerwear 0.16 0.4 813 Oilcake 0.12 0.5 8510* Footwear 0.12 0.5 111 Bovine meat 0.10 0.7 8219* Furniture parts N.E.S. 0.09 0.2 484 Bakery 0.08 0.9 1124 Alcoholic beverages 0.08 0.5 5530* Perfumery & cosmetics 0.07 0.3 1123 Beer 0.06 0.8 7821* Trucks & vans 0.06 0.2 7810* Cars 0.05 0.1 7822* Special purpose trucks & vans 0.05 0.9 7929* Aircraft equipment parts N.E.S. 0.04 0.4 7234* Construction & mining machinery 0.04 0.2 Source: Calculated by authors based on data from UN Comtrade. Note 1: * Indicates manufactured exports belonging to SITC Sections 5–8. 60 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS Annex 4: Exporter Dynamics in Selected previous years the entrants’ share of exports in total Products (2008–12) coffee exports surged. This suggests that entrants in 2012 were larger than entrants in previous years. Coffee43 Unequal survival rates across different key markets suggest a re-orientation of coffee exports. Coffee is the leading goods export for Ethiopia The one-year survival rates of new Ethiopian cof- although its share in total exports declined from fee exporters to the EU fell dramatically from 88 in 36 percent to 32 percent in 2008–2012. The export 2009 to 61 percent in 2011 and those to the U.S. unit price grew nearly 60 percent between 2008 and also fell during the same period from 73 to 55 per- 2011, but declined almost 20 percent between 2011 cent (Figure 3.4). In contrast, one-year survival rates and 2012. Despite this price decline, coffee export of new Ethiopian coffee exporters to Japan and to growth remained positive in that year as the quantity Saudi Arabia have increased. Due to a ban on coffee of coffee exports expanded. It is important to high- exports to Japan in 2009, the exit rate that year of light that while the value of coffee exports has grown coffee exporters was high (70 percent). But by 2010 every year between 2009 and 2012 the growth rate there was a high entry rate again combined with a has decelerated (Figure 3.3.1). low exit rate. Ethiopia’s coffee exporters are growing in numbers from 132 firms in 2008 to over 210 in Oil seeds44 2012. Coffee exporters represented only 11 percent of Ethiopia’s total number of exporters in 2012, but Oil seeds are the second largest merchandise export captured 32 percent of the country’s total exports for Ethiopia, representing 17 percent of total (Figure 3.3.2). exports in 2012. Among oil seeds, sesame is the most The largest exporters (selling more than important. Since 2009 oil seed exporters have enjoyed US$5 million per year) account for nearly 80 per- stable prices and the number of oil seed exporters has cent of Ethiopian coffee exports (Figure 3.3.3). remained practically unchanged at around 180 export- Rather small companies largely occupy the remaining ers (Figure 3.5.1). The share of Ethiopian exporters 20 percent of the coffee market with sales ranging that export oil seeds is low at around 10 percent. In from US$100 thousand to US$500 thousand per 2009 and 2012, the years with large increases in export year. Another way of showing the domination of large quantities, the share of entrants was larger than the companies in the coffee market is by looking at num- share of exiters (Figure 3.5.2). ber and market shares of companies with more or less The average size of oil seeds exporters than US$1 million sales per year. Forty-five percent increased from US$2 million in 2008 to of coffee exporters sold more than US$1 million per US$2.5 million in 2012. This increase in the year in 2012 and they accounted for over 95 percent average size of exporters was driven by the increase of Ethiopia’s coffee exports. The other 55 percent of in the average size of incumbent exporters, as the exporters barely accounted for 5 percent of exports average size of entrants actually declined over (Figure 3.3.4). the (Figure 3.5.3). The largest exporters selling The coffee market is relatively static without more than US$5 million represent 12 percent of many opportunities for new export firms to enter into business. Incumbent (established) exporters constitute around 70 percent of the total number of 43 Coffee exports are defined as exports of products within 4-digit HS code 0901. coffee exporters (Figures 3.3.5 to 3.3.6). Although 44 Oil Seeds exports are defined as exports of products within 4-digit the number of entrants declined in 2012 relative to HS code 1207. ANNEXES 61 FIGURE 3.3: Exporter Dynamics: Coffee 1. Trends of Coffee Exports and Exporters 2. Shares of Coffee Exports and Exporters 1.0 0.36 2008 0.09 0.24 2009 0.06 0.08 0.07 0.5 0.32 2010 0.09 0.06 0.09 0.33 2011 0.10 0.06 0.10 0 0.32 2012 0.11 0.08 0.11 0 0.1 0.2 0.3 0.4 –0.5 Share 2008 2009 2010 2011 2012 Share of exports Share of exporters Value Quantity Unit price Number exporters Share of entrants Share of exiters 3. Size Distribution of Value of Exports 4. Size Distribution of Number Exporters 81 80 82 80 78 40 68 33 29 60 30 26 26 27 25 26 26 22 22 Percent Percent 22 19 20 19 19 19 40 20 18 16 17 14 13 25 11 11 11 18 9 20 14 16 15 10 3 012 03 023 022 022 0 0 2008 2009 2010 2011 2012 2008 2009 2010 2011 2012 <100K 100−500K 500K−1M <100K 100−500K 500K−1M 1M−5M >5M 1M−5M >5M 5. Numbers of Entrants and Incumbents 6. Entrants’ Shares of Coffee Exports 0.12 46 2009 102 72 0.10 2010 114 Share of exports 61 2011 0.08 144 59 2012 152 0.06 0 50 100 150 Number of exporters 0.04 Entrants Incumbents 2008 2009 2010 2011 2012 Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graph 2 “Share of Exporters” shows the share of coffee exporters in the total number of exporters in Ethiopia, and “Share of Entrants” and “Share of Exiters” represent, respectively, shares of Ethiopian coffee exporter entrants and exiters in the total number of Ethiopian exporter entrants and exiters. 62 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 3.4: Dynamics of Coffee Exporters in Main Destinations EU JPN 0.88 0.8 0.8 0.78 0.69 0.70 0.67 0.61 0.6 0.6 0.57 0.50 0.49 0.40 0.39 0.4 0.4 0.33 0.35 0.25 0.26 0.24 0.27 0.21 0.24 0.22 0.19 0.2 0.15 0.2 0 0 2009 2010 2011 2012 2009 2010 2011 2012 SAU USA 0.8 0.8 0.73 0.6 0.6 0.57 0.55 0.52 0.44 0.43 0.45 0.46 0.4 0.36 0.33 0.33 0.38 0.36 0.36 0.4 0.38 0.34 0.30 0.27 0.28 0.25 0.25 0.21 0.2 0.2 0 0 2009 2010 2011 2012 2009 2010 2011 2012 Entry rate Exit rate Survival rate Source: Authors’ calculations based on data used for the Exporter Dynamics Database. exporters in 2012, a relatively high share for this After a big surge of Ethiopian exporter entrants size group (by comparison to other sectors) and into the Chinese market in 2009, the exit rate also those exporters account for 66 percent of total oil increased dramatically so somewhat correcting seed exports (Figure 3.5.4). for the previous influx. At the same time the U.S. After a large boom of entrants into oil seed seems to be a declining market for Ethiopia oil seeds exports in 2009, the entry rate declined rapidly exports as its share of exports has declined every year in 2010 and remained stable thereafter. At the since 2008. Survival rates are mixed across destination same time the exporter exit rate has been increasing. markets. Among all markets, the highest one-year Slightly more than 25 percent of oil seeds exporters survival rate in 2011 was verified for new exporters in 2009 did not export in the following year and that to Israel (61 percent) while the lowest was verified for rate increased to over 40 percent in 2011. However, new exporters to the U.S. (19 percent). Furthermore, the exit rate declined between 2011 and 2012, exit rates of Ethiopian oil seed exporters to the U.S. which allowed the entry rate to surpass the exit rate have been rising reaching 51 percent in 2012. In fact, and increasing the number of exporters in net terms entry rates were higher than exit rates in all main des- (Figures 3.5.5–3.5.6). tinations except the U.S. in 2012 (Figure 3.6). ANNEXES 63 FIGURE 3.5: Exporter Dynamics: Oil seeds 1. Trends of Oil seeds Exports and Exporters 2. Shares of Oil seeds Exports and Exporters 1.5 0.16 2008 0.08 0.25 2009 0.10 0.13 1.0 0.08 0.09 0.15 2010 0.08 0.11 0.5 0.09 0.14 2011 0.10 0.11 0.17 2012 0.10 0 0.10 0.13 0 0.05 0.10 0.15 0.20 0.25 –0.5 Share 2008 2009 2010 2011 2012 Share of Exports Share of Exporters Value Quantity Unit Price Number Exporters Share of Entrants Share of Exiters 3. Size Distribution of Value of Exports 4. Size Distribution of Number Exporters 30 30 30 31 31 29 29 80 30 70 26 26 27 66 66 63 62 21 60 20 19 20 17 18 17 Percent 15 Percent 40 13 12 11 12 28 28 28 10 10 22 22 10 8 8 20 45 46 47 43 0 0 1 0 033 0 0 2008 2009 2010 2011 2012 2008 2009 2010 2011 2012 <100K 100−500K 500K−1M <100K 100−500K 500K−1M 1M−5M >5M 1M−5M >5M 5. Numbers of Entrants and Incumbents 6. Entrants’ Shares of Oil seeds Exports 0.16 95 2009 89 0.14 69 2010 117 Share of Exports 0.12 67 2011 110 0.10 70 2012 110 0.08 0 50 100 150 Number of Exporters 0.06 Entrants Incumbents 2008 2009 2010 2011 2012 Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graph 2 “Share of Exporters” shows the share of oil seeds exporters in the total number of exporters in Ethiopia, and “Share of Entrants” and “Share of Exiters” represent, respectively, shares of Ethiopian oil seeds exporter entrants and exiters in the total number of Ethiopian exporter entrants and exiters. 64 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 3.6: Dynamics of Oil Seeds Exporters in Main Destinations CHN ISR 0.8 0.8 0.62 0.61 0.6 0.6 0.51 0.52 0.51 0.46 0.47 0.48 0.47 0.43 0.42 0.44 0.43 0.43 0.38 0.39 0.37 0.4 0.36 0.33 0.33 0.4 0.34 0.26 0.2 0.2 0 0 2009 2010 2011 2012 2009 2010 2011 2012 TUR USA 0.8 0.8 0.63 0.66 0.60 0.62 0.6 0.53 0.6 0.56 0.53 0.51 0.51 0.46 0.45 0.48 0.46 0.45 0.38 0.40 0.4 0.35 0.4 0.36 0.32 0.31 0.25 0.19 0.2 0.2 0 0 2009 2010 2011 2012 2009 2010 2011 2012 Entry rate Exit rate Survival rate Source: Authors’ calculations based on data used for the Exporter Dynamics Database. Cut flowers45 The average size of cut flowers exporters more than doubled between 2008 and 2012. Growth in Cut flowers account for approximately 7 percent average exporter size was entirely driven by growth of total Ethiopian exports and both the quanti- of incumbent exporters as the average size of entrants ties and values of exports of cut flowers increased actually declined (3.7.3). The share of the largest between 2008 and 2011. Unit prices of cut flowers exporters of cut flowers selling more than US$5 mil- exports have declined slightly since 2008. However, lion increased from 5 to 8 percent of the total num- the number of Ethiopian exporters of cut flowers has ber of exporters over the period. Furthermore, these declined substantially from 125 in 2009 to 80 in 2012 largest exporters increased substantially their share of (Figure 3.7.1). The share of the cut flowers sector in total cut flower exports from 37 percent in 2008 to the total number of exporters was 4 percent in 2012 59 percent in 2012. Moreover, the share of exporters and that sector’s share of total entrants as well as of selling US$1–5 million also increased substantially total exiters was a mere 1 percent. The smaller shares from 23 percent in 2008 to 36 percent in 2012 (3.7.4). of entrants and exiters relative to the shares of total exporters suggest less dynamism in the sector relative 45 Cut flowers exports are defined as exports of products within 4-digit to other exporting sectors in Ethiopia (Figure 3.7.2). HS code 0603. ANNEXES 65 FIGURE 3.7: Exporter Dynamics: Cut flowers 1. Trends of Cut Flowers Exports and Exporters 2. Shares of Cut Flowers Exports and Exporters 0.8 2008 0.07 0.07 0.09 2009 0.07 0.6 0.05 0.05 0.06 2010 0.02 0.04 0.4 0.09 0.07 0.2 2011 0.01 0.04 0.02 0.06 0 2012 0.01 0.04 0.01 –0.2 0 0.2 0.4 0.6 0.8 Share 2008 2009 2010 2011 2012 Share of exports Share of exporters Value Quantity Unit price Number exporters Share of entrants Share of exiters 3. Size Distribution of Value of Exports 4. Size Distribution of Number Exporters 44 60 59 58 59 40 36 35 34 46 46 30 28 29 40 39 25 25 37 24 23 Percent Percent 32 32 21 20 22 28 19 20 20 16 14 13 20 12 11 10 7 8 89 8 5 6 68 4 7 54 4 3 1 1 1 0 0 0 0 2008 2009 2010 2011 2012 2008 2009 2010 2011 2012 <100K 100−500K 500K−1M <100K 100−500K 500K−1M 1M−5M >5M 1M−5M >5M 5. Numbers of Entrants and Incumbents 6. Entrants’ Shares of Cut Flowers Exports 39 0.15 2009 85 14 2010 74 0.10 Share of exports 5 2011 71 6 0.05 2012 71 0 20 40 60 80 Number of exporters 0 Entrants Incumbents 2008 2009 2010 2011 2012 Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graph 2 “Share of Exporters” shows the share of cut flowers exporters in the total number of exporters in Ethiopia, and “Share of Entrants” and “Share of Exiters” represent, respectively, shares of Ethiopian oil seeds exporter entrants and exiters in the total number of Ethiopian exporter entrants and exiters. 66 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 3.8: Dynamics of Cut Flowers Exporters in Main Destinations ARE EU 0.8 0.8 0.65 0.67 0.59 0.57 0.56 0.60 0.60 0.6 0.6 0.56 0.48 0.39 0.4 0.36 0.4 0.31 0.30 0.29 0.20 0.21 0.2 0.2 0.17 0.14 0.13 0.13 0.06 0.07 0 0 2009 2010 2011 2012 2009 2010 2011 2012 NOR SAU 0.8 0.8 0.73 0.71 0.67 0.67 0.60 0.57 0.6 0.6 0.54 0.50 0.51 0.45 0.43 0.40 0.40 0.4 0.33 0.4 0.35 0.30 0.27 0.29 0.29 0.25 0.23 0.20 0.2 0.2 0 0 2009 2010 2011 2012 2009 2010 2011 2012 Entry rate Exit rate Survival rate Source: Authors’ calculations based on data used for the Exporter Dynamics Database. The Ethiopian cut flowers export sector is destinations (at more than 60 percent) and it increased becoming dominated by large incumbents. The every year since 2008. Norway exhibits the highest importance of entrants both in terms of the number exporter entry rates in every year but also the highest of exporters and of total exports has declined dramati- exit rates (Figure 3.8). cally, the latter falling from over 15 percent in 2009 to less than 1 percent in 2012. The number of cut flower Textile and Garments46 exporters in Ethiopia dropped dramatically between 2009 and 2012. Entry and exit rates show how the Within light manufacturing, textiles and gar- exit rate of cut flower exporters surpassed the entry ments represent an important industry, yet exports rate in 2009; this was only slightly reversed in 2012 accounted for only 2 percent of Ethiopia’s total. (Figures 3.7.5–3.7.6). The number of Ethiopian textilesexporters grew Looking at destinations, entry rates into the between 2009 and 2010 reaching 150 but then EU declined until 2012 while exit rates increased, making the EU a declining market for Ethiopian 46 Textiles and Garments exports are defined as exports of products with exporters. At the same time the one-year survival 2-digit HS codes between 50 and 63. For the remainder of this section rate of new exporters in the EU is largest among all the textiles and garments sector is referred to as “textiles”. ANNEXES 67 declined to 120 in 2012, which was the same num- Leather and Leather Products47 ber as in 2008. Although textiles account for only 2 percent of Ethiopia’s total exports, they account The share of leather exports has halved between for 7 percent of the total number of exporters 2008 and 2012. This was driven by a sharp decline (Figure 3.9.2) Furthermore, the share of textiles in in quantity (Figure 3.11.1), which could not be off- exporter entry and exit tends to be higher than their set by unit price increases of over 300 percent in the share in total exporters. same period. Ethiopian leather exporters initially The textile export sector is characterized by a seemed to have weathered the 2009 financial crisis high concentration of a few large firms (Figure 3.9.3 fairly well, as the number of firms remained almost and 3.9.4). The two largest categories of export- unchanged until 2010 and increased by 30 percent ers selling over US$1 million represented a total of (to reach 65) in 2011. But the number of exporters, 5 percent of exporters and 84 percent of Ethiopia’s export quantity, and value declined in 2012. This is textiles exports in 2012—with a rising trend over time. related to the overall policy environment. In 2011, In contrast, the large majority of exporters—more the government de facto fixed the price with the than 72 percent in any given year—exported less than goal of stemming price increases in the local leather US$100,000 and accounted for a mere 3 percent of market. This generated an opportunity for arbitrage, total textiles exports. with cheap skins and hides bought in the local mar- There was a high turnover in the textiles sec- ket and exported at a profit. In December 2011, the tor between 2008 and 2010, when the number of government imposed a 150 percent tax on exports entrants into textiles exports was larger than the of crust leather (which represented about 40 percent number of incumbent exporters (Figure 3.9.5). of Ethiopian leather exports in 2011), to protect the This picture changed more recently with less new domestic market entrants in 2011 and 2012, showing less dynamism The Ethiopian leather sector sees more exits in the sector. For instance, the entrants’ share in from their export businesses than the average total textiles exports declined from 23 percent in export community does. The share of Ethiopian 2010 to 10 percent in 2012 (Figure 3.9.6). So since leather exporter exiters in total exiters in 2009 2011 the exit rate has been higher than the entry was nearly double their share in total exporters rate explaining the recent decline in the number of (Figure 3.11.2) indicating a disproportionate presence textiles exporters. of exiters in this sector relative to other export sectors. Low survival rates among Ethiopian textiles In contrast, the share of leather exporter entrants in exporters in the EU are of particular concern given total entrants exceeded that in total exporters in 2011. given the overall importance of the EU market The fast growth in the number of leather exporters in those products (Figure 3.10). The exit rate has in 2011 accompanied by the quick fall in 2012 seem exceeded the entry rate in all years except 2010 for to be closely related to the described changes in the Ethiopian textile exporters in the EU. In 2012, for policy environment in 2011 and 2012. instance, the exit rate was 43 percent compared to an Leather exports are dominated by large export- entry rate of 33 percent. Furthermore, overall entry ers; in 2012, 94 percent of exports corresponded and exit rates into the EU are low. This is in contrast to exporters selling over US$1 million (which to other non-EU destination markets for Ethiopian represented 31 percent of exporters, Figure 3.11.3). textiles, where entry and exit rates are consistently higher than in the EU. This indicates more “churn- 47 The Leather and Leather Products sector is defined as exports of ing” in those smaller markets compared to the large products with 2-digit HS codes 41 and 42. For the remainder of this EU market. section the leather and leather products sector is referred to as “leather”. 68 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 3.9: Exporter Dynamics: Textile and Garments Exports 1. Trends of Textile Exports and Exporters 2. Shares of Textile Exports and Exporters 3 2008 0.02 0.08 0.02 2009 0.06 0.08 0.15 2 0.02 2010 0.07 0.12 0.10 0.03 2011 0.06 0.09 1 0.13 0.02 2012 0.07 0.08 0.06 0 0 0.05 0.10 0.15 Share 2008 2009 2010 2011 2012 Share of exports Share of exporters Value Quantity Unit price Number exporters Share of entrants Share of exiters 3. Size Distribution of Value of Exports 4. Size Distribution of Number Exporters 80 77 80 76 78 73 72 73 74 68 60 60 49 Percent Percent 42 40 40 27 22 19 17 20 1312 16 13 1412 16 20 12 15 10 11 9 66 67 4 5 1 5 4 6 5 3 2 3 3 3 2 3 3 1 2 2 0 0 0 0 2008 2009 2010 2011 2012 2008 2009 2010 2011 2012 <100K 100−500K 500K−1M <100K 100−500K 500K−1M 1M−5M >5M 1M−5M >5M 5. Numbers of Entrants and Incumbents 6. Entrants’ Shares of Textile Exports 0.25 58 2009 55 98 2010 53 0.20 Share of exports 58 2011 63 43 0.15 2012 78 0 20 40 60 80 100 0.10 Number of exporters Entrants Incumbents 2008 2009 2010 2011 2012 Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graph 2 “Share of Exporters” shows the share of textile exporters in the total number of exporters in Ethiopia, and “Share of Entrants” and “Share of Exiters” represent, respectively, shares of Ethiopian leather exporter entrants and exiters in the total number of Ethiopian exporter entrants and exiters. ANNEXES 69 FIGURE 3.10: Dynamics of Textile Exporters in Main Destinations CHN EU 1.00 1.00 1.0 1.0 0.8 0.8 0.67 0.60 0.6 0.6 0.56 0.54 0.48 0.47 0.45 0.45 0.40 0.43 0.43 0.4 0.33 0.4 0.32 0.33 0.25 0.25 0.20 0.2 0.2 0.00 0 0 2009 2010 2011 2012 2009 2010 2011 2012 SDN TUR 1.00 1.0 0.96 1.0 0.86 0.86 0.8 0.8 0.71 0.71 0.71 0.67 0.67 0.59 0.57 0.6 0.54 0.6 0.43 0.47 0.40 0.43 0.40 0.4 0.32 0.4 0.20 0.20 0.2 0.14 0.2 0.00 0 0 2009 2010 2011 2012 2009 2010 2011 2012 Entry rate Exit rate Survival rate Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database Yet, the majority of leather exporters are small, 50 China is a particular appealing market for percent or more sell less than US$100,000 in any Ethiopian leather exporters. In China, entry given year and they account for a minimal share of rates are very high relative to exit rates in every total leather exports. While skewed, the size distri- year except 2012 when exit rates exceed entry rates bution of leather exporters has been rather stable (Figure 3.12). Moreover, the Chinese market is across years. The only noticeable changes have been characterized by very high one-year survival rates an increase in the share of exports accounted for by for new exporters (above 70 percent), although the largest exporters (above US$5 million) and a fall survival rates declined over the past years. Similarly, in the share of exports accounted for by exporters entry rates have been substantially higher than exit selling US$1–5 million. rates over the last three years in Hong Kong, China, Leather is a fairly dynamic sector, where the and new exporters have also enjoyed high rates of number of new leather exporters exceeded the survival beyond their first year exporting to that number of incumbents both in 2009 and 2011 market. New leather exporters to the EU also have (Figure 3.11.5 and 3.11.6). Yet, the policy environ- high one-year survival rates but not as high as in ment triggered some of this recent activity. other destinations. 70 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 3.11: Exporter Dynamics: Leather and Leather Exports 1. Trends of Leather Exports and Exporters 2. Shares of Leather Exports and Exporters 2008 0.06 3 0.03 0.03 2009 0.03 0.03 0.06 2 0.02 0.03 2010 0.02 0.03 0.05 1 2011 0.03 0.05 0.02 0.03 2012 0.03 0 0.04 0.04 0 0.02 0.04 0.06 –1 Share 2008 2009 2010 2011 2012 Share of exports Share of exporters Value Quantity Unit price Number exporters Share of entrants Share of exiters 3. Size Distribution of Value of Exports 4. Size Distribution of Number Exporters 57 60 80 60 69 53 67 67 50 49 60 60 53 40 Percent Percent 30 40 37 34 26 26 27 20 20 17 17 17 20 12 13 11 10 9 11 6 10 10 9 9 7 6 4 47 6 4 2 011 1 10 01 133 0 0 2008 2009 2010 2011 2012 2008 2009 2010 2011 2012 <100K 100−500K 500K−1M <100K 100−500K 500K−1M 1M−5M >5M 1M−5M >5M 5. Numbers of Entrants and Incumbents 6. Entrants’ Shares of Leather Exports 24 2009 23 0.12 17 2010 31 0.10 Share of exports 35 2011 31 0.08 20 2012 41 0.06 0 10 20 30 40 0.04 Number of exporters Entrants Incumbents 2008 2009 2010 2011 2012 Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graph 2 “Share of Exporters” shows the share of leather exporters in the total number of exporters in Ethiopia, and “Share of Entrants” and “Share of Exiters” represent, respectively, shares of Ethiopian leather exporter entrants and exiters in the total number of Ethiopian exporter entrants and exiters. ANNEXES 71 FIGURE 3.12: Dynamics of Leather Exporters in Main Destinations CHN EU 1.00 1.0 1.0 0.86 0.8 0.71 0.8 0.62 0.6 0.6 0.55 0.54 0.50 0.44 0.43 0.40 0.37 0.4 0.35 0.4 0.31 0.32 0.28 0.29 0.26 0.28 0.21 0.23 0.2 0.2 0.07 0.10 0 0 2009 2010 2011 2012 2009 2010 2011 2012 HKG THA 1.0 1.0 0.83 0.80 0.8 0.75 0.8 0.75 0.75 0.67 0.67 0.6 0.57 0.6 0.50 0.50 0.46 0.38 0.38 0.40 0.4 0.36 0.35 0.4 0.33 0.25 0.25 0.2 0.19 0.2 0.09 0.00 0 0 2009 2010 2011 2012 2009 2010 2011 2012 Entry rate Exit rate Survival rate Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Live animals48 represented a particularly high—25 percent—share of total exiters in 2012 (Figure 3.13.2). Ethiopia’s exports of live animals represented The average size of live animals exporters has 3 percent of total exports in 2008 and increased to nearly unchanged between 2008 and 2012. On the 7 percent in 2012. Unit prices of live animals exports other hand, the average size of incumbents peaked in remained nearly flat during this period thus growth in 2010 and declined thereafter, whereas the average size export values was entirely driven by growth in quan- of entrants was mostly flat until 2010 then declined tities exported. The number of live animal exporters (3.13.3). In 2012 81 percent of live animal exporters exhibits the same exceptional growth from 100 in sold less than US$500 thousand while only 3 percent 2008 to 350 in 2011 (Figure 3.13.1). However, value, of exporters sold more than US$5 million. Both of quantity, and number of exporters experienced a sub- these shares were up from the corresponding values stantial decline in 2012 while the unit price increased. Entrants into live animals exports account for a larger share of total entrants than exporters of live animals 48 The Live Animals sector is defined as exports of products with 2-digit account for in total exporters. Also live animal exiters HS code 01. 72 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 3.13: Exporter Dynamics: Live Animals 1. Trends of Live Animals Exports and Exporters 2. Shares of Live Animals Exports and Exporters 3 2008 0.03 0.06 0.04 2009 0.05 0.06 0.11 0.06 2 2010 0.09 0.15 0.05 0.07 2011 0.18 0.35 0.08 1 0.07 2012 0.16 0.20 0.25 0 0 0.01 0.02 0.03 0.04 Share 2008 2009 2010 2011 2012 Share of exports Share of exporters Value Quantity Unit price Number exporters Share of entrants Share of exiters 3. Size Distribution of Value of Exports 4. Size Distribution of Number Exporters 44 42 44 61 60 40 37 53 52 35 35 35 34 47 31 32 30 40 34 Percent Percent 30 20 27 20 17 21 21 15 18 12 13 20 11 14 14 12 14 14 13 9 8 8 11 11 10 10 6 9 3 2 3 3 2 2 3 2 1 1 0 0 2008 2009 2010 2011 2012 2008 2009 2010 2011 2012 <100K 100−500K 500K−1M <100K 100−500K 500K−1M 1M−5M >5M 1M−5M >5M 5. Numbers of Entrants and Incumbents 6. Entrants’ Shares of Live Animals Exports 42 2009 0.35 47 124 2010 0.30 60 Share of exports 225 2011 0.25 126 109 2012 182 0.20 0 50 100 150 200 250 0.15 Number of exporters Entrants Incumbents 2008 2009 2010 2011 2012 Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graph 2 “Share of Exporters” shows the share of live animals exporters in the total number of exporters in Ethiopia, and “Share of Entrants” and “Share of Exiters” represent, respectively, shares of Ethiopian live animals exporter entrants and exiters in the total number of Ethiopian ex- porter entrants and exiters. ANNEXES 73 FIGURE 3.14: Dynamics of Live Animal Exporters in Main Destinations DJI EGY 1.0 0.93 1.0 0.90 0.89 0.81 0.8 0.8 0.75 0.70 0.70 0.71 0.66 0.67 0.59 0.61 0.6 0.57 0.6 0.50 0.48 0.38 0.41 0.4 0.4 0.37 0.29 0.25 0.27 0.2 0.2 0.10 0 0 2009 2010 2011 2012 2009 2010 2011 2012 SDN SOM 1.0 1.0 0.82 0.83 0.80 0.8 0.72 0.8 0.63 0.63 0.65 0.59 0.61 0.6 0.6 0.55 0.55 0.48 0.50 0.49 0.50 0.45 0.4 0.4 0.37 0.32 0.29 0.28 0.2 0.17 0.2 0.09 0 0 2009 2010 2011 2012 2009 2010 2011 2012 Entry rate Exit rate Survival rate Source: Authors’ calculations based on data used for the Exporter Dynamics Database. in 2008. So mid-sized exporters are essentially disap- and Sudan were higher than entry rates in 2012 indi- pearing (Figure 3.13.4). cating shrinking markets. The only main destination Live animal exports’ entry rates increased dra- where entry rates remained higher that exit rates was matically between 2009 and 2010 and remained Djibouti (Figure 3.14). high in 2011 but collapsed in 2012. Exit rates were the mirror image of entry rates. A rapid decline in the Meat and Meat Offal49 number of live animals exporters between 2011 and 2012 was accompanied by a rapid decline in entrants’ Ethiopian meat exports experienced an important share of exports. In 2012, exit rates surpassed entry expansion between 2008 and 2012 and represented rates thus explaining the fall in the number of live 2.5 percent of total exports in 2012. Growth in animals exporters in that year (Figures 3.13.5–3.13.6). the value of exports was mostly due to growth in the Somalia and Egypt are Ethiopia’s biggest des- tinations for live animal exports. But exporters in both markets have declining one-year survival rates 49 The Meat and meat offal sector is defined as exports of products within 2-digit HS code 02. For the remainder of this section we will refer to the since 2009. Also, exporter exit rates in Somalia, Egypt, meat and meat offal products sector as “meat”. 74 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 3.15: Exporter Dynamics: Meat 1. Trends of Meat Exports and Exporters 2. Shares of Meat Exports and Exporters 2.0 2008 0.02 0.01 0.02 2009 0.01 0.02 1.5 0.02 0.02 2010 0.02 0.03 0.02 1.0 0.03 2011 0.02 0.02 0.03 0.03 0.5 2012 0.01 0.02 0.02 0 0 0.01 0.02 0.03 Share 2008 2009 2010 2011 2012 Share of exports Share of exporters Value Quantity Unit price Number exporters Share of entrants Share of exiters 3. Size Distribution of Value of Exports 4. Size Distribution of Number Exporters 100 96 95 80 71 84 63 64 80 79 60 60 52 60 59 Percent Percent 40 40 37 19 19 19 20 20 18 20 13 20 13 10 1212 6 6 6 6 7 8 8 112 140 4 110 1120 000 0 0 0 0 0 0 2008 2009 2010 2011 2012 2008 2009 2010 2011 2012 <100K 100−500K 500K−1M <100K 100−500K 500K−1M 1M−5M >5M 1M−5M >5M 5. Numbers of Entrants and Incumbents 6. Entrants’ Shares of Meat Exports 0.08 12 2009 9 25 0.06 2010 9 Share of exports 14 2011 0.04 16 10 2012 15 0.02 0 5 10 15 20 25 Number of exporters 0 Entrants Incumbents 2008 2009 2010 2011 2012 Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graph 2 “Share of Exporters” shows the share of meat exporters in the total number of exporters in Ethiopia, and “Share of Entrants” and “Share of Exiters” represent, respectively, shares of Ethiopian live animals exporter entrants and exiters in the total number of Ethiopian exporter entrants and exiters. ANNEXES 75 FIGURE 3.16: Dynamics of Meat Exporters in Main Destinations ARE BHR 1.00 1.00 1.00 1.00 1.0 1.0 0.8 0.8 0.67 0.60 0.6 0.6 0.50 0.50 0.40 0.40 0.4 0.33 0.36 0.4 0.33 0.20 0.20 0.2 0.13 0.2 0.11 0.00 0.00 0.00 0.00 0 0 2009 2010 2011 2012 2009 2010 2011 2012 HKG SAU 1.00 1.0 1.0 0.8 0.8 0.67 0.63 0.6 0.6 0.56 0.50 0.43 0.42 0.42 0.4 0.4 0.20 0.20 0.2 0.2 0.14 0.00 0 0 2012 2009 2010 2011 2012 Entry rate Exit rate Survival rate Source: Authors’ calculations based on data used for the Exporter Dynamics Database. quantities exported, but the unit prices of meat exports the average size of entrants declined during the same also increased over 2010–2011. Unit prices contin- period (3.15.3). A large share of meat exporters in ued to increase over 2011–2012 but the quantities Ethiopia are very large: more than 20 percent of exported declined, leading the value of meat exports exporters in 2012 had annual exports of more than to fall in that period (Figure 3.15.1). An important US$5 million and they account for 95 percent of meat expansion in the number of meat exporters occurred exports. Those large exporters of meat seem to have between 2008 and 2010, which was prior to the very been affected by the global financial crisis of 2009 but fast export growth period: there were 16 meat export- have recovered since (Figure 3.15.4). ers in 2008 which increased to 34 in 2010 but declined The number of meat exporter entrants exceeded to 25 by 2012 (Figure 3.15.2). that of incumbent exporters between 2009 and The average size of meat exporters increased 2010. Although entry rates into meat exports were between 2008 and 2012 driven mostly by growth very high between 2009 and 2010, they declined since. in incumbent exporters’ size from US$2.75 million Exit rates also increased between 2009 and 2010 but to nearly US$5 million over the period. In contrast, declined thereafter. In 2011 the entry rate dropped 76 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS below the exit rate and the gap between the two has selling over US$5 million annually was almost zero widened indicating a shrinking meat export sector in before 2011 and increased to just 2 percent of total Ethiopia (Figures 3.15.5–3.15.6). pulses exporters in 2012. Prior to 2011 the largest The dynamics of meat exporters in its four share of exports corresponded to exporters selling major destinations show a worrying negative US$1–5 million annually but in 2012 that largest trend in exporter entry rates—from 60 percent share corresponded to large exporters. It is also worth in 2010 to only 13 percent in 2012. On the other noting the high share of exports corresponding to hand this was accompanied by a positive trend in mid-size exporters (Figure 3.17.4). exporter exit rates from 40 percent in 2010 to 50 The number of entrants and incumbent export- percent in 2012 in the Arab Emirates. One-year ers of pulses in Ethiopia was almost unchanged survival rates of new meat exporters in the Arab in 2008–2010. It only changed thereafter. In 2011 Emirates also declined from 2010 to 2011. Exporter the number of entrants declined and that of incum- entry rates into Saudi Arabia have also declined bents increased and the share of incumbents in total over the period while Hong Kong, which became a exporters increased to 61 percent. The entrants’ brand new destination for Ethiopian meat exporters share of pulses exports also increased in 2010 and in 2012, accounted already for 16 percent of total declined thereafter. There was a big decline in the exporters (Figure 3.16). exit rate of pulse exporters in 2010 but it rebounded in 2011. In contrast, entry rates increased slightly Pulses50 in 2009–2010 but declined in 2010–2011. While the gap between exit rates and entry rates narrowed Pulses accounted for 7 percent of Ethiopian exports in 2012, exit rates remained higher that entry rates in 2012, and while their export value declined in (Figures 3.17.5–3.17.6). 2009 during the global financial crisis it recovered Looking at destinations, Pakistan and India robustly thereafter. Export value, quantity, and unit became the largest importers of Ethiopian pulses prices of pulses have been on the rise since 2009. The in 2012. The entry rate of pulse exporters into number of Ethiopian pulses exporters was 380 in 2010 Pakistan was 59 percent while the exit rate was 30 but declined to 330 in 2012 (Figure 3.17.1). Pulses percent in 2012. For India the corresponding num- exporters accounted for 18 percent of total exporters bers were 68 percent (entry) and the 42 percent (exit). in Ethiopia in 2012. The pulses export sector appears Regarding Sudan, the exit rate remained higher than to have a higher level of churning than other export the entry rate over the period and the one-year sur- sectors in Ethiopia as its share of total entrants and vival rate of new exporters declined between 2009 of total exiters exceeded its share of total exporters in and 2011. In contrast, survival rates of new pulses all years (Figure 3.17.2). exporters in the EU are on the rise from 18 percent With increasing unit prices and a declining of new exporters in 2009 to 50 percent in 2012 number of exporters, the average size of pulses (Figure 3.18). exporters increased between 2008 and 2012. While the growth in the average size of pulses export- ers in 2012 was mostly driven by growth in the size of incumbent exporters, growth in the median size of pulses exporters is driven by the increase in the median size of entrants, which nearly doubled between 50 Pulses (a legume) exports are defined as exports of products within 2009 and 2012 (3.17.3). The share of large exporters 4-digit HS code 0713. ANNEXES 77 FIGURE 3.17: Exporter Dynamics: Pulses 1. Trends of Pulses Exports and Exporters 2. Shares of Pulses Exports and Exporters 0.8 2008 0.08 0.21 0.07 2009 0.16 0.19 0.6 0.35 0.06 2010 0.19 0.23 0.4 0.18 0.05 2011 0.17 0.21 0.2 0.25 0.07 2012 0.18 0.27 0 0.22 0 0.1 0.2 0.3 0.4 –0.2 Share 2008 2009 2010 2011 2012 Share of exports Share of exporters Value Quantity Unit price Number exporters Share of entrants Share of exiters 3. Size Distribution of Value of Exports 4. Size Distribution of Number Exporters 52 52 60 57 50 50 54 45 44 40 37 4040 40 37 37 31 32 35 29 30 Percent Percent 22 21 22 21 20 20 18 18 15 16 15 16 11 11 10 10 9 10 8 8 8 9 9 8 8 5 6 5 5 4 5 3 1 2 0 0 0 0 0 2008 2009 2010 2011 2012 2008 2009 2010 2011 2012 <100K 100−500K 500K−1M <100K 100−500K 500K−1M 1M−5M >5M 1M−5M >5M 5. Numbers of Entrants and Incumbents 6. Entrants’ Shares of Pulses Exports 0.24 140 2009 153 0.22 192 2010 189 Share of exports 0.20 132 2011 204 0.18 144 2012 188 0.16 0 50 100 150 200 Number of exporters 0.14 Entrants Incumbents 2008 2009 2010 2011 2012 Source: World Bank staff own calculations, based on data used for the Exporter Dynamics Database. Notes: Graph 2 “Share of Exporters” shows the share of pules exporters in the total number of exporters in Ethiopia, and “Share of Entrants” and “Share of Exiters” represent, respectively, shares of Ethiopian pulses exporter entrants and exiters in the total number of Ethiopian exporter entrants and exiters. 78 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS FIGURE 3.18: Dynamics of Pulses Exporters in Main Destinations EU IND 0.8 0.8 0.76 0.75 0.74 0.73 0.68 0.61 0.59 0.59 0.6 0.6 0.50 0.50 0.50 0.52 0.46 0.45 0.46 0.44 0.41 0.42 0.42 0.4 0.33 0.4 0.27 0.2 0.18 0.2 0 0 2009 2010 2011 2012 2009 2010 2011 2012 PAK SDN 0.8 0.8 0.72 0.69 0.64 0.59 0.59 0.59 0.60 0.6 0.6 0.51 0.52 0.45 0.46 0.45 0.45 0.42 0.42 0.41 0.4 0.33 0.4 0.29 0.30 0.29 0.28 0.25 0.2 0.2 0 0 2009 2010 2011 2012 2009 2010 2011 2012 Entry rate Exit rate Survival rate Source: Authors’ calculations based on data used for the Exporter Dynamics Database. ANNEXES 79 Annex 5: Additional details on Measuring a country’s RER misalignment RER calculations and theoretical considerations In the first step, we measure an RER misalignment index after controlling for the Balassa-Samuelson Examples of the theoretical relationship effect. The Balassa-Samuelson effect captures the effect between the real exchange rate and exports of an economy’s productivity on its non-tradable goods’ prices. In details, this can be explained as fol- Two examples show the theoretical impact of RER lows: We usually observe that the prices of services adjustments on exports: (like a haircut) are higher in developed countries than in developing countries, because wages are higher in  First, a reduction in domestic demand would developed countries. But why wages are higher in lower both the RER and the price of export. This developed countries? It is because the tradable sector is since the reduced domestic demand would of developed countries has higher productivity than lower the prices of both the domestic non-trad- that in developing countries. Given the law of one able and tradable goods. Since the price of the price on tradable goods, this implies that wages paid foreign tradable good does not change (much), to tradable-sector workers in developed countries have this implies that the domestic tradable good to be higher to commensurate their high productiv- would become relatively cheaper compared to the ity. In other words, low productivity explains a large foreign tradable good. This in turn would imply part why the tradable/non-tradable good price ratio cheaper exports. (i.e. the real exchange rate) in developing countries In addition, since both the prices of the non- is larger than that in developed countries. After the tradable good and tradable good decline com- Balassa-Samuelson effect is captured, the remaining pared to the foreign goods, the relative price of residual is considered the misaligned part. the domestic consumption basket becomes lower, We capture the Balassa-Samuelson effect as implying a depreciated RER. So there would be follows: cheaper export and a depreciated RER.  Second, consider a policy that would subsidize ( ) si ,t ∗ ln RERiW ,t = β si ,t ∗ lny i ,t − lny w ,t + ui ,t the production of domestic tradable goods. As a result, production of the domestic tradable It is a weighted regression (to take into account goods would expand and they would become the fact that larger countries have heavier weights in cheaper compared to the foreign tradable good, the regression). A country’s productivity is proxied by implying cheaper export. On the other hand, Yi ,t the unsubsidized domestic non-tradable goods its output per capita. si ,t = is the weight N would become relatively scarce and hence more ∑ j =1Y j ,t expensive. The increase in the non-tradable of country i at time t. Yi ,t is country i’ nominal out- goods’ prices could outweigh the decline of the put;. ln RERiW ,t is the log of the real exchange rate tradable goods’ prices, thus possibly making the of country i relative to the world; lny i ,t , lny w ,t are price of the domestic consumption basket to go country i and world average output per capita at time up compared to that of the foreign consump- t. Coefficient β captures the Balassa-Samuelson effect tion basket. The RER would appreciate. So the with an expected negative sign. The idea is that accord- subsidy makes export cheaper, but the RER to ing to Balassa-Samuelson effect, a country’s RER, at appreciate. any given time, is larger if its output per capita (a proxy 80 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS for productivity) is smaller compared to the world’s Measuring the impact of undervalued RERs output. This is slightly different to Rodrik approach, in on export and output growth the sense that he only regresses a country’s RER with its absolute output per capita. Since RER is a relative To examine econometrically the relationship between concept, we decide to add the world average output to a country’s RER undervaluation with its export and the right hand side of the equation to generate output output growth using international data the following differential, which is a relative concept as well. regression is used: Notice that there is no constant in the regression u 〉 and no time and country fixed effects. The regression growthratei ,t = α + γ i ,t + f t + f i + εi ,t si ,t is designed that the sum of the right hand side exactly equals the sum of the left hand side every period where growth rate is calculated for both real (i.e. sum of ui ,t equals 0 for all t). What it means is exports and real GDP, ui ,t / si ,t is the undervalua- 〉 uiu 〉 〉 growthrate growthrate α γ t ,t++f f ++f f ++ ,i ε, that at any given time, on average, the world RER is =α +γ t ,t measure, ition ,i = + and t t , i iareεi country i t ,t and time fixed sis,it ,t exactly aligned. effects. The time fixed effects is to control for global The results show that Balassa-Samuel effect is macroeconomic factors that affect all countries’ export highly significant with a negative sign. It shows that for in the same way at a given time. The country fixed each additional 1 percent output differential, Balassa effects is to control for country’s time-invariant char- Samuelson effect on average explains 0.317 percent of acteristics. Essentially, with the country fixed effects, RER appreciation. What this means is that for each 1 we essentially ask the following question: how a real percent output differential, the productivity differen- export growth changes within a country, given its RER tial accounts for 0.317 percent of the RER differential undervaluation index relative to the rest of the world? between countries. We also control for the initial value of export and out- ui ,t / si ,t will be our RER misalignment variable put levels. The expected sign of is positive: it implies 〉 of country i where ui ,t is the residual of the regres- that a more undervalued exchange rate (a larger 〉 ui ,t ) is associated with higher export and output growth. 〉 sion. A positive ui ,t implies an undervalued RER. That is, the RER is larger beyond the explanation of si ,t the Balassa-Samuelson effect. In other words, the trad- Two results are derived and shown in Tables 3.4 able/non-tradable good price ratio is larger, beyond the and 3.5. Simple descriptions of the results are in the explanation of the Balassa-Samuelson effect. main text. TABLE 3.3: Panel Estimation Effect on Exports Growth of Undervaluation First Stage Regression First-Stage Balassa-Samuelson wit*ln(RERiwt) Full Sample Weighted relative GDP growth –0.317*** (0.0117) Time Fixed Effect no Country Fixed Effect no Observations 8,184 R-squared 0.804 Notes: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 ANNEXES 81 TABLE 3.4: Undervalued RERs and Export Growth Second-Stage All Countries High Income Low Income Dln(real exports) Dln(real exports) Dln(real exports) Full Sample Full Sample 1950–1980 1981–2011 Full Sample 1950–1980 1981–2011 (1) (2) (3) (4) (5) (6) (7) ln(Initial Real –0.0999*** –0.0847*** –0.137*** –0.127*** –0.127*** –0.388*** –0.128*** Exports) (0.0225) (0.0151) (0.0233) (0.0187) (0.0437) (0.0631) (0.0250) Undervaluation 0.0598*** 0.0811*** 0.126*** 0.0497 0.0502** –0.0375 0.0699*** (0.0131) (0.0188) (0.0379) (0.0336) (0.0211) (0.0708) (0.0258) Constant 0.778*** 0.547*** 1.163*** 1.199*** 0.771*** 2.431*** 0.934*** (0.148) (0.104) (0.174) (0.168) (0.273) (0.401) (0.173) Time Fixed Effect yes yes yes yes yes yes yes Country Fixed Effect yes yes yes yes yes yes yes Observations 7,139 3,561 1,444 2,117 3,578 1,305 2,273 R-squared 0.139 0.159 0.157 0.209 0.148 0.334 0.159 156 75 63 75 81 63 81 Notes: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Low Income countries: if the real GDP per capita was below US$ 6,000/year in 2000. TABLE 3.5: Undervalued RERs and Output Growth Second-Stage All Countries High Income Low Income Dln(real GDP) Dln(real GDP) Dln(real GDP) Full Sample Full Sample 1950–1980 1981–2011 Full Sample 1950–1980 1981–2011 (1) (2) (3) (4) (5) (6) (7) ln(Initial Real Ex- –0.0916*** –0.141*** –0.215** –0.294** –0.0830*** –0.182*** –0.135*** ports) (0.0204) (0.0450) (0.0983) (0.114) (0.0128) (0.0372) (0.0171) Undervaluation 0.0884*** 0.138*** 0.195*** 0.183** 0.0764*** 0.121*** 0.114*** (0.0204) (0.0496) (0.0731) (0.0828) (0.0123) (0.0248) (0.0184) Weighted relative GDP growth Constant 0.778*** 0.547*** 1.163*** 1.199*** 0.771*** 2.431*** 0.934*** (0.148) (0.104) (0.174) (0.168) (0.273) (0.401) (0.173) Time Fixed Effect yes yes yes yes yes yes yes Country Fixed Effect yes yes yes yes yes yes yes Observations 8,020 3,925 1,570 2,355 4,095 1,579 2,516 R-squared 0.140 0.219 0.295 0.297 0.129 0.182 0.175 Number of countryid 165 80 68 80 85 73 85 Notes: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 Low Income countries: if the real GDP per capita was below US$ 6,000/year in 2000. 83 REFERENCES AfDB, OECD, UNDP and ECA. 2013. African Goods and Services.” World Bank, Washington, E c o n o m i c O u t l o o k 2 0 1 3 : St r u c t u r a l DC. Transformation and Natural Resources, African Collier, P and A. Venables. 2007. “Rethinking Development Bank, Organisation for Economic Trade Preferences: How Africa Can Diversify Co-operation and Development, United Nations Its Exports.” The World Economy 30(8): Development Programme and Economic 1326–1345 Commission for Africa. de Vries, Timmer and de Vries. 2013. “Structural Allaro, H. B. 2012. “The Effect of Export-Led Growth Transformation in Africa: Static Gains, Dynamic Strategy on the Ethiopian Economy” American Losses,” GGDC Research Memorandum 136. Journal of Economics 2012, 2(3): 50–56. Dinh, H., V. Palmade, V. Chandra, and F. Cossar. Aleksynska, M., and O. Havrylchyk. 2011. “FDI from 2012. “Light Manufacturing in Africa: Targeted the South: The Role of Institutional Distance and Policies to Enhance Private Investment and Create Natural Resources.” Working Paper 2011–05. Jobs.” World Bank, Washington, DC. Centre D’Études Prospectives et D’Informations Dinh, H., T. G. Rawski, A. Zafar, L. Wang, and E. Internationales, Paris. Mavroeidi. 2013. ”Tales from the Development Athukorala, P. and S. Waglé. 2013, “Export Frontier: How China and Other Countries Performance in Transition: The Case of Georgia.” Harness Light Manufacturing to Create Jobs and Working Papers in Trade and Development No. 2. Prosperity.” World Bank, Washington, DC. Australian National University, Canberra. Eden Maya and Ha Nguyen. 2012. “Correcting Real Bera, S., and S. Gupta. 2009. “South-South FDI vs. Exchange Rate Misalignment: Conceptual and North-South FDI: A Comparative Analysis in the Practical Issues.” World Bank Policy Research Context of India.” Working Paper 238. Indian Working Paper No. 6405. Council of Research in International Economic Davies, E. and L. Nilsson. 2013. “A Comparative Relations, New Delhi. Analysis of EU and US Trade Preferences for the Bernard, A., J. B. Jensen, S. J. Redding, and LDCs and the AGOA Beneficiaries.” European P. K. Schott. 2011. “The Empirics of Firm Commission Chief Economist’s Note, Issue 1. Heterogeneity and International Trade.” Annual Farole, T., J..G. Reis and S. Waglé. 2010. “Analyzing Review of Economics 4:283–313. Trade Competitiveness: A Diagnostics Approach,” Brenton, P., C. Saborowski, and E. von Uexkull. Policy Research Working Paper 5329, World 2009. What Explains the Low Survival Rate Bank, Washington, DC of Developing Country Export Flows? World Farole, T. and J. G. Reis. 2012. “Trade Competitiveness Bank, Policy Research Working Paper, No. 4951, Diagnostic Toolkit.” World Bank, Washington, Washington, DC. DC Brenton, P. and G. Isik, eds. 2012. “De-fragmenting Farole, T. and D. Winkler. 2014. “Making Foreign Africa: Deepening Regional Trade Integration in Direct Investment Work for Sub-Saharan Africa: 84 3RD ETHIOPIA ECONOMIC UPDATE – STRENGTHENING EXPORT PERFORMANCE THROUGH IMPROVED COMPETITIVENESS Local Spillovers and Competitiveness in Global Lederman, D. and W. Maloney. 2012. “Does What Value Chains.” World Bank, Washington, DC. You Export Matter? In Search for Empirical Ferro, E. and A. Fernandes. 2013. “A Note on Guidance for Industrial Policies.” World Bank, Ethiopian Exporters over the 2008–2012 Period,” Washington, D.C. Background Paper commissioned for this report. Lewis, Arthur. 1954. “Economic Development with World Bank. Processed. Unlimited Supplies of Labour.” The Manchester Freund, C. and D. Pierola. 2012. “Export Superstars.” School. Volume 22, Issue 2, pages 139–191. World Bank Policy Research Working Paper 6222. Martins, P. 2014. “Structural Change in Ethiopia: Gelb, A., C. Meyer, and V. Ramachandran. 2013. An Employment Perspective.” World Bank Policy Does Poor Mean Cheap? A Comparative Look at Research Working Paper 6749, Washington D.C. Africa’s Industrial Labor Costs. Center for Global McKinsey Global Institute. 2010. “Lions on the move: Development. Working Paper 325. May 2013, The progress and potential of African economies.” Washington DC. McMillan, M. and D. Rodrik. 2011. “Globalization, Gill, I., I. Izvorski, W. van Eeghen and D. De Rosa. Structural Change, and Productivity Growth.” 2014. “Diversified Development: Making the NBER Working Paper 17143. National Bureau Most of Natural Resources in Eurasia.” World of Economic Research. Bank, Washington, DC. Nguyen, H. 2014. “Real Exchange Rate and Export Hausmann, R., D. Rodrik, and J. Hwang. 2007. Growth in Ethiopia.” Background Paper com- “What You Export Matters,” Journal of Economic missioned for this report. World Bank. Processed. Growth 12(1). Papgeorgiu, C. and N. Spatafora. 2012. Economic Hausmann, R. and B. Klinger. 2007. “The Structure Diversification in LICs. Stylized Facts and Macro- of the Product Space and the Evolution of economic Implications. IMF Staff Discussion Comparative Advantage.” CID Working paper Note, SDN/12/13, December 14 No. 146, Harvard University. Rauch, J. E. 1999. “Networks versus markets in IMF. 2012. “Article IV Consultation 2012.” IMF international trade.” Journal of International Country Report No. 12/287. October. Economics: 48: 7–35. IMF. 2013. “Article IV Consultation 2013.” IMF Rodrik, D. 2013b. “Unconditional Convergence Country Report No. 13/308. October. in Manufacturing.” The Quarterly Journal of Kabeer, N., and S. Mahmud. 2004. “Rags, Riches Economics 128 (1): 165–204. and Women Workers: Export Oriented Garment Rodrik, D. 2013b. “The Past, Present and Future of Manufacturing in Bangladesh.” In M. Carr, ed., Economic Growth,” Global Citizen Foundation Chains of Fortune: Linking Women Producers Working Paper 1, June 2013. and Workers with Global Markets. London: Rodrik, D. 2008. “The Real Exchange Rate and Commonwealth Secretariat. Economic Growth.” Brookings Papers on Ko r i n e k , A n t o n a n d L u i s Se r v e n . 2 0 1 0 . Economic Activity. Fall 2008. “Undervaluation through Foreign Reserve Tensay, T.M. 2006. “Measuring the real exchange Accumulation: Static Losses, Dynamic Gains.” rate misalignment of Ethiopia: a dynamic OLS World Bank Policy Research Working Paper No approach.” Ethiopia Journal of Economics. 5250. UNCTAD. 2014, “World Investment Report Lee, Jaewoo, Gian Maria Milesi-Ferretti, Jonathan 2013.” Annex Tables, accessed on 2014/01/28 Ostry, Alessandro Prati, and Luca Antonio Ricci. http://unctad.org/en/pages/DIAE/World%20 2006. “Exchange rate assessments: CGER meth- Investment%20Report/Annex-Tables.aspx odologies,” IMF Occasional Paper 261. References 85 Waglé, S. 2007. “Geographical Indications as Trade- PREM Africa Region of the World Bank, Related Intellectual Property.” Discussion Paper. Washington, DC. UNDP Regional Center in Colombo. World Bank. 2012a. “Ethiopia Economic Update I: Waglé, S. 2011. “Investing Across Borders with Overcoming Inflation, Raising Competiveness” Heterogeneous Firms: Do FDI-Specific World Bank, Washington, DC. Regulations Matter?” Policy Research Working World Bank. 2012b. “Chinese FDI in Ethiopia: A Paper No. 5914. World Bank, Washington, DC. World Bank Survey.” World Bank, Washington, Waglé, S. 2014. “Export Performance and DC. Competitiveness: Analytical Inputs to the Ethiopia World Bank. 2013. “Ethiopia Economic Update II: Economic Update 2014.” Background Paper com- Laying the Foundation for Achieving Middle missioned for this report. World Bank. Processed. Income Status.” World Bank, Washington, DC. World Bank. 2011. “Southern Africa: Harnessing Regional Integration for Trade and Growth.” 1818 H Street, NW Washington, DC 20433