74385 Ethiopia Economic Update Overcoming Inflation, Raising Competitiveness the World Bank Ethiopia Economic Update Overcoming Inflation, Raising Competitiveness the World Bank Authors: Michael Geiger and Chorching Goh Contributors: Elisa Ticci, Taehyun Lee, Mesfin Bezawagaw, Nora Dihel, Juan Baron, Toru Nishiuchi, Xiaochen Fu, Jingyi Zhang November 2012 Table of Contents List of Acronyms......................................................................................................................................... v Economic overview................................................................................................................................... 1 The External Sector....................................................................................................................................................2 Fiscal Policy................................................................................................................................................................7 The Growth and Transformation Plan (GTP).............................................................................................................9 Monetary Policy.......................................................................................................................................................10 Inflation Rate...........................................................................................................................................................11 In depth: Inflation and its impact on the poor................................................................... 15 Background..............................................................................................................................................................15 The 2010/11 Inflation Situation in Ethiopia.............................................................................................................16 Distributive Impact of the Rise in Consumer Prices between July 2010 and October 2011......................................18 In depth: Agglomeration and competitiveness in Ethiopia......................................... 25 Agglomeration and Economic Growth.....................................................................................................................25 Financing the Development of Addis Ababa.............................................................................................................27 Improving Competitiveness in Ethiopia....................................................................................................................29 Lessons from East Asian Experiences........................................................................................................................31 Specific Recommendations to Increase Ethiopia’s Competitiveness in Light Manufacturing.....................................32 Summary: Unfolding the Potential of a Thriving Economy......................................................................................39 List of Boxes Box 1: Ethiopia’s Development Experience Compared to China and Korea: Selected Indicators and Policies..................4 Box 2: Ethiopia’s Services Exports...................................................................................................................................8 Box 3: Welfare Impact of Recent Price Shocks in Ethiopia based on Nationally Representative Datasets......................16 Box 4: Impact of Inflation: Methodology and Data......................................................................................................18 Box 5: Increases of Electricity Tariffs in Ethiopia – Insights on the Possible Effects on the Poor....................................20 Box 6: Business Services to Improve Competitiveness...................................................................................................31 Box 7: Korea’s Experience in Special Economic Zones: Masan Free Export Zone ........................................................34 Box 8: Privatization Strategies in China and Korea.......................................................................................................35 List of Figures Figure 1: Ethiopia GDP Growth Rates, 2004 to 2011....................................................................................................1 Figure 2: Ethiopian Exports by Commodity Groups, 2010/11 vs. 2011/12....................................................................2 Figure 3: Ethiopian Imports, by Import Categories, 2010/11 vs. 2011/12......................................................................3 Figure 4: Development of the Exchange Rate in Ethiopia...............................................................................................7 iv Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Figure 5: Services Exports from Ethiopia........................................................................................................................8 Figure 6: Ethiopia’s fiscal deficit and debt, 2009/10 to 2011/12.....................................................................................9 Figure 7: Base and Broad Money Growth in Ethiopia, Jul/Nov 2010 to May 2012......................................................11 Figure 8: Inflation Rate in Ethiopia, Jan 2005 to Oct 2012..........................................................................................12 Figure 9: Inflation and GDP Growth Rates in China, 1991–2010...............................................................................12 Figure 10: Inflation Rates in Ethiopia, Jan 2001 to Oct 2011.......................................................................................17 Figure 11: Food Prices between Jul 2010 and Oct 2011, by Item and Region, Percentage Change...............................17 Figure 12: Consumer Prices, Food Indices (2000 = 100)..............................................................................................18 Figure 13: Comparison of Average Effective Inflation Rates..........................................................................................22 Figure 14: Effective Inflation Rates between July 2010 and October 2011, by Region, Average and Min and Max by Quintile, based on HICES 2004/05...........................................23 Figure 15: Addis Ababa’s Economic Density Compared to its Immediate Neighbors....................................................26 Figure 16: Addis Ababa’s Economic Density Compared to Selected International Cities...............................................26 Figure 17: Addis Ababa’s per Capita Revenues, in Ethiopian Birr (ETB).......................................................................29 Figure 18: Average productivity of users vs. non-users of IT and professional services in Ethiopia................................31 Figure 19: Development Trends in Masan FEZ (Korea)................................................................................................34 List of Tables Table 1: Population below the National Poverty Line (less than US$0.6 per day)...........................................................2 Table 2: Ethiopia, China, Korea, and Vietnam: Main Growth Periods and their Indicators............................................4 Table 3: A Comparison of Chinese and Ethiopian FDI Regulation................................................................................5 Table 4: Total Service Trade and Service Subsector as a Share of Total Trade, 2000–2010...............................................8 Table 5: Revealed Comparative Advantage of Ethiopia’s Services Exports in Selected Subsectors.....................................9 Table 6: Selected Macroeconomic Targets of the GTP..................................................................................................10 Table 7: Self-food Expenditure as Share of Total Food Expenditure, 2005....................................................................19 Table 8: Food Expenditure Share by Area and Poverty Status........................................................................................19 Table 9: Simulated Changes in Poverty Indices due to Electricity Tariff Adjustments, Three Scenarios, 2005/06..........21 Table 10: Food Expenditure Share by Area and Quintiles of the Population.................................................................23 Table 11: Addis Ababa’s Competitiveness Relative to 8 other African Cities..................................................................30 Table 12: Challenges for Manufacturers in Six Selected Ethiopian Industries, 2010......................................................32 Table 13: Productivity and Cost-related Constraints for Firms and Industries in Ethiopia, 2010..................................32 References.................................................................................................................................................... 41 List of Acronyms AACAREP Addis Ababa City Administration KEPCO Korea Electric Power Company Revenue Enhancement Plan KT Korea Telecom Corporation COMESA Common Market for Eastern and MOFED Ministry of Finance and Economic Southern Africa Development CSA Central Statistics Agency of NBE National Bank of Ethiopia Ethiopia NEER Nominal Effective Exchange Rate EAC East African Community POSCO Pohang Steel and Iron Corporation EEPCO Ethiopia Electric Power Company RCA Revealed Competitive Advantage EFY Ethiopian Fiscal Year REER Real Effective Exchange Rate ETB Ethiopian Birr (currency) SADC Southern African Development EU European Union Community FDI Foreign Direct Investment SEZ Special Economic Zone FEZ Free Export Zone SNNPR Southern Nations, Nationalities, FIE Foreign Invested Enterprise and People‘s Region FTA Free Trade Agreement SOE State-owned Enterprise GDP Gross Domestic Product SSA Sub-Saharan Africa GoE Government of Ethiopia US United States GTP Growth and Transformation Plan VAT Value Added Tax HICES Household Income Consumption WTO World Trade Organization and Expenditure Survey MoFED Ministry of Finace and Economic IMF International Monetary Fund Development JV Joint Venture NBE National Bank of Ethiopia Economic overview 1 S ince 2004 (Ethiopian Fiscal Year (EFY) and government-led development investments. The 1997), Ethiopia has experienced strong and initial double digits growth rates have now manifested generally broad-based real economic growth slightly lower but remain at high single-digit levels. of around 10.6 percent on average between then The economy is expected to stabilize at around seven and 2011(Figure 1). Growth over the last nine years to eight percent in 2012, largely owing to improved was far beyond the growth rates recorded in aggre- performance in the agriculture sector. GDP growth is gate terms for Sub-Saharan Africa (SSA), which on likely to stay around that margin up until 2016 (EFY average only reached 5.2 percent, less than half of 2008) driven by rising foreign investment and exports Ethiopia’s average real GDP growth rate during that (Economist Intelligence Unit 2012). period. Inspired by the East Asian experiences (see High inflation persists, but is on a slightly Box 1 for a comparison of selected indicators and decreasing trend. Headline inflation rates, which policies of Ethiopia and China/Korea),1 growth was were beyond 33 percent on average over the full induced through a mix of factors including agri- year 2011, started to ease at the end of last year, and cultural modernization, the development of new reached 29.8 percent in April, stood at 20.8 percent export sectors, strong global commodity demand, in June and 20.2 percent in August 2012. It was 15.8 percent in October 2012. The reduction in inflation rates reflects the tightening fiscal stance and monetary Figure 1: Ethiopia GDP Growth Rates, 2004 base growth induced by the Government of Ethiopia to 2011 (GoE) in the second and third quarters of 2011. 16% Economic growth brought with it positive 14% trends in reducing poverty, in both urban and 12% rural areas. While 38.7 percent of Ethiopians 10% 8% lived in extreme poverty in 2004/05, five years 6% later this was 29.6 percent, which is a decrease 4% of 9.1 percentage points as measured according 2% the national poverty line, of less than US$0.6 per 0% –2% day (see Table 1). Using the Growth and Trans- 2004 2005 2006 2007 2008 2009 2010 2011 formation Plan (GTP), the target is to reduce this Ethiopia real GDP growth (annual %) further to 22.2 percent by 2014/15, which would Ethiopia real GDP per capita growth (annual %) be another decrease of 7.4 percentage points over SSA real GDP growth (annual %) five years. There is some concern that the recent SSA real GDP per capita growth (annual %) and high inflation rates would reverse some of this Source: World Bank, World Development Indicators (2012); and IMF, World Economic Outlook (2011). Note: World Bank staff estimate for SSA real GDP per capita growth in 2011. Government growth estimates and projections differ for 2010/11 and beyond and are above 11% per year (e.g.: 2010/11 estimate 11.4%, and 2012/13). 1 Korea in this document refers to the Republic of Korea. 2 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Table 1: Population below the National Poverty Line (less than US$0.6 per day) 1995/96 1999/00 2004/05 2010/11 Urban (%) 33.2 36.9 35.1 25.7 Rural (%) 47.5 45.4 39.3 30.4 Total (%) 45.5 44.2 38.7 29.6 Number of poor 25.6 million 28.1 million 27.5 million 25 million Number of food-poor 27.9 million 26.6 million 27.0 million 28.4 million Source: HICES (several issues). remarkable success made over the past five years fruits and vegetables all show very strong increases (Section II explores this issue). of volume over the period under consideration. A welcome development is also that (agriculture- The External Sector based) light manufacturing products are among the growing segments in the markets (middle Ethiopia follows a strategy of increasing exports part of Figure 2). Meat and meat products, and to facilitate growth. This is appropriate given textile and textile products show growth rates of the currently limited size of its domestic mar- 25 to 40 percent in 2011/12 compared to 2010/11 ket and it is consistent with the development (year-on-year). Ethiopia’s services exports, often experience of some of the recently successful overlooked, show a similarly strong performance, countries, particularly in East Asia (Box 1 and in fact Ethiopia is one of the few large, land- and Table 2). Ethiopian goods exports showed locked economies in the world that exports more a growth of 14.8 percent in EFY 2011/12, com- services than goods (Box 2). pared to the previous EFY 2010/11. Figure 2 shows the composition of these exports by commodity group (upper part of Figure 2). Coffee continues to be the largest export, now closely followed by Figure 2: Ethiopian Exports by Commodity gold, with the latter’s rise against the backdrop Groups, 2010/11 vs. 2011/12 of increasing unit prices for gold, while coffee’s Composition, in US$ million decline is driven by a sharp decrease in volume. 3,500 Coffee’s decline may be the result of the new ship- 3,000 ment regulation implemented in November, but 2,500 2,000 revoked in December 2011, which had mandated 1,500 coffee to be shipped in bulk containers rather than 1,000 the preferred 60 kg packs. It is to be expected that 500 coffee volume will rebound. 0 Export of goods growth is to a good extent 2010/11 2011/12 driven by volume growth across a variety of Bees Wax Pulses Chat product groups, which indicates that this Fruits & Vegetables Live Animals Oil Seeds Textile & Textile Products Others Gold growth is a result of recent efforts to increase Meat & Meat Products Flower Coffee Leather & Leather Products and diversify the export base (lowest part of Figure 2). Oil seeds, flowers, live animals, as well as (continued on next page) Growth by Product Group Oil Seeds Fruits & Vegetables Live Animals Textile & Textile Products Gold 1,500 1,000 500 0 2010/11 2011/12 Economic overview 3 Bees Wax Pulses Chat Fruits & Vegetables Live Animals Oil Seeds Textile & Textile Products Others Gold Figure Meat & 2: Ethiopian Meat Products exports by commodity Flower Coffee Overall export and import developments Leather & Leather Products groups, 2010/11 vs. 2011/12 (continued) result in a significantly increased trade deficit by Growth by Product Group 43 percent, up from US$5.5 billion in 2010/11 to US$7.9 billion. The trade balance is in deficit Oil Seeds Fruits & Vegetables due to the large recovery of imports (33 percent) Live Animals Textile & Textile Products against a rise in exports of 14.8 percent. Likewise, Gold the current account balance is negative, even Meat & Meat Products Pulses though it had improved temporarily in 2010/11, Flower when it recorded a small surplus. This was owed Leather & Leather Products Chat to a significant decline in imports of capital and Coffee consumer goods, reflecting a temporary slowing of –10% 0% 10% 20% 30% 40% 50% investment activities. Yet, all import categories have recovered since. Thus, the current account turned Changes in Unit Price and Volume by selected Product Groups into deficit again (around US$2.4 billion), and is Oil Seeds Fruits & Vegetables likely to stay there given the import requirements Live Animals of the GTP and increasing demand for imported Flower Gold consumer goods. The current account balance also Meat & Meat Products represents the excess of investment over savings (see Pulses Table 2), which is substantial given the very low Chat Coffee level of domestic savings (Box 1). Leather & Leather The nominal bilateral exchange to the US$ hit –20% 0% 20% 40% 60% its preliminary high in October 2012 with Birr/ Change in Unit Price Change in Volume US$ 18.0. The rate keeps on its steady deprecation Source: MoFED and NBE, as part of the Macro Update on the Current Inflation in Ethiopia. Figure 3: Ethiopian Imports, by Import Categories, 2010/11 vs. 2011/12 Ethiopian annual goods imports increased by 90% 33.5 percent in 2011/12. Capital goods is the only 80% major import category that recorded less than 8 70% percent growth (Figure 3). Within capital goods, 60% industrial goods showed a marginal increase of 1.3 50% percent, while transport and agricultural capital goods 40% increased strongly with 17.7 and 87.7 percent, respec- 30% tively. The latter is an important driver to facilitate 20% export diversification into agricultural based light 10% manufacturing. The growth rate recorded for con- 0% Raw Materials Semi-finished Fuel Capital Goods Consumer Miscellaneous sumer goods is high with 54 percent. This, however, Goods Goods is mainly driven by non-durable food imports such as cereals (increase of 233 percent) and other food items (increase by 53 percent). The largest portion Growth by categories (in %) Overall growth (in %) of durable imports are vehicles, which increased by Source: MoFED and NBE, as part of the Macro Update on the 24 percent. Current Inflation in Ethiopia. 4 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Box 1: Ethiopia’s Development Experience Compared to China and Korea: Selected Indicators and Policies Overview those comparisons. Table 2 shows a total of seven indicators in Ethiopia between 2004 and 2010—the seven years in which Ethiopia’s development policies are clearly inspired by the GDP per capita grew at the very high level of 8.6 percent per success of East Asian countries, and particular the two high year. This is unique for Ethiopia, but not unusually high in the growth countries from the past and the present, the Republic overall development experience of countries. China I from 1982 of Korea and the People’s Republic of China (New African to 1988 and China II from 1991 to 2010 showed 9.9 and 9.6 2011 and The Africa Report 2012). Given this focus, insight is percent GDP per capita growth respectively. Likewise, Korea provided by a look at some of the indicators and policies that reached GDP per capita rates of 7.0 and 7.3 percent during the are currently carried out in Ethiopia and comparing them to 1968 and 1979 period (Korea I) as well as from 1982 to 1996 the same outcomes during growth periods of Korea and China (Korea II). All these growth periods were identified by consecutive (and for reference also to Vietnam and the US). This box makes years of GDP per capita growth exceeding five percent per year. Table 2: Ethiopia, China, Korea, and Vietnam: Main Growth Periods and their Indicators Foreign Direct Inflation rate* Investment** Investment** reserves*** savings** Exports** domestic GDP per Country capita* Gross Time Total Ethiopia 2004–2010 8.6 15.1 12.9 22.9 2.0 3.2 2.5 Korea I 1968–1979 7.0 14.6 22.6 26.8 0.2 21.1 na (Break in 1972/75) Korea II 1982–1996 7.3 5.2 30.9 33.1 0.3 34.3 1.8 (Break in 1992) China I 1982–1988 9.9 na 11.5 30.0 0.6 35.7 7.7 China II 1991–2010 9.6 4.8 26.0 37.0 3.9 44.1. 11.1 Vietnam 2000–2010 6.0 6.9 66.8 33.1 5.7 28.3 2.7 (Break in 2009) Memo: US 1970–2010 1.8 4.5 9.4 18.6 0.9 17.0 2.6 SE Asia**** 1981–2000 4.3 5.3 78.7 31.5 4.5 36.0 4.1 Source: World Bank staff own calculations, based on World Development Indicators (WDI). Notes: Time definition: GDP per capita growth exceeding 5 percent per year. * average % per year ** average per year in % of GDP *** in months of export **** Indonesia, Singapore, Malaysia, and Thailand Investment seven years under consideration. This is just short of investment rates in Korea I, but it is markedly lower than investment rates Investment rates (measured by gross fixed capital formation) in Korea II, China I and China II, which were all above 30 reached an average of 22.9 percent of GDP per year of the percent of GDP . It is clear from Table 2, that high investment (continued on next page) Economic overview 5 Box 1: Ethiopia’s Development Experience Compared to China and Korea: Selected Indicators and Policies (continued) FDI policies in China in the 1990s Since 1978, China has implemented a series of policies and regulations to attract foreign direct investment (FDI). With rapid economic development and lessons learned, the focus of these policies evolved from quantity to quality. At the beginning of Reform and Opening-up Policy in late 70s to early 80s, China’s strategy was attracting FDI by establishing Special Economic Zones with preferential policies towards FDI. This period is characterized by a top-down policy targeting at reform and opening up. From late 80s to early 90s, China’s FDI policy focused on increasing its volume by establishing more Economic Development Zones and formulating preferential policies in the aspects of tax, land leasing fees, labor fees and profit distribution. Since 1993, China became the second largest outward FDI destination country after the US. Since that time, China’s strategy on FDI changed so that the focus was on the quality of FDI. Policies were formulated to direct FDI into strategic sectors, including agriculture, energy, transport, new material, and new technology. In addition, special attention was paid to setting up research and development centers for technology transfer. From the joining of the World Trade Organization (WTO) to the present, China has emphasized on attracting new technology, human capital and management methods. Since 2010, in response to the initiative on industrial upgrade, new policies are directing FDI to provinces in Western China. Source: Xiao and Zhao (1998); MOFCOM (2012); and China FDI (2012). ratios in China and Korea were (and are) driven by high gross largely decentralized character that allowed local authorities to savings rates—in China II beyond 40 percent of GDP , and attract foreign investors through localized incentives. While this beyond 30 percent in both Korea I and Korea II. In contrast, is partly also the case in Ethiopia, where regional governments Ethiopia’s gross savings rate was a mere 3.2 percent in the can provide variations in local incentive packages, such as in period 2004 to 2010. Total reserves are relatively lower than in the area of land access, the general FDI regime seems to be China, and by not small margins, mainly because the Chinese more centralized in nature. This is a difference by design due reserve accumulation was driven by a heavily managed and to the rather low capacity of regional governments; in fact, somewhat undervalued exchange rate. In Ethiopia, however, according to the federal government, regional states requested the exchange rate tends to be overvalued and hence has been the federal level to administer FDI issues to overcome those on a deprecation path since 2007/08, with limitation of further capacity constraints. Another difference is the openness of depreciation due to the practice of selling foreign exchange in the economy as such. In China in the 1990s, geographical the market to soak up liquidity in much of 2011/12. and sectoral restrictions had been largely eliminated, while in Ethiopia 25 sectors are still closed for foreign investments. Foreign Direct Investment (FDI) It is to be expected, however, that an opening up of various sectors could come about once the ongoing WTO accession FDI in percentage of GDP in Ethiopia is at a relatively mid-level negotiations of Ethiopia are finalized. of 2.0 percent, higher than in Korea I and China I, both times of Based on policies observed in China in the 1990s, there lower levels of globalization and generally lower FDI levels across are a series of “quick-wins� that could be used to increase the world than nowadays observable. During the overlapping the potential for FDI inflows into Ethiopia. These could entail: period of China II with Ethiopia, however, there is a significant decentralizing the approval authority of small-scale FDI projects gap between the two countries: China’s FDI in percentage of GDP to the provincial level; introducing more discretional power for reached 3.9 percent, which was double the value of Ethiopia’s. local levels to negotiate the terms and incentives; providing Vietnam stands out with 5.7 percent of GDP for incoming FDI more longer-term incentives such as favorable taxation in between 2000 and 2010. FDI is not only important to sustain special economic zones (SEZs), which currently is time bound; high investment rates, a key feature of the East Asian development encouraging more joint-ventures to happen between state- experience, but also for knowledge and technology transfer. owned enterprises (SOEs) and foreign investment enterprises Table 3 compares the FDI policies of the 1990s in China (FIEs) to enhance technological transfer; and to broaden the with the current policies in Ethiopia. FDI policies in China had a FDI base by opening-up more sectors (continued on next page) 6 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Box 1: Ethiopia’s Development Experience Compared to China and Korea: Selected Indicators and Policies (continued) Table 3: A Comparison of Chinese and Ethiopian FDI Regulation China in 1990s Ethiopia now Motivation and driver for • Private sector constrained by credits • Low savings rates in the economy at large the government to attract • Technology transfer • Technology transfer more FDI • Focus on strategic industries • Focus on strategic industries Enforcement of FDI • Decentralized FDI approval framework • Rather centralized FDI approval framework regulations • Discretional administrative framework, • Rather centralized administration of incentives, such as favorable taxation in SEZs which are largely time bound (e.g. two-year tax holidays) Benefits for foreign invested • Encourage joint-ventures (JVs) between • Lower capital requirement for JVs, but JVs barely enterprises (FIE) state-owned enterprises (SOE) and FIEs happen between SOEs and FIEs • Removable of geographic and sectoral • 25 sectors are still closed for foreign investors restrictions on the FDI activities • Guarantee against expropriation • Improved private property rights protection Source: World Bank staff compilation, based on Huang 1998, and EthioInvest 2012. Domestic Savings interest rate, indicates that persistent negative real interest rates in fact have reduced the private savings rate further. In Developments in the savings rate in Ethiopia is one of the areas Korea, on the other hand, the authorities managed to maintain in Table 2 where a fundamental difference to the Chinese positive real interest rates through much of the development and Korean growth periods can be observed. Gross domestic period by keeping inflation low and flexibly adjusting nominal savings—as reported in the World Development Indicators—in interest rates; and this was achieved even during the two oil both East Asian countries were beyond 25 percent of GDP shock periods in the 1970s. Looking at China, gives another and even close to 45 percent of GDP in the case of China example of keeping inflation low even during very high growth II (in Vietnam 28.3 percent of GDP). In contrast, in Ethiopia periods. The write-up of the main text about inflation provides the gross savings rate was a mere 3.2 percent in the period some insights on that above. between 2004 and 2010 (there are indications that this has increased to levels between 9 and 13 percent in 2011/12). On Trade and Industrial Policy the other hand, and given the ambitious investment program in the current GTP , which is clearly inspired by the East Asian Exports played a major role in the East Asian development experience, the investment rates in Ethiopia are expected to get experience, and developing a larger export base in an overall closer to the ones observed in China and Korea. To achieve this, market system provides a unique opportunity for Ethiopia. But however, more domestic resource mobilization than in the past Table 2 shows that the country’s exports measured in percent is required. A higher gross domestic savings rate would be an of GDP really lack Korea’s experiences, and also China’s integral part of such an approach. In fact, it will be impossible development path since the 1990s. China’s exports in the 1980s to follow the East Asian experience of high investment-cum- were not higher than Ethiopia’s are, which is not surprising given state-led development policy if the savings rate is close to the the large focus on domestic agriculture in China in the 1980s. one in the US. Ethiopia has had a similar focus on the beginning of its reforms Academic literature widely agrees that saving behavior a decade ago, but it is time to shift gears for another level. can be explained by income growth beyond subsistence Section III of this Economic Update provides various insights level of consumption, demographic change and, to a lesser of a recent analysis carried out to identify growth sectors in degree, the level of financial development. The superb saving light manufacturing as a way to respond to the relatively low performances in China and Korea can be mostly explained levels of exports in Ethiopia. There are six sectors (garment, by these factors. Even though Ethiopia achieved high income leather products, wood products, metal products, wheat flour, growth it is striking that the country has not reached the and processed milk), which could lead the way based on the subsistence level beyond which people start to save in earnest. insight that industries can be developed if a country identifies Much more, the level of financial development in Ethiopia, and follows its comparative advantage in an overall market often measured by the extent of monetization and the real environment. Economic overview 7 Figure 4: Development of the Exchange Rate in Ethiopia Nominal bilateral Birr/US$ (end of period), Jul 2000 to Oct 2012 Real and Nominal Effective Exchange Rates, Sep 2008 to Dec 2011 20 300 18 16 250 14 200 12 10 150 8 6 100 4 50 2 0 0 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Sep-08 Dec-08 Mar-09 Jun-09 Sep-09 Dec-09 Mar-10 Jun-10 Sep-10 Dec-10 Mar-11 Jun-11 Sep-11 Dec-11 Birr/USD (end of period) REER NEER Source: MoFED and NBE, as part of the Macro Update on the Current Inflation in Ethiopia. path since September 2010, which followed the 20 reforms are ongoing to strengthen the expenditure percent nominal deprecation in August 2010. Overall side. This includes plans to move to a program-based the Birr has depreciated by 100 percent since January budgeting approach. Public debt is on a declining 2007 (left side of Figure 4). These developments are trend and expected to be below 35 percent of GDP in largely mirrored in the Nominal Effective Exchange 2011/12. The latest sustainability analysis for external Rate (NEER), which shows a steady path of depre- debt carried out in 2012 indicates that Ethiopia stays ciation between September 2008 and December at a low risk of external debt distress—similar to the 2011 (right side of Figure 4). The Real Effective findings in 2009/10 and 2010/11. Exchange Rate, however, is on a rising appreciating trend since December 2010, which coincides with The Growth and Transformation Plan and largely originates in the recently observed high (GTP) inflation rates. The GTP sets out a series of ambitious targets for Fiscal Policy the country to achieve between 2010 and 2015 (Table 6). Key to success for the plan is to find a way Ethiopia’s fiscal performance appears to be adequate to sustain the extraordinary high financing require- given the current state of the economy and financ- ments—US$57.4 billion2 from 2010/11 to 2014/15 ing requirements for development (Figure 6). The for on-budget capital and off-budget other expendi- overall general government deficit (including grants) ture—for the plan to be implemented. Most promis- in 2010/11 was 1.6 percent of GDP and improved to ing of options is to leverage high public investment 1.2 percent in 2011/12. To a large extent the deficits and attract more private domestic and foreign sources. are financed through external project loans, and to Opportunities to reach this include to improvement a lesser extent also from receipts through privatiza- of the investment climate and the creation of greater tion. On the revenue side, a tax reform was started in room for private sector activity to multiply the 2010 and is ongoing but already showing significant improvements in tax collection. At the same time, implementation of public financial management 2 As indicated in the GTP, and using an exchange rate of 17 Birr/US$. 8 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Box 2: Ethiopia’s Services Exports Ethiopia is one of the few large, land-locked economies in the above the sample average conditional on the level of per world that exports more services than goods (left side of Figure capita income (despite a slight decrease in Ethiopia’s services 5). Also, in 2010 Ethiopia’s services exports as a percentage exports between 2009 and 2010). Yet, there is widespread of GDP were higher than the ratios registered by countries perception that the comparative advantage of a low-income at similar levels of development (right side of Figure 5). In country like Ethiopia lies in export of primary products and the scatter diagram that plots each country‘s overall services labor intensive, low-skill manufacturing goods. There has openness measured as the percentage of services exports in been, however, no rigorous assessment of Ethiopia’s capacity GDP against its per capita income Ethiopia lies above the to become a niche player in global services market. trend line, implying that the country’s services exports are Figure 5: Services Exports from Ethiopia Ethiopia’s services exports and imports as shares of total exports and imports, 2000–2010 Service export openness and per capita income, 2010 70% 1 Service Export Openness (%) 60% 50% 0.1 40% 0.01 Ethiopia 30% 20% 0.001 10% 0% 0.0001 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 0.05 0.5 5 50 % Serv X/Total X %Serv M/Total M GDP Per Capita (1,000 of Current USD) Source: World Bank National Account Data (2012); Source: World Bank staff own calculations, based on IMF Balance of and IMF Balance of Payment Statistics (2011). Payment Statistics (2011) and World Bank WDI (2012) At this stage of development, Ethiopia’s services exports 2000 to 2010, while Table 5 shows the revealed comparative consist mainly of travel and transportation services, whereas the advantage (RCA) of Ethiopia’s services exports in selected country tends to be less competitive in exporting higher value subsectors. These facts confirm that Ethiopia has potential to added services such as computer or business services. Table develop its trade in services but the country has only just begun 4 presents the sectoral composition of services exports and to take advantage of the growing opportunities awarded by the compound growth rate of several service subsectors from global trade in business services. Table 4: Total Service Trade and Service Subsector as a Share of Total Trade, 2000–2010 Service sectors 2000 2005 2010 Compound annual growth rates 2000–2010 Total Services exports (millions US$) 506 1012 2244 16.1 Transportation (%) 42.5 46.0 52.4 18.5 Travel (%) 11.2 16.6 23.3 24.8 Communication services (%) 3.5 4.1 4.5 19.0 Construction services (%) 2.1 1.2 0.6 3.0 Insurance services (%) 0.2 0.5 0.2 12.5 Financial services (%) 0.7 2.5 0.0 –25.4 Computer & information services (%) 0.1 0.0 0.0 4.6 Other business services (%) 15.8 6.9 7.7 7.9 Source: World Bank staff own calculations, based on IMF Balance of Payment Statistics (2011). (continued on next page) Economic overview 9 Box 2: Ethiopia’s Services Exports (continued) Table 5: Revealed Comparative Advantage of Ethiopia’s Services Exports in Selected Subsectors Service sectors RCA Transportation 42.5 Travel 11.2 Communication services 3.5 Financial services 0.7 Computer & information services 15.8 Source: World Bank staff own calculations, based on IMF Balance of Payment Statistics (2011). impact of the large public infrastructure program on in making investment decisions. More decentralized growth, employment, and tax revenue. Likewise, more decision-making of FDI projects could tap more into FDI inflows could fill some of the gap in domestic the local specifics through the provision of localized resources. In both areas—unleashing the private sector incentives. Likewise, the existence of a large number and attracting FDI—the experiences in East Asia, as of sectors still closed to foreign investment could be discussed in Box 1, provide unique insights to better rethought to provide greater opportunities for invest- manage the transition. ment from foreign sources. There is an opportunity to adjust some poli- There is another opportunity to trigger higher cies that are framing the environment for FDI domestic savings through efforts to increase the real into Ethiopia. Compared to China, for instance, in interest rate, which currently is negative. In fact, Ethiopia many of the incentives established for foreign it will be impossible to follow the East Asian experi- investments seem to be more centralized nature, which ence of high investment-cum-state-led development may inhibit ability to respond flexibly to foreign needs policy, if the savings rate stays close to the one in the Figure 6: Ethiopia’s fiscal deficit and debt, 2009/10 to 2011/12 Fiscal deficit (% of GDP) Public debt (% of GDP) 6% 40% 5.3% 39.0% 4.8% 39% 5% 4.6% 38% 37.4% 4% 37% 36% 3% 35% 34.3% 2% 1.6% 34% 1.3% 1.2% 33% 1% 32% 0% 31% 09/10 10/11 11/12 09/10 10/11 11/12 Fiscal balance (incl. grants) Fiscal balance (excl. grants) Public debt Source: International Monetary Fund; and MoFED and NBE, as part of the Macro Update on the Current Inflation in Ethiopia. 10 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Table 6: Selected Macroeconomic Targets of the GTP Targets Start (2010) End (2015) Real GDP growth 10.1 11.5 Real Per capita income (US$) 235.0 354.7 Inflation (%) 36.0 < 10.0 Share of total consumption in GDP (%) 90.6 82.6 Share of total national investment in GDP (%) 23.7 31.5 Share of total export in GDP (%) 10.5 12.5 Share of total import in GDP (%) 273.0 26.7 Share of national saving in GDP (%) 9.4 17.4 Share of national revenue in GDP (%) 12.9 17.3 Share of tax revenue in GDP (%) 9.7 15.3 Share of total poverty reduction expenditure in GDP (%) 12.5 14.2 Percentage of the population living below poverty line 29.6 22.2 Source: Growth and Transformation Plan of Ethiopia. US. Increasing the real interest rate would not only over the ten months from July 2011 to May 2012, require higher nominal interest rates, but also better base money declined by a total of ETB 5.9 billion management of those rates to allow setting interest or 8.5 percent. This was largely driven by a decline rates according to changes in the prevalent economic of the credits from the National Bank of Ethiopia to situation. Granting the central bank more flexibility the Government. in the management of the nominal interest rate is, for While broad money growth has lowered its rate example, one of the Korean ingredients to maintaining of increase compared to the previous year, it did not its relatively high savings rate. Another insight from follow the declining trend seen in base money growth. China is that a holistic plan for managing inflation Still, the high growth of broad money in 2010/11 was makes a big difference during a transition period that stopped (right side of Figure 7). Over ten months could possibly extent beyond the central bank. In fact, from July 2011 to May 2012, base money increased by given the current stage of Ethiopia’s development the 21 percent compared to 31 percent between November Chinese experiences especially in the 1990s could be 2010 and June 2011. Still, the rates of increase in broad very relevant. money originate from continued strong credit growth to public enterprises. Monetary Policy The lowering of the reserve requirement ratio by 5 percentage points is also indicative of a slightly Starting July 2011 a policy was implemented to looser stance of monetary policy. This move, which use base money (reserve money) as a nominal occurred in January 2012, may weaken the tightening anchor for monetary policy and to stabilize base effects on base money as it frees resources to increase money growth for better inflation control. As a lending through the banking system. This is a welcome result, over the second half of 2011 the base money policy change for private banks, which are subject to a growth has reversed its previous trend (left side of restrictive policy that has, since April 2011, required Figure 7), after having increased by more than 42 them to hold certain levels of central bank bills. But it percent from July 2010 to June 2011. In contrast, will also enable the Commercial Bank of Ethiopia to Economic overview 11 Figure 7: Base and Broad Money Growth in Ethiopia, Jul/Nov 2010 to May 2012 Base money growth Broad money growth 80,000 200,000 70,000 180,000 160,000 60,000 140,000 50,000 120,000 40,000 100,000 30,000 80,000 60,000 20,000 40,000 10,000 20,000 0 0 Jul-10 Jun-11 Jul-11 Aug-11 Sep-11 Oct-11 Nov-11 Dec-11 Jan-12 Feb-12 Mar-12 Apr-12 May-12 Nov-10 Jun-11 Jul-11 Aug-11 Sep-11 Oct-11 Nov-11 Dec-11 Jan-12 Feb-12 Mar-12 Apr-12 May-12 Currency in circulation Commercial Bank reserves with NBE Money Quasi money Broad money Base money (reserve money) Source: MoFED and NBE, as part of the Macro Update on the Current Inflation in Ethiopia. increase its lending volume, which in turn may refuel 2011, core inflation was between 10 and 20 percent broad money growth. in 34 months, and above 20 percent in 25 months. Non-food inflation could potentially have several Inflation Rate drivers in Ethiopia and may include: the overall relatively high monetary growth in the system; Headline inflation rates in Ethiopia are largely the role of inflation expectations (which could be driven by developments in food price inflation. further fuelled by food price inflation); and cur- Since 2005, there are two major spikes observ- rency devaluation, which increases the prices for able—in 2008 and 2011, respectively (Figure 8). imported goods. In 2008, headline inflation peaked at 61.6 percent in Recent high inflation rates are a particular August, driven by 79.2 percent in food price inflation; concern for Ethiopia and its economy. Inflation likewise, inflation in 2011 peaked at 40.7 and 40.2 has a negative impact on poverty, possibly pushing percent in August and September, respectively, at a poverty up by as much as 0.5 percent for each increase time when food inflation reached almost 50 percent. in the inflation rate by 1 percent (Lopez, 2004). Fur- Food inflation kept rising to more than 50 percent in ther, inflation negatively affects the domestic savings November 2011; but then headline inflation started rate, which for the past decade (2000–10) has been to ease slowly already, driven by a tightening stance of in single-digit levels (Table 2), as well as the private monetary conditions (see Figure 7, which shows the investment rate, both of which are vital to finance stabilization of base money growth at the end of 2011). the much needed infrastructure in Ethiopia. Finally, Looking at non-food inflation, a measure of it is important to note that high growth rates and low core inflation shows a relatively high “socket� inflation rates are possible—as for instance is shown in level, which indicates that inflation in Ethiopia many of the East Asian experiences in Table 2. Look- is not purely a food price problem. In both spikes ing specifically at the case of China, Figure 9 provides of 2008 and 2011 non-food inflation passed 20 per- a close illustration of this and shows that there really cent, yet did not hit the 30 percent marker. In the does not need to be a conflict between growth and 84 months between January 2005 and December inflation. 12 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Figure 8: Inflation Rate in Ethiopia, Jan 2005 to Oct 2012 100% 80% 60% 40% 20% 0% –20% Jan-05 Mar-05 May-05 Jul-05 Sep-05 Nov-05 Jan-06 Mar-06 May-06 Jul-06 Sep-06 Nov-06 Jan-07 Mar-07 May-07 Jul-07 Sep-07 Nov-07 Jan-08 Mar-08 May-08 Jul-08 Sep-08 Nov-08 Jan-09 Mar-09 May-09 Jul-09 Sep-09 Nov-09 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 Food inflation Non-food inflation Headline inflation Source: Central Statistics Agency of Ethiopia (CSA). Comparing experiences in Ethiopia again Figure 9: Inflation and GDP Growth Rates with developments in East Asia at times of simi- in China, 1991–2010 lar high growth rates shows that inflation rates 20% in Ethiopia at the moment are indeed at a very high end (Box 1). In Korea, inflation between 1982 15% and 1996 was a mere 5.2 percent at a time with 10% GDP per capita growth rates on average reaching 7.3 percent (Table 2). Likewise, in China between 5% 1991 and 2010, inflation rates reached 4.8 percent with GDP per capita growing at 9.6 percent, even 0% higher than the current rates in Ethiopia at 8.6 percent. This creates an opportunity to look at the –5% 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 developments there and draw lessons for Ethiopia on how to manage a high growth economy and Inflation rate (%) GDP growth (%) at the same time keep inflation under control. In Source: World Bank, World Development Indicators. China the authorities, over much of the 1990s and Note: 3 year moving average. into the new millennium, were able to manage the inflation rate with non-orthodox instruments that essentially freed monetary policy from the strict of the central bank, and there are non-central bank obligation of inflation control without sacrificing policy instruments. The former includes price-based the inflation target as such (Flassbeck, Dullien, indirect and quantity-based direct instruments, and and Geiger 2005). the latter includes non-central bank policy instruments, For instance in China, in order to keep inflation especially in the earlier years of the 1990s. There is low in a high growth environment, monetary policy no doubt that with the ever-increasing marketization applied (and to some extent still applies) two broad of the economy in China, the simultaneous usage of sets of monetary policy instruments in an overall all these instruments leads to various distortions that market environment but against the backdrop of a ultimately prevent monetary transmission to function tightly managed exchange rate: There are instruments properly and that a shift away from the current hybrid Economic overview 13 model is needed (Geiger 2010). However, given the development period and especially in the 1990s could current stage of Ethiopia’s development, the Chinese well be very relevant. experiences in controlling inflation over much of its In depth: Inflation and its impact on the poor 2 Background Scope of this Analysis As was described in the earlier section, Ethiopia • July 2010 to October 2011 experienced a strong acceleration in inflation in • Data from the 2004/05 WMS-HICES, the latest nationally representative household survey available 2010 and 2011. While inflationary pressures have for this analysis. slightly, it is important to analyze the impact of the • No consideration of substitution effects (consumers’ recent high inflation on poverty reduction efforts in demand) and second round effects (change in labor market participation, in saving and investment patterns, Ethiopia. For the illustration and analysis in this sec- in production costs, and in returns to economic activities tion, data is being used from July 2010 to October due to inflation). 2011. During this period national consumer retail • Comparison of 2007/08 situation (analyzed in Ticci, 2011) with 2010/11. prices increased by 48 percent; rates at this magnitude call to mind the 2008 price shock; back then consumer prices rose by 72 percent between June 2007 and Sep- In Ethiopia, and in line with evidence of other tember 2008. Comparing both situations—2011 and countries, the negative effects of food price shocks 2008—food prices saw even more marked upswings are usually more evident in urban areas where with increases by 60 and 100 percent, respectively. people are net food consumers; this is particularly High inflation poses a serious threat to pov- true for the poor and the near-poor who spend erty reduction and food security. High inflation larger shares of their budget on food. For instance, causes protracted stress on household budgets a food security assessments conducted by the World and increases the vulnerability of any household, Food Program in Addis Ababa in 2008 indicated that especially poor ones. In addition, coping behaviors the proportion of households consuming an adequate of households, such as withdrawing children from diet decreased from 64 to 40 percent between January school—even temporarily—might have long-term and July 2008 (World Food Program 2008), which consequences. Moreover, despite some geographical exactly coincides with the past peak of food prices. differentiation, price shocks represent covariate shocks Similarly, analyzing a panel dataset on 567 households to which households and other community enti- ties have only limited ability to respond. Faced with repeated price shocks of such a magnitude in 2008 and 3 Indeed, based on household panel data from 1994 to 2004, Bigsten 2011, Ethiopian households likely had compromised and Shimeles (2008) found that the likelihood of escaping poverty their ability to rely on sustainable and effective coping considerably declines as the time spent in the state of poverty increases. 4 In Ethiopia, an estimated 38.7 percent of the population was living strategies, which in turn may have increased their risk below the official poverty line in 2004/05. Rural Ethiopia, where about of falling into a poverty trap.3 The combination of all 80 percent of the population lives, had a higher poverty incidence (39.3) than urban Ethiopia (35.1). Yet rural areas experienced robust poverty these factors may have hindered(or halted) progress reduction (a decline of nine percentage points since 1995/96), while a in rural poverty reduction, slowed down any improve- strong rise in inequality resulted in a marginal increase in poverty in urban areas (two percentage points since 1995/96). Despite the overall ments in urban poverty reduction, and exacerbated reduction in poverty, the number of poor people grew by an estimated urban inequality.4 1.9 million people in the decade up to 2005. 16 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness in four major cities (Addis Ababa, Awassa, Dessie and percent reported that their food consumption was Mekelle) Alem and Söderbom (2010) find that 89 very negatively affected and 38 percent stated that percent of the households interviewed in 2008/2009 they reduced the quantity of food consumed because regarded food price inflation as the main adverse of the price shock.5 shock facing them between 2004 and 2008; 60 The impact of food price inflation in rural areas is less straightforward, especially in situations where food prices dominate and overshoot overall Box 3: Welfare impact of Recent Price inflation rates. Netsellers of food, for instance, could Shocks in Ethiopia based on Nationally experience an increase in their incomes.6 Rising food Representative Datasets prices could help farmers to accumulate savings and they could generate positive incentives to expand Empirical literature on the welfare impact of recent price shocks based on nationally representative datasets, has not production, and to invest in agricultural productivity led to univocal findings. Using the Almost Ideal Demand and in high value crops. Yet, the positive potential of System model and data from the 2000 WMS/HICES, increasing food prices might be offset by other factors Ulimwengu et al. (2009) derive demand elasticities with respect to income and to food price and indirectly estimate such as growing production costs, uncertain returns the overall change in food consumption and calorie intake due to price volatility, rising non-food prices, higher due to hypothetical food-price increases. They found that shares of net food buyers also in rural areas7 and the welfare losses show a great deal of heterogeneity across regions, are relatively higher in rural than in urban areas concentration of net food sellers among the rural bet- (with the exception of SNNPR) and for a price rise of cereals ter off.8 Overall, however, it is important to note that than for other food items. Household cereal consumption, effects of food price shocks on the welfare of rural and for instance, is estimated to fall by 0.82 and 0.86 percent in rural and urban areas, respectively, for a one percent urban populations are varied and complex (Box 3). increase in the price. Under these conditions, the 2008 cereal price shock might have had severe costs for food The 2010/11 Inflation Situation in security in both rural and urban areas. Loening and Oseni (2007), based on data from the 2000 WMS/HICES and Ethiopia applying a net benefit ratio approach, found that, on average, the rise in food prices occurred between 2000 to After the spike in inflation rates in 2007 and 2008, 2007 was likely to produce a positive effect in rural areas and a negative impact in urban areas but vulnerability to inflation slowed down in 2009 and then remained food inflation varies across expenditure quintile. In rural areas better-off households were likely to benefit relatively more than lower-income households, while in urban centers the middle-income groups were hit the most. In rural areas, 5 Alem and Söderbom’s regression results also indicate that households moreover, the households’ net food market positions are with little assets and workers with uncertain earnings, such as casual crucial in determining the sign of welfare effect due to the workers, were more adversely affected by the 2008 food price shock price increase: net buyer rural households, especially the suggesting that in urban areas the poor were the most vulnerable to the poorest, were most harmed by food inflation, while net food price spurt. 6 Based on a panel data set on 354 households in six rural communities sellers gained from the food price increase. Ticci (2011) in Ethiopia and collected in 1989 and in 1994–1995, Dercon (2006), applied a similar approach but it included the effect of for instance, found that favorable changes in the terms of trade for crop non-food prices and of cereal production increase, used producers (average real producer prices increased by 26 percent) explain more recent data, and focused on the period between 2006 about 60 percent of growth in real food consumption among the house- and 2008. The work concluded that in urban areas, high holds in the sample over the 1989–1995 period. inflation between 2007 and 2008 significantly worsened all 7 The 2005 Poverty Assessment (World Bank, 2005) and Loening and poverty dimensions (incidence, depth, and severity), while Oseni (2007) have drawn attention to the low share of net food sellers even in rural areas. the results were more ambiguous in rural areas: incidence 8 Ticci (2011) finds that, according to data from the 2000 WMS/HICES, of poverty might have decreased but the severity of poverty the share of net cereal sellers is increasing in income. Analogously, based increased even under the most optimistic hypotheses. on the same data source, the analysis carried out by Levinsohn and Mac- According these results, therefore, the main risk in rural Millan (2007) indicates that there are more net buyers of wheat than net areas was a further impoverishment of the poorest. sellers of wheat at all levels of income and the proportion of wheat net sellers is increasing in living standards. In depth: Inflation and its Impact on the Poor 17 Figure 10: Inflation Rates in Ethiopia, Jan 2001 to Oct 2011 Food and non-food inflation rates National Retail Price Indices (end of period) (December 2006 = 100) 100% 300 83.4% 80% 250 60% 51.9% 200 40% 24.2% 22.8% 150 20% 100 0% –20% 50 –40% 0 Jan-01 Aug01 Mar-02 Oct-02 May-03 Dec-03 Jul-04 Feb-05 Sep-05 Apr-06 Nov-06 Jun-07 Jan-08 Aug-08 Mar-09 Oct-09 May-10 Dec-10 Jul-11 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Food inflation Non-food inflation Overall inflation National retail food price index National retail non-food price index Source: Central Statistical Agency of Ethiopia. stable for the larger part of 2010. Within 2009, Figure 11: Food Prices between Jul 2010 food prices showed more pronounced variations and Oct 2011, by Item and Region, than non-food prices: food inflation became nega- Percentage Change tive in May 2009 fluctuating around negative or 250% low positive rates until July 2010 when it started to 200% climb again. Annual food inflation (end-of-period) 150% jumped from –1.9 percent in July 2010 to 8.7 per- 100% cent in December 2010, 38.6 percent in May 2011 50% and to 51.9 percent in October 2011 (Figure 10, left side). Moreover, despite the reversal in inflation 0% Tigray Harari Addis Ababa Dire Dawa Afar Oromyia Somali SNNP Gambela Amahara Benishangul Gumuz in 2009, both food and non-food prices have never returned to levels in line with price dynamics prior to the 2008 peak (Figure 10, right side). In addition, Cereals Pulses Meat Milk, Cheese and Egg the inflation rates are particularly high for those items that relatively weigh more on the budget of the poor. Source: Central Statistical Agency of Ethiopia. Figure 11, shows that, over the period of consideration (July 2010 to October 2011), the price of cereals and had already experienced higher growth rates than in pulses, on average, has grown more rapidly than the several African countries and this gap has further price of other food items such as meat, milk, eggs, and widened since the beginning of 2011. Such a gap sug- cheese; these latter items tend to be more intensively gests that the inflation acceleration might be due to a used by the better off.9 combination of external and country-specific factors. Similarly to the situation observed in 2008, Ethiopia’s domestic food prices in 2011 have increased more rapidly than international food 9 Berhane et al. (2011) find that the main cereals account for 46 and 47 percent of total food expenditure of the bottom 40 percent of income dis- prices and prices in other African countries tribution in urban and rural areas, compared to 36 and 43 percent of food (Figure 12). In 2010, Ethiopian food consumer prices expenditure of the top 60 percent rural and urban population, respectively. 18 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Figure 12: Consumer Prices, Food Indices higher than in most other African countries (FAO- (2000 = 100) GIEWS 2011). 600 500 Distributive Impact of the Rise in 400 Consumer Prices between July 2010 300 and October 2011 200 100 The distributional impact—at the aggregate 0 level—of a rise in consumer prices depends on the Jan-10 Feb-10 Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10 Dec-10 Jan-11 Feb-11 Mar-11 Apr-11 May-11 Jun-11 consumption patterns of different income groups and on the pace at which the prices of different Botswana Burkina Faso, Ouagadougou Gambia, Banjul goods increase. Figure 13 presents the effective infla- Ethiopia Gabon, Libreville Côte d'Ivoire, Abidjan Ghana Madagascar Malawi Mali, Bamako tion experienced by individuals in different quintiles Mozambique Niger, Niamey Nigeria Rwanda of the adult equivalent expenditure from July 2010 Senegal Sierra Leone Uganda to October 2011 and from December 2006 to July Source: FAOSTAT (2011). 2008, based on the consumption bundles households were consuming in 2004/05 net of own-consumed items. The figure shows both the national effective The latter may include monetary policy related factors inflation rates (upper side of Figure 13) as well as the (see Section I), but also factors such as the unusual rural/urban effective rates (lower side of Figure 13). pace of increase in fuel prices in Ethiopia (fuel prices During both periods of high inflation, poorer parts drive crop prices), which at the end of 2011 were of the population faced significantly higher inflation Box 4: Impact of Inflation: Methodology and Data The standard approach (Deaton 1997) to examine the impact simplifications due to data issues. The most recent nationally of the rise in the price of good x is based on the assumption representative survey that is available for this analysis is the that a household will be affected in proportion to the share of HICES survey from 2004/2005, which was collected by the its total expenditure devoted to good x. Such an impact will be Central Statistical Agency of Ethiopia (CSA). The survey provides at least partially offset if the household produces good x or if detailed information on household expenditures, but income the household’s labor income increases due to the increase in information has not been released and thus limits the analysis on the price of x. This methodology can be generalized to the case the income side. This analysis, therefore analyzes the distributive of overall inflation by summing up the effects of the increases impact of the rise in consumer prices alone, focusing only on in prices of the various goods (or groups of goods). the expenditure side. Since positive effects of inflation on income This approach makes a number of simplifying assumptions sources and households’ defensive strategies are not accounted such as the absence of substitution effects (i.e., adjustments for, the estimates provided can be regarded as the upper limit of in the consumption choices across goods), of changes in the poverty impact caused by the food price shock in 2010/11. marketing behavior of agricultural goods (i.e., the assumption In a low income and largely agricultural country such as that, following the price shock, households relied just as much Ethiopia, many households rely on their own production for a as before on food that they had produced themselves), or in significant share of their consumption, particularly of food. In production. These assumptions make it possible to extrapolate rural areas, where the majority of population lives, in-kind food the impact of inflation on the consumption bundles that consumption represents about 30 percent of total expenditure households were consuming/producing before the inflation and more than half of food expenditure, compared with an spurt, thereby providing a first order approximation of the in-kind food share of four per cent among urban households effects. This analysis follows this approach, but with further (Table 7). (continued on next page) In depth: Inflation and its Impact on the Poor 19 Box 4: Methodology and Data (continued) Table 7: Self-food Expenditure as Share of Total Food Expenditure, 2005 Should the CPI in Ethiopia include households’ own consumption? Quintiles Rural Urban 1 0.53 0.05 For national accounts, the International Labor Organization (ILO) suggests treating “…the goods acquired by 2 0.56 0.06 households on the market for use as inputs into the various 3 0.57 0.06 kinds of households production activities as if they were themselves final outcomes� (ILO, 2004). For the calculation 4 0.56 0.04 of the CPI, however, the ILO suggests that this approach 5 0.52 0.03 is doubtful. In simple terms, the ILO suggests that the CPI records the actual input prices or the imputed output All 0.55 0.04 prices, but not both. Also, if the imputed output prices for Source: HICES-WMS, 2004/2005. “subsistence agriculture� are included in the CPI, the prices Note: Food self-expenditure defined as all in kind food expenditure of purchased inputs should be excluded. from household agricultural enterprises. In Ethiopia, the way the CPI is measured, it appears, incorporates own consumption given how similar results are The analysis excludes own-consumption from the impact of when we compare the weight of the food component of the inflation. The main reason for doing so is that own-consumed CPI to the food expenditure shares with own consumption goods can be seen as “insulated� from fluctuations in market prices (calculated from the household survey). This in principle as they contribute to household budgets from both perspectives of is not a problem, but it would require that acquisitions consumption and income.a This way of proceeding should ensure of goods that are used as inputs in the production of that estimates of the impact of inflation are more plausible. At the the staples for household consumption are eliminated same time, this is likely to affect the relative impact between rural from the calculations. This is quite complex and for that and urban areas of food inflation on consumption only (i.e. whether reason current practice in most countries is not to include rural or urban areas are more affected). Table 8 helps illustrating households’ own consumption, and instead to consider all this point: Excluding food own-consumption lowers the food share goods purchased by the household in the market. by areas and by poverty status. But it results in food expenditure Incorporating own consumption into the calculation of shares which are higher in urban than in rural areas. Finally, it is the CPI has implications not only for the comparison of how worth observing that the poor spend a higher share of their budget much more inflation the poorest households face, but also on food than the non-poor, regardless of the area of residence and for the calculation of the inflation rate in Ethiopia. the treatment of food own-consumption. This implies they are more vulnerable to food price shocks than the non-poor. a It is assumed that all households spend in final goods all additional Table 8: Food Expenditure Share by Area incomes that they earn thanks to real economic growth and inflation and that income rises at the same rate as the nominal GDP . This sce- and Poverty Status nario starts out with an optimistic initial estimate of poverty. In fact, it assumes equally distributed gains of growth, while empirical evidence Poverty Without own- With own- suggest that, in the past, inequality increased reducing the pro-poor status consumption consumption impact of economic growth (DPRD\MOFED 2008). In urban areas, National Non-poor 33.7% 51.4% the opportunities coming from economic expansion between 1995 and 2004 were benefited mainly by the better off who experienced Poor 40.7% 59.4% higher consumption growth rates than the rest of urban population (World Bank, 2010). Thus, sustained economic growth in the decade Rural Non-poor 32.4% 53.0% up to 2004/05 was associated with poverty reduction stagnation and a Poor 38.3% 59.5% marked rise in income inequality in urban areas where Gini coefficient increased by 10 percentage points (from 0.34 to 044). Urban Non-poor 41.6% 42.6% b Of course households can change their consumption and selling Poor 57.2% 58.9% behaviors in response to the different dynamics of consumer and Source: HICES-WMS, 2004/2005. producer prices—ote however that the assumption that they do not Note: Food self-expenditure defined as all in kind food expenditure (which we are implicitly adopting) is consistent with the methodology from household agricultural enterprises. broader assumption that households do not substitute items in their consumption baskets and do not change their selling behavior. 20 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness rates than the richest quintile of the population. In In 2008, the rise in prices was more marked and the the 2006/08 period the difference between the poorest distribution of effective inflation rates was less uniform and the richest quintile was 9.1 percentage points, and with a larger gap between rural and urban areas and in the 2010/11 period the difference was 3.1 percent- between the top and the bottom expenditure quintiles. age points. Overall, rural areas experienced slightly In 2008, therefore, the price shock was probably more lower inflation than urban ones. The rural-urban gap adverse for the poor and the urban households than is less evident for individuals in the highest quintile of the 2011 price acceleration has been up to now. These expenditure distribution.10 Given the current debate differences in effective relative price variation across on the possible need of revisions in the electricity tar- iff structure, Box 5 also gives a short overview of the 10 Note that if own-consumed items were included inflation would be impact of tariff increases on inflation and the poor. higher, ranging from 48 percent (for the second quintile) to 43 percent The current price shock, as that of 2008, is (in the top quintile) in rural areas and from 44–45 percent (for individuals in the lowest three quintiles) to 38 percent in the top quintile in urban likely to hit the poor and the households close to areas. The basic pattern with lower inflation for the top quintile would the poverty line relatively harder than the better off. however be the same. Box 5: Increases of Electricity Tariffs in Ethiopia – Insights on the Possible Effects on the Poor Background on the consumers of EEPCO’s services and especially the poor segments of the population. A short analysis of the three Ethiopia Electric Power Company (EEPCO) is coming under hypothetical tariff increases mentioned—196, 233, and 390 financial pressure in its efforts to serve the ambitious targets percent—shows the potential impact on both inflation and laid out in the GTP; this may create significant fiscal challenges poverty. for the GoE. Under the GTP , EEPCO has been leading an Impact US$11 billion energy sector expansion program, of which US$3.5 billion has already been raised. In the coming years, An increase of the electricity tariff by 196, 233 and 390 percent, these maturing loans may amount to such a large annual debt respectively, is found to have limited effects on inflation and the service obligation that EEPCO will not be able to meet from its poverty incidence. Simulated inflation rates in 2005/06 would current and future operating cash flows. This shortfall would have reached 13.3 percent (196 percent increase in tariff) and then place an additional burden of about US$2 billion on the 13.8 percent (with a 390 percent increase in tariffs) compared GOE’s budget in FY2012–20 (cumulatively), which is equivalent to the actual inflation rate of 12.7 percent. Likewise, doubling to three to four percent of the GOE’s annual budget expenditure. or quadrupling the electricity tariffs would have negligible effects Policy Options on the poverty incidence because most of the poor and near poor do actually not have access to electricity; in 2009, only Alternatively, the shortfall could be met with additional tariff 17 percent of population had access. for services of EEPCO of about US$0.03–0.04/kWh. Such The main findings on poverty impact under the three an increase over time would allow the tariff to reflect full cost scenarios are reported in Table 9. The results suggest that recovery (estimated to be US$ 0.06–0.07/kWh) and thereby the implications of the simulated tariff structure revision ensure financial health of EEPCO and the sector. So to reach in terms of rural poverty incidence, depth and severity are cost recovery levels, this would translate into an increase of 233 minimal. In urban areas, the impacts are higher but they are percent from the current effective electricity tariff rate of US$ still quite limited. A 233 percent increase in electricity prices, 0.03/kWh to US$ 0.06–0.07/kWh. At the current exchange rate, for instance, would lead to a rise by 0.23, 0.27 and 0.11 the revised tariff would be equivalent to 1.21 Birr/kWh, which percentage points in urban headcount poverty, poverty gap would mean a 390 percent increase with respect to the 2005 and squared poverty gap indices, respectively and the impact tariff (0.31 Birr/kWh). But if the proposed tariff adjustment is of a 390 percent rise on the same poverty indices is estimated reported in 2005 prices, instead, it would correspond to 0.61 to be 0.27, 0.32 and 0.13 percentage points. Since most of Birr/kWh, which would be a 196 percent rise. The question people live in rural areas, the effects at national level are also then is how an increase of such a magnitude would impact expected to be marginal. (continued on next page) In depth: Inflation and its Impact on the Poor 21 Box 5: Increases of Electricity Tariffs in Ethiopia – Insights on the Possible Effects on the Poor (continued) Table 9: Simulated Changes in Poverty Indices due to Electricity Tariff Adjustments, Three Scenarios, 2005/06 After the hypothetical Change in percentage Before the rise in tariffs rise in tariffs points Scenario Poverty index Rural Urban National Rural Urban National Rural Urban National 196 Headcount 39.4 35.2 38.8 39.4 35.4 38.8 0.0 0.2 0.0 percent Poverty Index increase Poverty Gap 8.5 7.7 8.4 8.5 8.0 8.4 0.0 0.3 0.0 Index Severity Poverty 2.7 2.5 2.7 2.7 2.7 2.7 0.0 0.1 0.0 Index 233 Headcount 39.4 35.2 38.8 39.4 35.4 38.8 0.0 0.2 0.0 percent Poverty Index increase Poverty Gap 8.5 7.7 8.4 8.5 8.0 8.4 0.0 0.3 0.0 Index Severity Poverty 2.7 2.5 2.7 2.7 2.7 2.7 0.0 0.1 0.0 Index 390 Headcount 39.4 35.2 38.8 39.4 35.4 38.8 0.0 0.3 0.0 percent Poverty Index increase Poverty Gap 8.5 7.7 8.4 8.5 8.0 8.4 0.0 0.3 0.1 Index Severity Poverty 2.7 2.5 2.7 2.7 2.7 2.7 0.0 0.1 0.0 Index Source: World Bank staff calculations, based on HICES 2004/05. Analysis Approach into account demand responses, such as substitution effects, which households could undertake as a reaction to the The impact of the electricity tariff revision is assessed—using increase in electricity price. This simplification does not alter HICES 2004/05 data—through an estimation of the effect on the interpretation of the results since the share of electricity rural and urban poverty rates and the impact on the national expenditures is very low. The counterfactual analysis is set at consumer price index (CPI) due to the three different increases the time of the latest national survey. It therefore estimates what in electricity tariff for residential uses. The analysis focuses on would have been the poverty implications, if the electricity tariff observed consumption patterns in 2004/05 without taking had been revised in 2005. areas and expenditure quintiles are likely to be driven (Figure 8) and food shares are lower for the highest by the incidence of food expenditure in the budgets two quintiles, regardless of the area of residence; of different households. the same is true for rural households once food own- Food inflation presents higher rates than non- consumption is excluded (Table 10). Moreover, com- food inflation in both periods of soaring prices pared to the latest rise in prices, the 2008 price shock 22 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Figure 13: Comparison of Average Effective Inflation Rates National effective inflation rates Inflation rates, by quintile Inflation rates, by quintile Dec 2006 to Jul 2008 July 2010–October 2011 65.5% 65.4% 42.9% 42.8% 63.5% 42.2% 60.7% 41.2% 56.4% 39.8% 1 2 3 4 5 1 2 3 4 5 Quintiles Quintiles Quintile of expenditure Quintile of expenditure Urban and rural effective inflation rates Inflation rates, by quintile Inflation rates, by quintile Dec 2006 to Jul 2008 July 2010–October 2011 75% 76% 75% 69% 63% 65% 63% 58% 54% 57% 44% 40% 43% 40% 44% 39% 42% 37% 36% 38% 1 2 3 4 5 1 2 3 4 5 Quintiles Quintiles Rural Urban Rural Urban Source: World Bank staff calculations, based on HICES 2004/05. Note: Effective inflation rates are based on the consumption bundles that households were consuming in 2004/05 net of own-consumed items, and constant spending patterns and constant own-consumption is assumed since 2004/05. Quintiles are calculated in terms of quintiles of the adult equivalent expenditure. Effective inflation rates refer to the relative price variation in the entire periods from December 2006 and July 2008 and from July 2010 to October 2011. was characterized by a larger divergence between non- Regional location contributes to variations in food prices and food inflation rates. In July 2008, food the effective inflation rates households experienced inflation jumped to more than 90 percent (annual between July 2010 and October 2011. Figure 14 end-of period inflation) while non-food inflation rates details average inflation by region and the minimum stood constantly below 30 percent; and there was a and maximum rates by quintile and region. With the gap between the two in the order of above 50 percent- exception of Harari, in all regions the minimum infla- age points between June and September 2008. In the tion was experienced by the richest quintiles, while in period under consideration 2011, on the contrary, the eight cases the maximum rates were experienced by difference between food and non-food inflation has either of the two bottom quintiles, but regional het- never surpassed 28 percentage points. erogeneity was even higher than differences by quintile. In depth: Inflation and its Impact on the Poor 23 Table 10: Food Expenditure Share by Area even if marketing infrastructure has improved and and Quintiles of the Population transaction costs have been reduced (Rashid 2011), Quintiles Rural Urban National market integration among regions faces continues to face barriers12 with some regional marketsbeing better 1 37.7% 57.3% 40.1% integrated than others.13 In this context, not all regions 2 38.0% 56.3% 40.1% might be equally capable of transmitting or capturing 3 36.6% 54.9% 37.9% price signals and shocks from other regions. Differ- 4 32.6% 48.7% 34.5% ences in weather conditions and, as a consequence, on 5 27.2% 35.6% 29.5% harvests, might also help to explain the regional price All 34.7% 47.1% 36.4% heterogeneity, at least for food products. All these fac- Source: World Bank staff calculations, based on HICES 2004/05. tors can contribute to regional heterogeneity of infla- Note: Food self-expenditure is excluded. tion effective rates, but a better understanding of the differences in price dynamics across regions requires further analysis.14 These regional differences are probably related with a Overall, the impact of rising consumer prices range of factors from the level of market integration on expenditures in both rural and urban areas is and spatial connectivity to the degree of reliance on non-negligible and leads to an estimated increase in local production11 and households’ production for the number of poor people of around 1.8 million.15 meeting their own food requirements. Moreover, Much more, it is to be expected that the consumer price inflation between July 2010 and October 2011 led to a greater increase in both the depth (the aver- Figure 14: Effective Inflation Rates between age difference between the income of the poor and July 2010 and October 2011, by Region, the poverty line) and the severity of poverty (i.e. the Average and Min and Max by Quintile, extent to which some of the poor are very far from the based on HICES 2004/05 poverty line) in urban than in rural areas. The latter is 0.55 likely to be more pronounced in urban areas, however, 0.50 0.45 11 Rashid (2010) notes that Amhara and Oromoyia account for 87 0.40 percent of teff and wheat production, and for 82 percent of maize pro- duction. These surplus regions, therefore, can rely on local production 0.35 to a greater extent than cereal deficit areas that are more dependent on 0.30 cereals transported in from other regions. This might translate on dif- ferent pressure on food prices. Tigray Benishangul Gumuz Somale Amahara SNNP Oromyia Afar Dire Dawa Addis Ababa Harari 12 Ulimwengu et al. (2009), for instance, find no evidence of market integration among Ethiopian regional maize markets. 13 Rashid (2011) finds that the maize, wheat, and teff markets in the center, southwest, and south central areas more integrated than those Minimum inflation rate Max inflation rate in the north and east. Regional inflation rate 14 In the light of the regional differences observed, it is important to note that this analysis relies on regional price indices. More precisely, regional Source: World Bank staff calculations, based on HICES 2004/05 monthly consumer prices are used, disaggregated by item group to deflate Note: Effective inflation rates are based on the consumption bundles consumption expenditure from July 2010 to October 2011. This choice that households were consuming in 2004/05 net of own-consumed is also in line with earlier research (Ulimwengu et al. 2009), which found items, and they assume constant spending patterns and constant a great deal of heterogeneity across regions in terms of consumption and own-consumption since 2004/05. Quintiles are calculated in terms calorie-intake loss due to foodprice increases. of quintiles of the adult equivalent expenditure. For each region, the 15 This is estimated by calculating the elasticity of growth and poverty Figure shows average inflation and the inflation rates experienced by reduction in Ethiopia based on the last 10 years, which is around 0.65. the two quintiles that faced the smallest and greatest effective price rise. This is then compared to an estimated loss in real income for the poorest Effective inflation rates refer to the relative price variation in the entire quintiles by comparing the national effective inflation rates by quintile period from July 2010 to October 2011. (Figure 13) with the national CPI (excluding own consumption). 24 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness suggesting that the urban poorest are among the most The greater poverty impact of food prices vulnerable to the food price shock due to the fact that on poverty in urban than in rural areas is driven urban poor tend to be net food consumers. by the fact that this analysis considers own-con- Comparing these estimates for 2010/11 with sumption as “unaffected� by the price dynamics. earlier findings based on a similar methodology It is noteworthy, however, that even using a defini- (Ticci 2011) shows that the recent surge in prices tion of food consumption which “insulates� rural has had a slightly lower poverty impact than the households more than urban ones from the impact protracted period of rising prices in 2007 and of very high food prices, the overall poverty impact 2008. This is not surprising since the inflation bout is considerable also in rural areas where in-kind food was much more pronounced at that time. But simi- expenditure from household agricultural enterprises lar to the past shock, the largest risk due to the price account for more than half of food expenditure. This acceleration is to lead to further impoverishment of underscores the importance of non-food inflation in the already poor segments of population in urban areas rural areas. and a greater vulnerability to poverty for those near to the poverty line in rural areas. In depth: Agglomeration and competitiveness in Ethiopia 3 S ince its foundation in 1886, Addis Ababa has revenues and desirable investments. It has not drawn emerged as Ethiopia’s main political, admin- sufficient investment to provide jobs for its expand- istrative, economic, financial, and communi- ing population and the weak infrastructure has been a cation center. About 2.7 million people live in Addis deterrent. The problems that have deterred investment Ababa. This is 3.7 percent of Ethiopia’s population are not confined to Addis Ababa, but are evident in low of 73 million, and 22.5 percent of Ethiopia’s urban investment levels throughout the country. population of 12 million. Addis Ababa is an impor- tant center of production, a gateway to information Agglomeration and Economic Growth and technology, and the major internal market for goods and services. To serve its role, Addis Ababa is No country has advanced to middle income sta- transforming to a service economy—financial services, tus without significant urbanization and indus- wholesale and retail trade, hotel and restaurants, trans- trialization (World Bank 2009). Rising densities port and communications, and real estate service—and of human settlements and mobility of workers and the service sector accounts for an estimated 71 percent entrepreneurs shorten the distances to markets and are of Addis Ababa’s output (EEA 2010). essential for economic development and prosperity. Addis Ababa is central to Ethiopia’s growth Growing cities, mobile people, and vigorous internal prospects. It is both the hub that connects inter- and external trade were the catalyst for progress in the nal markets to each other and the gateway to developed world over the last two centuries. And the external markets, making its markets central to same market forces of agglomeration, migration, and determining national competitiveness. For rural specialization are changing the economic landscape areas that depend on agricultural production, Addis of today’s most successful developing countries such is the reference market where domestic agricultural as Brazil, China, and India. prices are set. Many of the opportunities for growth Concentration of economic activity increases in agriculture, therefore, depend on Addis Ababa. the productivity of all resources: land, labor, and Similarly, the scope and growth of the manufacturing capital. When resources are in close proximity, they sector depend on linkages to urban markets, and the generate external economies through specialization, size and purchasing power of Addis Ababa make it a diversity and depth of skills, concentrated demand, vital catalyst in development. innovation, technology transfers, and other mutually Addis Ababa is in the early stages of expansion reinforcing factors. Firms and workers are synergistic, and development and it has yet to reach the eco- becoming mutually more productive and prosperous nomic size and density to reap the benefits it poten- when they are located in concentrated areas, rather tially could for its inhabitants and, in fact, for Ethiopia than in isolation. at large. It is the principal destination for internal Addis Ababa and Ethiopia are early in the pro- migration as it offers the best prospects to young people. cess of agglomeration. Figure 15 shows East Africa’s Its growth, however, has led to problems of inadequate economic density (GDP per square kilometer), where public services and a financial gap between public the vertical lines represent the value of production per 26 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Figure 15: Addis Ababa’s Economic Density lacks the economic density that is a characteristic of Compared to its Immediate Neighbors more dynamic and prosperous economies. There are important synergies between Addis Ababa and the rest of the country, and as Ethiopia develops the interdependence and inter-linkages will deepen. Addis Ababa has two significant roles in Ethiopia’s economic growth. First, it introduces new sources of growth because of its scale economies and agglomeration, skilled workers and entrepreneurs, supe- rior infrastructure and financial sectors, and its gateway to external markets. It is where goods and services are Addis Ababa most readily conceived, financed, and developed—be it rose flower exports, bottled water for domestic con- sumption, or ATMs and internet banking. Second, Source: World Bank (2009). Addis Ababa, along with other urban centers, links pro- Note: Economic density is measured with GDP per square kilometer. ducers in rural areas with urban consumers and export markets. It has an inherent geographical advantage as square kilometer of land. There are tall peaks where the center of the radial structure of major roads, with large production volumes occur in small areas: Cairo other urban agglomerations at the periphery connect- and the Nile Valley, Tripoli and Benghazi. There are ing the rural areas, and is the site of Ethiopia’s only modest peaks around Khartoum, Nairobi, and Kam- international airport. For Ethiopian goods and services pala. Addis Ababa’s small peak dominates Ethiopia’s to become more competitive, Addis Ababa’s economy, economy, but trails these other more significant centers. along with other urban centers, must flourish. Figure 16 the East Africa panorama in a larger context, Addis Ababa’s role in sustaining Ethiopia’s showing economic densities in Africa, the Middle East, double-digit growth should not be underesti- South Asia, and Australia. Ethiopia, like most of Africa, mated. Addis Ababa’s ability to introduce services and Figure 16: Addis Ababa’s Economic Density Compared to Selected International Cities Addis Ababa Source: World Bank (2009). Note: Economic density is measured with GDP per square kilometer. In depth: Agglomeration and Competitiveness in Ethiopia 27 products could be strengthened through better access of the Ethiopian services and industrial output. Addis of its citizens to city-specific modern and innovative Ababa’s real GDP growth has averaged about 10 per- technologies such as reliable and low-cost transport cent since 2003 (similar to the national economy). and communication, modern banking systems, or With population growth rates at around two percent first-rate educational institutions and hospitals. Like- this overall economic growth in Addis Ababa has wise, in agriculture, reducing transportation, logistics, implied significant gains in productivity and in living and marketing costs along with improving linkages standards (AACAREP 2011). Per capita income of between rural producers and urban consumers and Addis Ababa is estimated to about US$1,200.17 investors are equally important to crop-centric efforts Yet, the city’s economic performance has been such as reducing production costs and increasing farm uneven, and growth volatile. The gap between rich efficiency. And the Government is active to address and poor in the city has widened, and the Gini coef- those challenges. For instance, Addis Ababa and its ficient increased from 0.34 in 1995/96 to 0.44 in surrounding area makes one of six growth corridors 2004/05. Addis Ababa’s inhabitants have a literacy rate in Ethiopia, based on a National Framework for of 86.2 percent, but at the same time more than half Economic Growth Corridors (MoFED 2006, and of Addis Ababa’s population lives in absolute poverty. MoFED 2009). At the same time the role of urbaniza- Still, pressures to migrate to Addis Ababa—the eco- tion in transforming agriculture is being recognized nomic center of the nation—are on the rise. This is the government’s core strategy of Agricultural Devel- driven by factors such as changing aspirations of the opment Led Industrialization (ADLI). young, population pressure on rural land, environ- mental degradation, biased road and transportation Financing the Development of Addis networks, and the general momentum of development Ababa that requires people. Despite Addis Ababa’s growth, its public services Ethiopia has one of the lowest levels of urbaniza- and utilities are lagging behind. For instance, the tion in Africa. It is predominantly agrarian with an 2007 national census found that 32 percent of residents urban population of 14.3 million, less than 20 per- have to get water from private sources of some kind, cent of the total population. Yet the urban areas have and 38 percent need to access water from public taps accounted for about 80 percent of recent economic due to lack of household connections to water systems. growth and over 58 percent of GDP is produced in The same census showed that in 2007, 14 percent of towns and cities. And urban areas attract the produc- households had no sanitary facilities. Since residents use tive and educated labor force: 90 percent of degree rivers, ditches, and open spaces instead, the lack of sani- and above graduates live in urban areas. Ethiopia’s real tary facilities contributes to environmental pollution, GDP growth accelerated from a five-year average rate contamination, and health problems. Household waste of 6.3 percent before 2004/05 to an 11 percent aver- in Addis Ababa is either collected by the municipality age during 2005/06–2009/10. This was largely driven (for 34 percent of households) or private firms (for 36 by the service sector, which accounted for 55 percent of the overall GDP growth. Agriculture grew at an average rate of 8.4 percent, contributing 33 percent 16 According to the 2007 Population and Housing Census, Addis Ababa’s to growth, and the industrial sector accounted for the population stood at 2.74 million people, while the total urban population stood at 11.96 million people. balance, 12 percent of the overall growth (EEA 2010). 17 With a nominal GDP of US$31.27 billion and a population of 77.7 Addis Ababa is home to about a quarter of Ethi- million in 2009, Ethiopia’s per capita income stands at about US$400. On average, most urbanized centers tend to have a per capita income that is opia’s urban population.16 Its city economy accounts three times larger than the national level. Using this insight, Addis Ababa’s for about 30 percent and 25 percent, respectively, per capital income can be estimated to have reached around US$1,200. 28 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness percent), but the 2007 census found that the remainder 11.5 billion in EFY 2004 (2011/12), a steep planned of households (up to 30 percent) must dispose waste increase from ETB 5 billion in EFY 2003 (2010/11). elsewhere or burn or dump it. Less than 50 percent of This, in fact, raises real concern over the actual revenue households in Addis Ababa use electricity for cooking, increase, which probably was smaller in EFY 2004. with kerosene, charcoal, firewood, and dung among the An evaluation of per capita revenues for the city more widely used alternatives.18 Overall, and according of Addis Ababa shows yet a different picture with to data from EEPCO, households consumed 45 percent per capita revenues in real terms essentially being of electricity in the city,19 industries consumed 23 per- static (Figure 17); since EFY 2000 these remained just cent, and other service sectors accounted for 39 percent under ETB 2,000 (in EFY 2003 price levels). Much of consumption. The share of industrial consumption more, Addis Ababa’s revenues from its own sources in Addis Ababa has stagnated in recent years perhaps dominate and increased from 93 percent of total rev- due to widespread power rationing. enues in EFY 1996 to 98 percent in EFY 2002. This Addis Ababa has not had sufficient resources to is in contrast to other cities in the world. Especially in invest in new projects to transform and modern- capital cities, quite often 30 to 50 percent of revenues ize the city. Four percent of Ethiopians live in Addis are from external sources. Ababa but they generate nearly 70 percent of national There are four principal sources of revenue: finances. The city does not receive direct fiscal trans- taxes on wages and salaries (personal income tax), taxes fers from the federal government, but can access value on profit to individuals (income or profit tax), urban added tax proceeds that are collected on the federal land leases, and value added tax (VAT) on services. In level from businesses and public enterprises within its fact, Addis Ababa has most of the revenue sources at jurisdiction. Addis Ababa can borrow but only from its disposal that are used by cities around the world.20 domestic sources, which significantly limits the scope But the actual revenue performance seems far below of debt financing (AACAREP 2011). As a result the potential. A recent study from Ernst & Young esti- city almost entirely relies on its own resources, which mated that Addis Ababa has a revenue potential up to haven’t grown as rapidly as the needs of the city. In Birr 20 billion (equivalent to US$1.15 billion) in the fact, Addis Ababa faces an imbalance between its medium- to long-term from tax and non-tax revenue revenues and the rising demands for public invest- streams (Ernst and Young 2010). To reach this, how- ment to maintain and improve public infrastructure ever, improved and better collection processes, inte- and services. Capital investment has been lacking for grated data management systems, and increased fiscal new infrastructure and there is limited financing to improve services. Insufficient investment may have deterred private investors and reduced the current 18 Household, Income, Consumption and Expenditure Survey, Central Statistical Agency (CSA), 2004/05. growth rate below its potential. 19 Nationwide more than 80 percent of the population has no access to While Addis Ababa’s revenues grew by 225 electricity. So electricity is just one of the sources of households access to energy and other sources dominate: kerosene, charcoal, firewood, percent in nominal terms from Birr 1.6 billion to dung, gas, and bio-gas. Birr 5.2 billion between EFY 1996 (2003/04) and 20 In the current charter of the Addis Ababa city government, which was approved in EFY 2003 (2010/11), a totality of 16 sources of city revenues EFY 2002 (2009/10), in real terms growth over the are defined. These include: income tax from employment; land use fees; period was only 33 percent, which in fact is below the tax on income from agriculture; profit, excise and turnover taxes from businesses; urban land rent; tax on income from rented houses and other rate of growth of the city’s economy. The city’s budget properties; stamp duty on contracts, agreements, and title deeds; profit, shows that the nominal increase from EFY 2002 to excise and turnover taxes from city-owned public enterprises; road user- vehicle charges; rentals from city-owned housing and other property; EFY 2003 was about seven percent, indicating a real income tax, royalty and land rentals on small scale mining; royalty on slowdown in the rate of increase of the growth rate. But the use of forest resources; fees on licenses issued and services delivered by the city; municipal taxes, duties and service charges; income tax on then, total city revenues are budgeted to reach ETB gains from renting patent rights; and capital gains tax on property. In depth: Agglomeration and Competitiveness in Ethiopia 29 Figure 17: Addis Ababa’s per Capita countries, is presented by Hobson (2010), which Revenues, in Ethiopian Birr (ETB) compared Addis Ababa with eight other African cities 2,240 and seven successful, so called “good practice� cities 1,862 1,839 1,904 1,962 1,962 around the world. The dimensions for the comparison 1,647 are dived into economic structure and productivity (two indicators) and competitiveness (six indicators). The results of this comparison, from the perspective of Addis Ababa, are shown in Table 11. As a general observation it is noteworthy that African cities sur- EFY EFY EFY EFY EFY EFY EFY 1996 1997 1998 1999 2000 2001 2002 veyed somewhat lag behind the “good practice� cities around the world. Within Africa, Addis Ababa scores Addis Ababa per capita revenues (in ETB) very high on certain indicators, such as airport passen- Source: Addis Ababa City Administration Revenue Enhancement Plan gers, but below the average in others, such as days spent (April 2011). on export procedures or the level of capital requirements to start a business relative to income. With this, Table 11 transparency would be imperative. Additionally efforts shows that Addis Ababa performs well on the dimen- to unlock land development opportunities would pos- sions of connective infrastructure, unemployment, and sibly increase the city’s overall financial capabilities. tax rates, but it performs less well on agglomeration, Strengthening of existing systems, for instance higher education, and technology. And Addis Ababa by improving collection of revenues and streamlin- markedly lags behind its African peer cities in access to ing processes and procedures to increase appliance, finance, productive utilities, telecommunications, and is needed to increase the financial power of the city generally in the city’s business regulatory environment. to serve its population and businesses. The recently A recent survey of private firms in Addis Ababa drafted Addis Ababa City Administration Revenue and its periphery has supplemented and reinforced Enhancement Plan (April 2011) is a great step towards some of the results reported in Table 11. This sur- this goal—financed through the World Bank’s IDA vey was designed to find firms’ binding constraints support. And in fact, past experience from around to growth and investment (Addis Ababa Firm Survey the world indicate that increasing collection of exist- 2010). Data was collected from 100 firms in Addis ing taxes is generally more successful than introduc- Ababa and the surrounding zones of Oromia. Targeted ing new sources. Yet, the city government may also firms included those in sectors such as manufactur- want to consider also a limited number of new and ing, services, floriculture, agribusiness, or commercial targeted taxes to increase overall revenues, such as: An farms. They all had been in business for at least one additional hotel tax or luxury tax; airport tax; or rental year; the sample together accounted for 25 percent car tax; livestock tax; or a public safety tax (Ernst and of capital investment. Generally, firms’ top constraint Young 2010). against growth was seen to be access to bank credit and the high cost associated with such credit, followed by Improving Competitiveness in Ethiopia foreign exchange controls and the unreliability/high cost of electricity. Other significant constraints were Ethiopia ranks 110th of 146 countries worldwide in mentioned as: the tax regime, customs regulations, the World Economic Forum’s Global Competitive- high cost of land, shortages of skilled labor, and mac- ness Index (2011/12). This gives it a place in the roeconomic instability. middle of sub-Saharan African nations. A different Foreign investors say the most important rea- comparison, one that is focusing on cities rather than sons for establishing firms in Ethiopia and Addis 30 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Table 11: Addis Ababa’s Competitiveness Relative to 8 other African Cities Addis Ababa Competitiveness Relative to African Comparators Dimension Indicator (#1=best) Economic structure and productivity Agglomeration Population growth rate over previous 10 years #8 of 9 Unemployment Unemployment as % of working age population #3 of 6 Drivers of Competitiveness Infrastructure Total terminal passengers in main airport #1 of 8 Percent of households with access to telephone #5 of 8 Electricity per capita consumption #4 of 4 Financial sector Number of private banks #7 of 7 Human capital Percent of population with post-secondary education #4 of 4 Technology Internet users per 100 people Low (inadequate comparison data) Institutional quality Corruption perceptions index #5 of 9 Regulatory environment for Days spent on export procedures #9 of 9 business Capital requirement to start a business as a percentage #9 of 9 of income per capita Corporate income statutory tax rate #3 of 9 Source: Hobson (2010). Ababa were family or social links to the city. countries to achieve the gradual shift from purely Other factors that investors cited included duty- agrarian-based economies towards the develop- free import of investment capital, low labor costs, ment of a industrial sector. Light manufacturing and relatively good infrastructure with access to employs low cost labor in export production and foreign markets. The emphasis on social linkages is develops skills, technology, and business acumen for somewhat worrying. It appears that many firms in the more advanced production. The World Bank recently sample invested despite the bad investment climate, analyzed the integrated value chains for six Ethiopian not because of favorable conditions. This indicates products: garments, leather products, wood products, that Addis Ababa really could attract investors from metal products, wheat flour, and processed milk a larger pool than those with social ties to the city (World Bank 2011). Table 12 summarizes the findings through a more favorable investment climate. Firms for the six product chains, and it can be used to rank were also asked to rank the policies that they think possible priorities for government action. The six most have facilitated their operations; strikingly, most firms widespread and costly problems relate to inputs (cost, in the sample did not respond to the question. Those quality, and availability), trade logistics, competition that responded said that providing incentives related policies, and firms’ abilities to produce and deliver to investment was among the most important support large volumes of timely and export-quality outputs. from the government. The next mentioned policies The study estimated product cost structures were related to taxation, logistics, and infrastructure. and looked at problems on factory floors as Light manufacturing, including agribusiness, well as in supply chains and policies. The chal- has been used as an escalator in many developing lenges subsumed under firm and industry level In depth: Agglomeration and Competitiveness in Ethiopia 31 Box 6: Business Services to Improve Competitiveness The World Bank recently conducted an enterprise survey: Survey provided in-house—have higher average labor productivity than of Users of Business Services,a which shows the importance of firms without professional or IT services linkages. The average these services for Ethiopia’s competitiveness. Greater usage of labor productivityb of Ethiopian users of professional and IT IT and professional services (accounting, legal, and engineering) services is higher than that of non-users by about 40 percent to is associated with higher labor productivity for firms in Ethiopia. 50 percent for accounting and legal services and about 75% for Data from the survey show that firms that use accounting, legal, engineering and IT services. engineering, and IT services—whether externally outsourced or Figure 18: Average Productivity of Users vs. Non-users of IT and Professional Services in Ethiopia Usage in IT 1,583,692 Non-usage IT 347,305 Usage in Engineering 811,697 Non-usage Engineering 229,734 Usage in Accounting 723,656 Non-usage Accounting 349,455 Usage in Legal 595,626 Non-usage Legal 317,477 0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000 1,600,000 1,800,000 Labor Productivity (revenue/number of employees) Source: World Bank Survey of Users of Accounting, Legal, Engineering and IT Services in Ethiopia (2011). Similar results have been reported by Li (2011). The paper analy- users of professional and IT services in Ethiopia. For the profes- ses the usage of several services, including selected profession- sional service users’ survey, the sample frames are the Addis Ababa al services, for a larger sample of 2020 manufacturing firms. Li Business Directory 2008–2010 and the Ethiopian Information and (2011) shows that total outputs and net value added per worker Shopping Directory—Third Edition 2010/11. The survey instrument of users of accounting and legal services is higher than that of was developed by the World Bank and was implemented by Precise Consult International PLC. The survey covers the following: IT/com- non-users and finds that the estimated premiums for users’ labor puter, accounting, legal, and engineering business services. productivity is 33 percent for accounting and legal services. b Labor productivity is computed as the revenue of a firm divided by the number of its employees. a In 2010–2011 the World Bank conducted firm surveys covering 72 in Table 12 essentially are about productivity level constraints in Ethiopia; these are highlighted in and cost. Firms can usually control productivity if Table 13. the policy environment is conducive and enabling enough to adopt good practices. Causes for costs Lessons from East Asian Experiences (other than those due to management decisions), are typically beyond the control of firms, but they can The light manufacturing study also provides useful be affected by public policies related to taxes, tariffs, insights and comparisons to experiences in China, labor, and capital goods. The study delivers a series Vietnam and other East Asian economies, many of noteworthy insights both on productivity and cost of which, in the 1980’s, were similar to Ethiopia 32 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Table 12: Challenges for Manufacturers in Six Selected Ethiopian Industries, 2010 Issue Not a problem Problem Important Problem Firm and industry level # of industries Input quality, cost, and volume — — 6 Input efficiency — 3 3 Quality, volume, delivery time — 4 2 Labor efficiency — 5 1 Indirect costs — 6 — Capital technology & efficiency 1 4 1 Capital cost 1 5 — Labor cost 6 — — Policy level # of industries Input industry policies — — 6 Competition policies — 3 3 Trade logistics — 3 3 Land and infrastructure policies — 6 — Labor and technology policies — 6 — Financing policies — 6 — Source: World Bank (2011): 20. Note: Summary statistics of the findings of a World Bank study of six light manufacturing industries: garment, leather products, wood products, metal products, wheat flour, and processed milk. Table 13: Productivity and Cost-related Constraints for Firms and Industries in Ethiopia, 2010 Productivity related issues Cost related issues Labor efficiency. On average, labor efficiency in medium and large Labor costs. The wages for low-skilled Ethiopian workers are Ethiopian firms is about 50 percent of Chinese and Vietnamese levels. a tenth of those in China and less than half those in Vietnam. This is largely due to training, smaller-scale operations, poor organiza- The wage gaps for skilled labor are much smaller, but this is tion, and productivity bonuses. only a small share of the workforce in light manufacturing. Capital efficiency. Capacity utilization in Ethiopian firms averages 60 Capital costs. Industrial machines are imported but they percent, compared with 90 percent in China and about 80 percent cost about 50 percent more in Africa than in Asia because in Vietnam. The differences stem from less continuous order flow, old import duties and taxes are high. equipment, and power shortages. Input costs. Ethiopian firms enjoy a significant cost advan- Input efficiency. Average waste and product rejection rates are often tage in leather inputs, but they have difficulties in finding up to 15 percent in Ethiopia, compared with around 5 percent in China large volumes at consistent quality. Textiles are cheaper and Vietnam. This is mostly due mostly to poor training. but they cannot be exported because of poor quality. Product quality. Ethiopian products are sold at low prices on world Wood costs twice as much in Ethiopia as in China, steel markets. For polo shirts the price is 40 percent lower than for Chinese more than 25 percent more (most must be imported), and shirts; for high-end men’s leather shoes it is 25 percent lower. The lower domestically produced wheat 74 percent more because prices are due to lower product quality, longer and uncertain delivery of lower yields. Milk costs are high because of lower- times, inability to fulfill large and diverse orders, and poor reputation. yielding breeds. Source: World Bank (2011). In depth: Agglomeration and Competitiveness in Ethiopia 33 today. Through these comparisons, the study finds economy benefitted from promoting competition that “many of the root causes of the productivity and linkages between the two groups. Large firms and cost issues (…) can be traced to policy problems could reduce their costs and increase their flexibility relating to competition and input industries. These by building networks of small subcontractors, while problems cannot be solved by the private sector alone. small firms could grow stronger from working and Nor can they be solved by the public sector alone� competing with larger ones. And fourth, there is a (World Bank 2011). range of policy issues for the light manufacturing Looking at East Asian countries, there are a industries, but not all of them are equally impor- series of lessons that may help to unveil the poten- tant. The government should set priorities based on tial of light manufacturing in Ethiopia. First, for- careful analysis to achieve the best results. Global eign direct investment is important to introduce competitors have increased scale and scope such that good managerial practices and of competition to dif- Ethiopia can learn from their successes to improve its fuse those practices, particularly among small firms. policies. Finally, it appears that the light manufactur- Special Economic Zones can play a role in this, and ing industries with the greatest new-term potential are there is an opportunity to adopt proven insights from in leather, garment, and metal products. other countries, such as Korea (see Box 7 for Korea’s experience in the Masan Free Export Zone). It also Specific Recommendations to Increase found that good trade logistics is a must to be glob- Ethiopia’s Competitiveness in Light ally competitive, both in cost and quality. By con- Manufacturing trast, policies facilitating bank financing and those facilitating skills training in the formal sector were Ethiopian competition policies create an uneven less important. Second, the cost and quality of inputs playing field and discriminate between a few large represent significant challenges across the six product exporters and the many small informal firms pro- chains. Problems can be traced to policies related to ducing low-quality products for the domestic mar- entry barriers to competitive producers and to tariffs ket. Exporters enjoy government support through on imported inputs. Potentially competitive industries duty-free imports of inputs, tax incentives, access to such as agricultural products, wood, livestock, and export-oriented industrial zones, land, and finance. leather would benefit significantly from lower-cost However they pay for this with limited access to and higher quality inputs. the domestic market. Large firms, particularly state Third, there is a deep divide between the few enterprises, may also benefit from government pro- formal exporting firms and the larger number curement practices and subsidies. By contrast, small of small, informal industrial firms. The Chinese non-exporting firms are subject to numerous taxes Raising the level of competition • Improve the investment climate and reduce regulatory burdens (including tax policy and administration). • Exit from state-owned enterprises and abolish the investment licensing list (see Box 8 on privatization strategies in China and Korea). • Liberalize and remove or reduce tariffs, taxes, and duties on imported inputs and machinery for all firms whether for the domestic market or for export. • Harmonize of domestic taxes for both exporters and non-exporters, and reduce delays in reimbursement of value added tax. • Accede to the World Trade Organization. Source: World Bank (2011). 34 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Box 7: Korea’s Experience in Special Economic Zones: Masan Free Export Zone Special Economic Zones (SEZs) in Korea were introduced in show its success (Figure 19), especially during the early years the early 1970s and contributed to promoting exports and of the 1970s and 1980s: technological spillovers through attracting foreign direct investment. During that time Korea was suffering from serious • Foreign investment and employment. Foreign capital external imbalances and high employment, leading to internal invested in the zone started with meager US$0.5 million reserves drain and increasing external debt problems. Masan in 1971 but jumped to US$100 million in 1980, with the Free Export Zone (FEZ) was established in 1971 on the annual average new foreign investment amounting to Southern coastal area of the Korean peninsula as the first FEZ, US$11 million during the period. The majority of the foreign with the objective of promoting export and employment and investors were from Japan, followed by American firms. The inducing technological knowledge transfer among local firms number of workers in zone climbed to 31,000 in 1979 and through attracting FDIs in the manufacturing and chemical reached its peak of 36,000 in 1987. Thus, the zone achieved industries (see 2008). its objective of job creation in the local area. Foreign firms operating in the zone were given wide range of benefit including special exemptions from various regulations Figure 19: Development Trends in Masan on foreign investment that prevailed during the time. Incentives FEZ (Korea) given to foreign firms in Masan FEZ include the following: Trends of FDI and Employment in Masan FEZ • Tax incentives. Tax breaks and holidays were most typical (1971–1980) types of incentives offered to foreign investors in the zone. 120 35,000 The Korean government provided various tax incentives: (i) Exemption from custom duties was given to imported 100 30,000 capital goods; (ii) Complete exemption from national and 25,000 local taxes was provided during the first five years, including 80 20,000 corporate income tax, individual business income tax and 60 property tax, and then 50 percent tax reduction was provided 15,000 for the next three years. 40 10,000 • Reduced user fees and basic service cost. Rental 20 fees for factory sites or standard factory building were 5,000 completely exempted (for firms with high level technology) 0 0 or significantly reduced. User fees for basic services (i.e., 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 electricity, water) were also greatly discounted and basic facilities (i.e. fire station, postal service, bank) were provided Net FDI (millions of US$) – Left Axis within the zone. Number of workers – Right Axis • One-stop service system. An administrative office centrally handled various administrative processes (permits Export from Masan FEZ and incentives), significantly reducing firms’ administrative (millions of US$, 1971–1989) cost from dealing with different central and local government 2,000 departments. 1,800 • Restriction on labor dispute. Endowing foreign companies 1,600 in the zone with the legal status as public enterprises, the 1,400 government in effect banned labor dispute in the zone. 1,200 1,000 Masan FEZ became one of the most successful free economic 800 zones in Korea in the 1970s and 1980s (in fact until the late 600 1980s when broad-based growth of the Korean economy 400 gained momentum and the relative economic contribution of 200 the zone started to decline). Business activities in Masan FEZ grew rapidly with the number of foreign companies operating 0 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 in the zone reaching its fully capacity of 115 firms in 1973 only three years after its inception in 1971. The number of companies remains in the range of 70–100 ever since 1974 and currently 100 firms are operating in the zone. Key indicators of the zone Source: Administration Agency of Masan Free Trade Zone. (continued on next page) In depth: Agglomeration and Competitiveness in Ethiopia 35 Box 7: Korea’s Experience in Special Economic Zones: Masan Free Export Zone (continued) • Contribution to export. Foreign exchange earnings of foreign investors to limit the access of local workers to through exports is one of the primary gain expected from their technological assets. Then transfers of managerial SEZs in many developing and low income countries. Masan know-how and techniques became more significant in FEZ was not an exception to this. In fact, exports from the Masan FEZs through training of local employees provided zone increased greatly from mere US$70 million in 1973 to by foreign companies operating in the zone (Warr 1984). over US$628 million in 1980. Exports continued the steady Finally, in the 1980s, stronger backward linkages between growth trend thereafter and until 1989. The export share of foreign invested firms in the zone and local firms in the Masan FEZ among Korea’s total exports reached its peak of region (suppliers) provided an appropriate channel to transfer 4.1 percent of total exports in 1979. technologies; much more, the zone eventually became a driving force for job creation outside its boundaries. By Technological spillovers were made through strong 1988, 525 local firms with 16,686 workers were involved in backward linkages. One of the important benefits expected the production of basic materials/goods needed (inputs) for from foreign investment is to realize positive spillovers of the final manufacturing process of foreign firms operating in advanced technology and knowledge into the broader the zone (Cho 1990). Additionally, many Korean companies economy. But technological spillovers in Masan FEZ were accumulated technological capabilities through complying with not significant during the 1970s due to the labor-intensive the technological standards and demands from their business nature of most foreign invested production and the tendency counterparts in the zone. and high tariffs on machines and inputs. This forces cost is for moving it the 700 kilometers between the many firms into the informal sector, making it even port of Djibouti and Addis Ababa. Moving the same more difficult for them to compete in formal markets container from Guangzhou to Los Angeles costs only as sellers or as purchasers of skills, land, and utility $2,000. Poor trade logistics increase production costs, or financial services. delivery times, and uncertainty, which forces Ethio- Ethiopia’s trade logistics are weak. For example, pian manufacturers to accept lower prices for their moving a 20-foot, 10-ton container between Addis exports. Other countries have built their export sectors Ababa and Guangzhou costs $4,000, and half that on the back of good trade logistics. Box 8: Privatization Strategies in China and Korea … in China (1984–2010): Particular bold reforms under the 9th Five Year Plan (5YP) 1995–2000 led to a greatly expanded role of the private and Reform of China’s state-owned enterprises (SOEs) has been other non-state sectors (World Bank 2012a). With this, reform a major aim since urban reforms began in 1984. Although has accelerated and acquired some qualitatively new features. there were calls to privatize the SOEs, the Government’s First, the scale of change has expanded to affect almost every initial emphasis was on boosting performance by changing kind of SOE—small, medium, large, and very big; under both the internal governance of SOEs and improving the market central and local control. Second, ownership diversification has environment in which they operated. By the late 1980s the been so extensive that the wholly state-owned non-financial Government had decided that the best way to reform small company has become an endangered species in China. Third, SOEs was to lease them out, with the manager paying the the range of restructuring mechanisms being used has expanded state a fixed proportion of the firm’s profit. Incorporation was dramatically to include bankruptcies, liquidations, listings another significant measure that led to privatization (Kikeri and and de-listings, debt-for-equity swaps, sales to private parties Kolo 2005, referring to Garnaut et al. 2005). (domestic and foreign), auctioning of state firms and their assets (continued on next page) 36 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Box 8: Privatization Strategies in China and Korea (continued) or liabilities, standard corporate governance techniques, and in privatization occurred only after the 1997/98 Asian crisis when so on. Finally, mass layoffs—unheard off in the earlier years of the economy was in immediate danger, and supporting voices the reforms—have become a widespread phenomenon (Kikeri for public sector reform were particularly strong. Looking at more and Kolo 2005, referring to Garnaut et al. 2005). than three decades of development, privatization efforts in Korea In sectors of strategic importance, such as infrastructure can be summarized in five phases (KDI 2010; and MSF 2008): and energy, where the regulatory framework keeps evolving, monopolies have been broken and competition has been • 1st Phase (1968–1973): The number of SOEs in Korea had introduced. Many companies have been corporatized, and rapidly increased from 35 at the end of 1950 to 120 in some have been listed on local and international exchanges. 1970. Back then, SOEs existed and played a dominant Through this China has nurtured over 20 giant corporations role in most key sectors of the economy, including: and conglomerates that have proven competitive in the transportation, energy, mining, manufacturing, chemical international market. Some of these companies are laying production, construction, and finance. As the strong off tens—or even hundreds—of thousands of employees, not market intervention of the government through SOEs because they are in financial distress (some of them are hugely had increasingly been criticized for creating inefficiency profitable) but because they wish to position themselves as in the market, the first earnest privatization plan in Korea important international players (Kikeri and Kolo 2005, referring was implemented in 1968, largely through the sale of to Garnaut et al. 2005). government shares to strategic buyers. Eleven SOEs in Many small and medium-sized SOEs became privately aviation, surface transportation, mining and manufacturing owned. In line with these developments, the new policy were privatized between 1968–1973 including Korean Air direction has been to diversify the ownership of state Lines and one large commercial bank (out of five). enterprises. Indeed, many large state enterprises have been • 2nd Phase (1980–1983): Privatization in the 1980s put large “corporatized� and some of the biggest (including those emphasis on privatizing four remaining commercial banks directly monitored by the central government) are now not owned by the government through general competitive only listed on stock exchanges but have also improved their bidding process of the government shares. However, the governance structure, managerial professionalism, and original purpose of banking sector privatization was only profitability. But while the profitability of state enterprises has partially fulfilled as government intervention continued improved, it has still remained well below non-state firms, in the operation of the privatized banks through strong including pure private firms (World Bank 2012a). As a result, the state sector’s share in the total number of industrial enterprises (with annual sales over 5mn RMB) fell from 39.2 percent in 1998 to 4.5 percent in 2010. During this Chaebols are large corporate groups in Korea and at same period, SOEs’ share in total industrial assets fell from the heart of the unique development path of the country. 68.8 percent to 42.4 percent, while their share in employment Many of them were family controlled, but all of them was slashed from 60.5 percent to 19.4 percent. Their share are fully private companies. Famous current examples in China’s exports fell from 57 percent in 1997 to 15 percent include global enterprises such as Samsung, Hyundai in 2010. Thus, the non-state sector has become not only the and LG. Since the 1960s, Chaebols increased their main generator of output (an estimated 70 percent of GDP) and influence on the Korean economy with the support of the employment, and strongest growth engine, but also the most government at times of credit rationing and industrial active sector for innovation (World Bank 2012a). policy. Economic dominance of Chaebols in Korea was largely regarded as one of the underlying causes … in Korea (1968–2002): of the 1997 economic crisis. In the aftermath of the crisis, the Korean government announced its intention The privatization of SOEs in Korea went through several phases to break the myth of “too-big-to-fail� and implemented and the objective of privatization also evolved as the economic strong restructuring processes in the area of corporate and social environment changed along with development. governance. As a result, many Chaebols went bankrupt Korea’s experience showcases the difficulty of privatization or were dismantled into smaller corporate groups. One especially amidst strong political resistance; in fact, ambitious of the more famous examples is Daewoo, which grew privatization plans announced by the government achieved only rapidly during the 1980s but eventually went out of partial successes or had to be delayed and/or modified in many business in the early 2000s. cases because of political opposition. The most notable progress (continued on next page) In depth: Agglomeration and Competitiveness in Ethiopia 37 Box 8: Privatization Strategies in China and Korea (continued) political influence on management appointment, and asset Due to rising concerns on the growing economic power management until the 1990s. The oil industry was another concentration to Chaebols, the plan excluded the energy key sector of the privatization plan, which led to privatizing and telecommunication sector from the privatization the Korea Petroleum Corporation (1980). target. Concerns over the economic dominance of a few • 3rd Phase (1987–1989): A grand privatization plan was large conglomerates again led to strong political resistance established in 1987 to target SOEs with significant size and and the implementation of the privatization plan fell far economic influence. The plan included privatization through short of the original design: only 22 SOEs were fully or stock market issuances of Korea Telecom Corporation (KT), partially privatized by 1997. But the plan eventually led Korea Electric Power Corporation (KEPCO), Pohang Steel to the adoption of “Public Enterprise Privatization Act� in and Iron Corporation (POSCO) and Korea Stock Exchange. 1997. However, the ambitious original plan was again only partially • 5th Phase (1998–2002): Privatization was more extensive completed, due to two primary concerns: (i) the supply of and effective than in any past experience, strongly driven stocks from the large SOEs would put too large burden for by the urgent needs to address the then omnipresent the still weak stock market to bear; and (ii) there was strong economic crisis of 1997/98. A Government Reform Bureau opposition from political circles and labor unions due to was established for that purpose. It envisioned immediate potential economic power concentration to Chaebols which full privatization of five large SOEs in steel and heavy were considered most likely buyers of the large SOEs. By manufacturing industries and a phased-in privatization of 1989 the Korea Stock Exchange was fully privatized as its six large SOEs in telecommunication, energy and tobacco government shares were sold to financial companies and industries. Plans were announced for fifteen large SOEs to the government shares of KEPCO and POSCO were partially go through a so-called self-restructuring process. However, sold to the general public. re-emergence of public concern about some large SOEs • 4 th Phase (1993–1997): The new administration in that were regarded as public goods providers halted 1993 announced another full-scale privatization plan privatization in electricity, gas industries, and railroad envisioning the sales of government shares in 55 SOEs. services. Local inputs are of low quality so Ethiopia is not using its natural resources as a competitive advantage. Improving trade logistics for apparel exports This was the case in each of the six value chains that • Ethiopia has the potential to compete globally in were studied. Government control and protection has apparel thanks to a significant and growing labor cost undermined the competitiveness of textiles. Diseases advantage, access to a state-of-the-art and well-located container port in Djibouti, and duty-free access to the and access to land have led to low quality and low US and EU markets. volumes of leather skins. Poor management of wood • To address the binding constraint of trade logistics, plantations has raised the price of wood and threat- the government could consider establishing a green channel for apparel at customs, providing free and ened natural forests. Import tariffs on steel penalize immediate access to foreign exchange, reducing the domestic buyers. Low quality seeds, fertilizers, and cost of letters of credit, and setting up an industrial irrigation lead to a rise in the cost of domestically pro- zone closer to Djibouti. duced wheat. And low-yield breeds and entry barriers Source: World Bank (2012b). for large-scale animal farms increase the cost of milk. Lack of access to rural land has hindered poten- tial investments to produce inputs such as leather, Finance and investment are required when a firm wood, wheat, and milk. Similarly competitive, reli- starts up and when it expands. For light manufactur- able utility services are not widely available, with the ing firms in Ethiopia the main source of financing has power and telecommunications sectors still to be fully been retained earnings. The second biggest source has liberalized and opened to foreign investors. been supplier or customer credit. Industry value chains 38 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Overcoming input shortages for leather products Addressing land and finance issues for wood and metal products • Ethiopia has even greater potential in leather, which is more labor-intensive than apparel. But the immediate • Ethiopia has the natural resources to develop a constraint is limited access to high-quality processed competitive wood supply, which currently is unreliable, leather. unsustainable, and poorly organized. • Enabling imports of high-quality processed leather • The price of steel in Ethiopia is 30 percent above the would be an easy and immediate solution to the acute price in China due to poor trade logistics and high shortage of leather in the industry. import tariffs. Even in the better-managed firms, labor • Removing restrictions on the exports of leather would productivity is low. increase the incentives to invest in the Ethiopian leather • The potential lies not in exports (at least initially) but in supply chain. the growing domestic market and in the high weight-to- • Facilitating access to rural land for good practice animal value ratio of finished wood and metal imports. farms, promoting better breeds, controlling cattle • For wood, the government should facilitate access to disease, and enabling the use of cattle as collateral rural land and financing for private wood plantations. would further stimulate the supply of quality leather. • For metals, eliminating the 10 percent import tariff on steel should reduce the cost of inputs, and enabling Source: World Bank (2012b). the exploitation of the vast proven reserve of iron ore combined with hydro-energy, could make the domestic steel industry competitive. and industrial clusters generate the trust and peer pres- Source: World Bank (2012b). sure for this to work effectively. Financing from family and friends as well as the diaspora has been important, especially at start-up. Bank financing plays virtually liberalization and labor mobility. Ethiopia has made no role in the start-up phase in any country. It would commitments to pursue regional integration in the be extremely risky for banks to finance start-ups with context of the COMESA and the Tripartite21 pro- little collateral and no track record. It has had a bigger cess. Ethiopia’s participation in the WTO accession role in expansions. The Ethiopian government has a process, the COMESA negotiations on professional policy of facilitating access to finance only for export- services and the recently established Professional Ser- ers in preferred industries. vices Knowledge Platform can help the country with Additional, Ethiopia has considerable potential the development of a meaningful reform program to export services but sustaining and broadening of that includes the elimination of explicit barriers and this kind of export growth will require a more sup- regulatory, education and immigration reforms. portive policy framework. At the same time, services are crucial for Ethiopia’s overall competitiveness, and Summary: Unfolding the Potential of a opening up services can help diversify into exporting Thriving Economy agricultural and manufactured goods with higher value added as previously recommended. A clear services Growth requires agglomeration and the scale strategy that focuses on better coordination across economies it offers: specialization, diversity and ministries and deeper dialogue with stakeholders may depth of skills, concentrated demand, innovation, provide the required stimulus of such exports. and technology transfers. This section shows that A recent study provides concrete recommen- Addis Ababa is the only urban area in Ethiopia that dations for the reform of business services (World Bank 2012c). The report shows that a successful 21 The Tripartite agreement between COMESA, SADC and EAC aims reform of business services requires policy action at forming a common FTA among the three regional groupings. The at the national and international levels in the fol- decision by the three RECs to form the FTA has been described as a big political commitment and a landmark decision in Africa’s regional lowing areas: education, domestic regulation, trade integration efforts. In depth: Agglomeration and Competitiveness in Ethiopia 39 is capable of delivering these scale economies, but its of foreign investment into the country. Ethiopia ranks small economic density is insufficient to realize the 110th of 146 countries in the World Economic Forum’s economies that are necessary for a strong impact on Global Competitiveness Index and in the middle of Ethiopia’s growth. Addis Ababa’s economic profile is sub-Saharan African nations. A developing nation’s insignificant compared to productive and prosperous goal is to attract investors who are drawn by a favor- cities in East Africa, the Middle East, South Asia, and able investment climate, and there are lessons from Australia (Figure 17). For Ethiopia to develop, Addis East Asia on how to better achieve this. Ababa will have to increase its economic density as Light manufacturing, including agribusi- other, more dynamic cities have. ness, has been used to raise developing coun- Greater investment, public and private, will tries from agrarian-based economies. It employs lead to greater economic density as land and other low-cost labor in export production and develops factors are used more intensively. Addis Ababa’s skills, technology, and business methods for more public services and infrastructure—water, sanitation, advanced production. Ethiopia has had some success sewerage, electricity and communications—are inad- with light manufacturing but analysis of six product equate. They do not compare well to other African chains revealed several problems that are reducing cities, and they substantially trail comparator cities productivity and competitiveness: inputs (cost, outside Africa. Addis Ababa has a gap between the quality, and availability), trade logistics, competition revenues it is collecting and the investments that policies, and firms’ abilities to produce and deliver would be required to improve public services. There large volumes of timely and export-quality outputs. are several options for raising revenues from existing There are several policy options to alleviate these taxes to adequate levels by modernizing and improving problems. These policies would not only improve the city’s revenue collection systems, policies, proce- the competitiveness of these firms and product dures, and staffing. lines, but they would also improve the investment Attracting new private investment will require climate, attract new investors, and build develop- a better investment climate. Addis Ababa does not ment momentum to raise the economic profile of stand out as a city that can facilitate large-scale inflow Addis Ababa and Ethiopia. 4 References Section I The Africa Report. 2012. “The rise and rise of Ethio- pia: From starvation to industrialization in two Economist Intelligence Unit (EIU). 2012. Country decades, Ethiopia is clearly booming. But can it Report Ethiopia, March. avoid the curse of the oligarchs.� May. EthioInvest. 2012. Available at: http://www.ethioin- Xiao, Geng, and Haiying Zhao. 1998. “Characteristics vest.org/Investment_Regime.php in Chinese FDI and Trade Policy: Lessons Learnt Flassbeck, H., S. Dullien, and M. Geiger. 2005. and Policy Implications.� in International Trade China’s Spectacular Growth Since the Mid-1990s, Issues. July. Available at: http://www.econ.hku. Ma-croeconomic Conditions and Economic Policy hk/~xiaogeng/research/Paper/Chinese/China-trade.pdf Challenges. In: China in a Globalizing World, United Nations Conference on Trade and Devel- Section II opment. New York and Geneva. Geiger, M. 2010. Monetary Policy in China: Institu- AfDB, OECD, UNDP, UNECA. 2011. African tions, Targets, Instruments and Strategies. Doctoral Economic Outlook 2011, Ethiopia Country Note. Dissertation. University of Würzburg, Würzburg. Available at: http://www.africaneconomicoutlook. HICES: Household Income Consumption and org/en/countries/east-africa/ethiopia/ Expenditure Survey. 1995/96, 1999/00, 2004/05 Alem, Y. and M. Söderbom. 2010. “Household-Level and 2010/11. Central Statistics Agency (CSA), Consumption in Urban Ethiopia: The Impact of Addis Ababa, Ethiopia. Food Price Inflation and Idiosyncratic Shocks.� Huang,Yasheng. 1998. FDI in China: An Asian Per- CSAE Working Paper Series 2010–24, Oxford: spective. The Chinese University, Singapore. Centre for the Study of African Economies. Lopez, J. Humberto. 2004. Pro-growth, Pro-poor: Berhane, G., Z. Paulos, K. Tafere, and S. Tamiru. Is there a Tradeoff? The World Bank (PRMPR). 2011. “Foodgrain Consumption and Calorie Available at: http://www-wds.worldbank.org/ Intake Patterns in Ethiopia.� Ethiopia Strategy external/default/WDSContentServer/WDSP/IB/2 Support Program II (ESSP II) Working Paper No. 004/09/22/000160016_20040922152917/addi- 23. Washington, D.C.: International Food Policy tional/130530322_20041117143606.pdf Research Institute. Ministry of Commerce of the People’s Republic of Bigsten, A. and A. Shimeles. 2008. “Poverty Transition China. 2012. China Foreign Direct Investment and Persistence in Ethiopia: 1994–2004.� World (FDI). Available at: http://www.fdi.gov.cn/ Development, 36(9): 1559–1584. New African. 2011. Interview with Prime Minister Deaton, A. 1997. The Analysis of Household Surveys: Zenawi. Available at: http://www.newafricanmaga- A Microeconometric Approach to Development. zine.com/special-reports/country-reports/ethiopia- Washington DC: World Bank and John Hopkins focus-on-investment/zenawi-were-on-the-right-track University Press. 42 Ethiopia Economic Update — Overcoming Inflation, Raising Competitiveness Dercon, S. 2006. “Economic Reform, Growth and Rashid, S. 2010. “Staple Food Markets in Ethiopia.� Paper the Poor: Evidence from Rural Ethiopia.� Journal prepared for the for the COMESA policy workshop of Development Studies. 81 (1): 1–24. on food price variability, January 25–26, Maputo. DPRD and MOFED. 2008. Dynamics of Growth and Available at: http://www.aec.msu.edu/fs2/aamp/semi- Poverty in Ethiopia (1995/96–2004/05), Addis nar_3/AAMP_Maputo_27_Ethiopia_ppr.pdf Ababa: Development Planning and Research Rashid, S. 2011. “Inter-commodity Price Transmis- Department and Ministry of Finance and Eco- sion and Food Price Policies: An Analysis of nomic Development. Ethiopian Cereals Markets.� IFPRI Discussion Durevall, D., J. L. Loening, and Y.A. Birru. 2010. paper 1079. Washington, D.C.: International “Inflation Dynamics and Food Prices in Ethiopia�. Food Policy Research Institute. Working Papers in Economics 478, Göteborg Tafere, K., P. Zelekawork, A. S. Taffesse, T. Nigussie, University, Department of Economics. and T. Seneshaw. 2009. “Food Demand Elastici- FAO-GIEWS. 2011. Global food price monitor. Food ties in Ethiopia: Estimates using 2004/05 House- and Agriculture Organization, under the Global hold Income, Consumption Expenditure (HICE) Information and Early Warning System on Food Survey Data.� Ethiopia Strategy Support Program and Agriculture. 9 December: Rome II Discussion Paper 11. Addis Ababa: Interna- FAOSTAT. 2011. http://faostat.fao.org/. Statistics tional Food Policy Research Institute. downloaded in December 2011. Ticci, E. 2011. “Can inflation be a good thing for the ILO. 2004. “Consumer Price Index Manual: Theory poor? Evidence from Ethiopia.� Development in and Practice.� Available at http://www.ilo.org/pub- Practice. 21(4–5): 629–641. lic/english/bureau/stat/guides/cpi/index.htm. The Ulimwengu, J. M., S. Workneh, and Z. Paulos. chapter Concepts and Scope, number 3, includes 2009. “Impact of Soaring Food Price in Ethiopia. explicit recommendations on “subsistence agri- Does Location Matter?� IFPRI Discussion Paper culture� in the calculation of the CPI in page 47. 00846, Washington, D.C.: International Food International Monetary Fund. 2011. “The Federal Policy Research Institute. Democratic Republic of Ethiopia: Joint Staff World Bank. 2005. Ethiopia Well-Being and Poverty Advisory Note on the Growth and Transformation in Ethiopia. The Role of Agriculture and Agency, Plan 2010/11–2014/15.� IMF Country Report Washington D.C.: World Bank. No. 11/303, Washington D.C.: International World Bank. 2010. Re-Igniting Poverty Reduction in Monetary Fund. Urban Ethiopia through Inclusive Growth, Report Levinsohn, J. and M. MacMillan. 2007. “Does Food No. 52848-ET. Washington D.C.: World Bank. Aid Harm the Poor? Household Evidence from World Food Program. 2008. “Summary of Price Ethiopia.� In A. Harrison (ed.), Globalization and Impact Security Assessment Findings.� Food Poverty. University of Chicago Press. Security Analysis Group. Available at: http:// Loening, J. L. and G. Oseni. 2007. “Approximating home.wfp.org/stellent/groups/public/documents/ena/ Rural and Urban Welfare Effects of Food Price wfp186310.pdf Inflation in Ethiopia.� Unpublished paper, Wash- ington D.C.: World Bank. Section III Loening, J. L., D. Durevall, and Y.A. Birru. 2009. “Inflation Dynamics and Food Prices in an Agri- AACAREP. 2001. Addis Ababa City Administration cultural Economy: The Case of Ethiopia.� Policy Revenue Enhancement Plan, Draft Final Report, Research Working Paper Series 4969, Washington April 2011, p. 3. (project background report – D.C.: World Bank. unpublished). References 43 Ethiopian Economics Association, Ethiopian Eco- why?—Evidence from the manufacturing indus- nomic Policy Research Institute. 2010. Unleashing tries in Ethiopia.� World Bank, mimeo. Addis Growth Potential: Recent Economic and Social Ministry of Finance and Economic Development. Trends in Addis Ababa, p. 21. (project background 2006. A Plan for Accelerated and Sustained Devel- report – unpublished). opment to End Poverty. Ernst & Young. 2010. Project: Study of Revenue Gap Ministry of Finance and Economic Development. and Potential of the City of Addis Ababa. Ernst & 2009. Terms of Reference for Preparing a National Young, Addis Ababa (project background report Framework for Development and Operation of Eco- – unpublished). nomic Growth Corridors in Ethiopia. First Consult. 2010. Addis Ababa Firm Survey: Identi- Ministry of Strategy and Finance. 2008. Progress in fying the Binding Constraints to Firm Growth and Public Sector Reform Plan: Republic of Korea Investment, November, 2010. (project background Warr, Peter G. 1984. “Korea’s Masan Free Export report – unpublished). Zone: Benefits and Costs.� The Developing Eco- Garnaut, Ross, Ligang Song, Stoyan Tenev, and Yang nomics. Yao. 2005. China’s Ownership Transformation: Pro- World Bank. 2009. World Development Report 2009: cess, Outcomes, Prospects. IFC: Washington D.C. Reshaping Economic Geography. The World Bank: Hobson, Emma Wadie. 2010. Addis Ababa: Towards a Washington, DC. Globally Competitive City, August, 2010. (project World Bank. 2011. Development Economics, Opera- background report – unpublished). tions and Strategy; Africa Finance & Private KDI 2010. The Korean Economy – Six Decades of Sector Development, Light Manufacturing in Growth and Development. Africa: Creating Jobs and Prosperity by Fostering Kikeri, Sunita, and Aishetu Fatima Kolo. 2005. Competition, Volume I, The World Bank Group, “Privatization: Trends And Recent Developments� January 31, 2011 (draft unpublished report for World Bank Policy Research Working Paper discussion). 3765, November 2005. Available at: http://rru. World Bank. 2012a. China 2030: Building a Mod- worldbank.org/Documents/Privatization/Privatiza- ern, Harmonious, and Creative High-Income tion_paper.pdf Society. Available at: http://www.worldbank.org/en/ Kwang-kuk, Cho. 1990. “The Outprocessing Linkages news/2012/02/27/china-2030-executive-summary. in Korean EPZs�, in R. L. Bolin (ed.) Linking the World Bank. 2012b. Development Economics, Export Processing Zone to Local industry. The Operations and Strategy; Africa Finance & Pri- Flagstaff Institute. vate Sector Development, Light Manufacturing in Lee. 2008. “Feasibility Study on Establishing Special Africa: Creating Jobs and Prosperity by Fostering Economic Zones in Uzbekistan.� Policy Research Competition. Volume I. The World Bank Group. Issue Paper, Ministry of Finance and Economy, World Bank. 2012c. Developing Business Services May. in Ethiopia: How trade and International Inte- Li, Yue. 2011. “Are the users of services inputs gration Can Help. World Bank, Africa Region in Sub-Sahara African countries better and (forthcoming). the World Bank 1818 H Street, Nw Washington, DC 20433