35725 AID AND REFORM IN AFRICA PROJECT OF THE WORLD BANK Aid and Reform in Ethiopia Berhanu Abegaz Department of Economics The College of William and Mary Williamsburg, VA 23187 (USA) ABSTRACT Ethiopia managed to attract over US$17 billion in official development assistance in 1980-97. This average annual inflow of 16 percent of GDP is equally divided between multilateral donors (IDA, AfDB, U.N. and EU) and bilateral donors (mainly Scandinavia, Italy, Germany and U.S.). The country's onerous external debt stood at US$10 billion in 1997, almost equally divided between Russia (defense loans) and DAC lenders. Ethiopia's post-socialist track record shows that the country has made a decisive break with its command economy in many, but not all, respects. Price controls have been largely dismantled, tax and tariff rates reduced, and the inflation rate kept low. The EPRDF Government has a strong sense of ownership of the reforms, and it is committed to several measures that go beyond stabilization and liberalization. In response, well-coordinated donor assistance now focuses on sector-wide investment programs in roads, health and education. Important initiatives in energy and food security are also in the pipeline. Foreign aid has recently played a critical role in implementing first-generation reforms, and in nudging the Government to undertake deeper reforms regarding parastatals, finance and trade than it was initially willing to contemplate. The remarkable spurt of economic growth after 1992 is, however, under a constant threat of derailment as the recent return of war and famine underscores. The reasons are hardly elusive. First, the Government has yet to institute an inclusive governance system that empowers all fundamental interest groups. Second, it is reluctant to undertake second-generation institutional and policy reforms that would constitute an irreversible break with a discredited past. These include reprivatization of land and housing, more vigorous banking and trade reforms, and an even-handed competition policy. Third, an effective mechanism for mobilizing private savings has yet to be devised in order to raise the low investment rate and to reduce aid dependency. A credible commitment to a private-sector-led development strategy and the rule of law is essential for ensuring durable social peace, effectively absorbing additional resources from an increasingly selective foreign aid regime, and utilizing such aid as leverage for crowding-in foreign private investment. August 1999 Table of Contents I. Introduction 3 1.1. Aid: Form, Effectiveness and Selectivity 1.2. Country Background 1.3. Economic Performance, 1980-98 II. Aid: Level, Types and Sources 8 2.1. Size of Aid Flow 2.2. Sources and Uses of Aid 2.3. External Debt: Stock and Service 2.4. Timing and Stability of Aid Flows III. Institutional Development and Policy Reform 23 3.1. Institutional Transformation under the Derg 3.2. Institutional Reform under the EPRDF 3.3. Policy Reform under the EPRDF 3.4. Reform Dialog: Government and Stakeholders 3.5. Donor Assistance Strategy 3.6. Donor Assessment of Reform Record IV. Relationship Between Aid and Reform 37 4.1. Where Aid Succeeds in Inducing Reform 4.2. Where Aid Fails to Generate Reform 4.3. The Coupling of Money and Ideas 4.4. Rethinking the Donor-Beneficiary Relationship V. Concluding Thoughts 48 VI. References 51 Appendix A: 55 Table A3.1: Chronology of Recent Economic Reforms, 1974-98 Table A3.2: Highlights of Donor Assistance Strategy, 1980-2001 Table A3.3: Highlights of the Policy Reform Matrices of PFP1-FFP5, 1992-2001 2 I. Introduction Development aid involves the transfer of resources from official or private institutions to low-income economies in the form of loans on concessional terms, technical assistance, and outright grants. External economic assistance has played an important role in the economic development programs of successive Ethiopian governments since 1960. The primary objectives of donors in Ethiopia have been the promotion of economic growth through support for investment and reform, alleviation of the unacceptably high rate of absolute poverty, and reduction of the vulnerability of the economy to adverse natural and terms of trade shocks. Other donor motives, especially geostrategic ones, have also generated military assistance which at times dwarfed development aid and threatened to nullify the latter's positive impact. 1.1. Aid: Form, Effectiveness and Selectivity The primary concern of this paper is the relationship between development aid and the process of economic reform in Ethiopia in 1980-98. However, a few words on issues pertaining to aid effectiveness and selectivity would provide a useful context. The effectiveness of official development assistance has recently been the focus of comprehensive cross-country studies, most notably by the Task Force on Concessional Flows (Cassen, 1994) and the World Bank (1998b). The Cassen report, relying primarily on country case studies, concluded that (a) the majority of official aid intended for long-term development has been successful; and (b) aid has not necessarily gone where it is most efficiently utilized. For example, South Asia has lost its share of aid to Africa where aid works relatively less well. The recent World Bank study, focusing on domestic factors, notes that (a) financial assistance has a big, positive impact where recipients have already undertaken sensible economic reforms; (b) financial aid flows do not discriminate between recipients with good policy and recipients with mediocre policy, (c) financial aid is generally fungible thereby making the assessment of aid effectiveness difficult; (d) program finance enhances the prospects of success for sectoral and project aid; (e) non-financial aid is more effective than money in inducing non-reforming governments to seriously contemplate removing distortionary policies; and (f) effective aid often comes as a package of finance and ideas which fosters the emergence of good institutions and sound policies. These findings plus the decline of aid by one-third in the 1990s have triggered a serious rethinking of the optimal composition and recipient-selectivity of aid programs. This trend is particularly of concern for Sub-Saharan Africa where aid intensity is high and aid effectiveness is 3 considered low (van de Walle and Johnson, 1996; World Bank, 1994; 1998b; O'Connell and Soludo, 1998). The mixed record on the effectiveness of external economic assistance to Ethiopia is reviewed in Maxwell (1998) and IDS/IDR (1996) for EU aid, in World Bank (1997a) for the IDA loan portfolio, and in Clay, Molla and Habtewold (1998) for food aid. The main conclusions of this paper can be summarized as follows. Aid had a minimal impact on growth or poverty in the 1980s since Ethiopia was embroiled in a protracted civil war along with a bold program of building a socialist economy. Market-oriented reforms were initiated in 1990 in response to profound economic and political crises. The reform program initially attracted a sizeable increase in external assistance in support of liberalization, stabilization and rehabilitation. Subsequent aid helped to deepen reform commitments and supported high growth rates. The residual dirigisme of the ruling party, however, appears to frustrate the enactment of important second-generation reforms. The rest of the paper is organized as follows. Following a brief look at the economic profile of the country, I present data on the level, source and composition of aid in part 2. The political economy of institutional and policy reform under the Derg and EPRDF governments is discussed in part 3. Hypotheses concerning the relationship between aid and reform are then taken up in part 4. Part 5 offers a summary and concluding thoughts. A time profile of reforms undertaken during the period under review is supplied as an appendix. 1.2. Country Background Ethiopia has a number of peculiar characteristics, which must be kept in mind while evaluating the record on the role of aid in the country's development strategy. They include a high demographic and ecological diversity, an extremely high rate of absolute poverty and economic insecurity, a predominantly subsistence and coffee-based monocrop economy, and an old but resilient state subjected to autocratic rule and endemic social unrest. Its 60 million people are highly diverse in ethnic, religious, and geographic terms. According to census figures (GOE, 1998b), there are over eighty distinct ethnic groups. However, two equal-sized ethnic groups (the Oromo and the Amhara) together account for over two-thirds of the population. About 60 percent of Ethiopians are Christian, a third are Muslims, and the remainder are adherents of various indigenous religions. In the urban areas, there are distinct occupational specializations among the various ethnic groups. Ecological diversity is unusually high (with 18 agro-ecological zones), especially between the densely-settled mixed farming regions of the temperate central highlands and the transhumant lowlands. Ethiopia is arguably the poorest country of its size by almost all measures of poverty. Despite its good agricultural endowments, it has yet to attain sustained food security. According 4 to a 1981-82 survey, nearly 90 percent of the population was below the poverty line of US$2 per day. In 1995/96, 43 percent of farm households were food insecure. Famines, regional and nation-wide, occur with distressing frequency thereby converting much of the transitory poverty into chronic poverty (Ramachandran, 1997; Clay, Molla and Habteworld, 1998). The worst rural poverty is found in the famine-prone northern highlands (especially in parts of Tigray and Wollo) and in the lowlands that envelop the country's international borders (most notably, the Afar, Somali, Borana, Gedeo and Omo areas). Interestingly, the rate of urban poverty may very well have approached the level of rural poverty in the 1990s (see Middlebrook and Corzato, 1999 for poverty in Addis Ababa). An assessment of the relationship between growth and poverty reduction is frustrated by the lack of reliable data on the historical trends of poverty rates. The country's social indicators reflect its low per-capita income. Half the population does not have access to health facilities within ten kilometers. Life expectancy at less than 50 years is below the average for Sub-Saharan Africa, as are school enrollment ratios. Only 40 percent of relevant school-age children are enrolled in primary schools and just 15 percent attend secondary schools. 1.3. Economic Performance, 1980-98 Ethiopian economic data are generally incomplete or inconsistently reported, especially for periods longer than five years. I have constructed a number of economic series for the period under review, piecing together the more reliable Birr-based series from disparate sources.1 The basic macroeconomic indicators are reported in Table 1.1. They are averaged over distinct sub- periods paralleling notable reforms or shocks to the economy (see also Table A3.1). Seven time segments are identified with the intention of capturing major institutional changes or policy episodes: 1980-83 (zemetcha years), 1984-85 (nation-wide famine), 1986-89 (central planning), 1990-91 (liberalization and political change), 1992-94 (recovery and stabilization), 1995-97 (structural reforms), and 1998-2001 (sector development programs). The growth rate of real GDP over the past twenty years was modest but respectable. The predominantly agricultural economy (accounting for 55% of GDP) faced a number of shocks, including two waves of systemic change, and several adverse shocks (fluctuations in commodity 1 For this reason, the growth rates reported here do differ (mostly insignificantly) from those reported in official publications for shorter periods. The Ethiopian economy is among the least researched in Africa. 5 Table 1.1 Ethiopia: Basic Economic Indicators Item 1980-83 1984-85 1986-89 1990-91 1992-94 1995-97 1998-00* Annual Avg. Growth: (1990 Birr) Real GDP 5.5 -5.3 5.3 3.0 7.6 7.8 2.5 Population 2.8 2.8 3.0 3.1 2.9 2.9 3.0 GDP per head 2.7 -8.1 2.3 -0.1 4.7 4.9 -0.5 Exports 8.8 28.0 3.1 -18.0 89.4 12.1 3.2 Imports 3.8 3.3 -1.2 -32.6 -16.9 12.6 5.4 Gross Dom. Invest. 30.0 -40.2 0.0 -5.0 32.3 27.5 .. Money Supply, M1 12.2 17.0 10.7 19.6 13.0 4.4 10.0 GDP Deflator 6.5 26.0 0.8 21.4 11.7 3.4 2.4 CPI** 3.8 29.1 4.2 35.7 5.7 -4.3 3.8 As % of GDP: Gross Dom. Investment 12.5 13.8 16.7 11.5 12.9 18.0 20.3 Private Consumption 78.8 78.6 75.4 76.5 84.8 81.7 80.0 Govt. Consumption 15.5 16.3 16.4 17.9 11.6 12.0 11.0 Net Exports & NFR -6.8 -8.8 -8.0 -6.5 -9.6 -11.8 -11.7 Gross Dom. Savings 5.7 5.0 8.7 5.4 3.3 6.2 8.5 (continues) (**) The CPI is for Addis Ababa only, and it excludes housing cost. A national CPI has been introduced in 1995/96. For 1990-97, the growth rates of the three price indicators are as follows: GDP deflator (8.9%), CPI (6.9%) and Food Price Index (12.4%). The complete data for the GDP-based inflation rate are as follows: 1970-98 (7.8%), 1970-75 (4.8%), 1976-80 (15.0%), 1980-90 (5.2%), 1991-92 (21.0%) and 1993-98 (3.6%). 6 Table 1.1: Ethiopia--Economic Indicators (continued) 1980 1997 Levels (millions of 1990 Birr): Real GDP 13485 28361 Population (millions) 38 60 GDP per head 358 473 Real Exports 1912 4238 Real Imports 10555 20033 Exchange Rate (ETB/US$) 2.1 6.5 Quality of Life Indicators: Poverty Rate (1981-82 survey): Below US$1/day (% of population) 46.0 44.0 (1995) Poverty Gap (% GDP) 12.4 na Below US$2/day 89.0 na Poverty Gap (% GDP) 42.7 na Under-5 Mortality Rate (per `000) 213.0 177.0 Gross Enrollment Ratio (% ages 6-23) 16.0 18.0 Country Policy and Institutional Assessment (CPIA): (average scores for 1998) Ethiopia Africa (avg.) Macroeconomic Management 4.3 3.2 Sustainable and Equitable Growth 3.1 3.1 Reduction of Inequality 3.8 2.9 Public Sector Management 3.9 2.8 Overall Rating 3.7 3.0 Notes and Sources: IMF, International Financial Statistics 1998; February 1999; Ethiopia: Statistical Appendix, 1998. World Bank, World Tables 1995; World Development Report 1998/99. UNDP, Human Development Report 1998. World Bank (CPIA data): 3 = Moderately Unsatisfactory, 4 = Moderately Satisfactory, 5 = Good. (*) Projected: based on constant factor cost; GOE, Policy Framework Paper (October 1998). ETB Ethiopian Birr (1 US$ 8.0 ETB in July 1999) Poverty Gap Size of income transfer needed to lift the poor above the poverty line, expressed as a fraction of GDP. 7 prices, civil strife and droughts). During 'normal'years, real per capita income grew at an annual rate of 2.6-3.3 percent. The corresponding growth rate for GDP was 5.5 percent in the 1980s and 6.2 percent in the 1990s. These improvement, though hard to disentangle, is attributable to good weather, higher coffee prices, and allocational efficiency gains from the zemetcha (or mass mobilization) campaigns, economic liberalization, and substantial inflows of development assistance. When the crisis years (1984-85 and 1990-91) are included, however, real GDP grew at the annual rate of only 2.6 percent during 1980-97--replicating the performance of the 1970s. Agricultural growth, which carries half the sectoral weight, averaged 0.6 percent per annum in the 1980s and 3.0 percent in the 1990s. According to the IMF, the annual average growth rate of real GDP for 1970-98 was 2.86 percent. In light of the high rate of population growth (variously estimated at 2.5-3.0 percent per annum), the average Ethiopian is at best as well off today as she was in 1970. On the other hand, the record on macroeconomic stability has been enviable. The inflation rate rarely exceeded single digits (annual average of 7.8 percent for 1970-98), and the fiscal deficit was kept within manageable levels with some exceptions (see tables 1.1 and 1.2). II. Aid: Level, Types and Sources I use two broad measures of official aid. The first, Net ODA, is the sum of official grants (including technical assistance and food aid) and concessional loans with a grant element of at least 25 percent. It is calculated net of amortization payments. The second, Effective Development Assistance (EDA) is of the sum of discounted official grants and the grant equivalent of concessional loans (i.e., the dollar value of the grant element). Dissatisfied with the shortcomings of Net ODA, Chang, Fernandez-Arias and Serven (1999) constructed EDA as a superior measure of the real cost of aid to donors (though not necessarily to the recipient). 2.1 Size of Aid Flow The data on the time pattern of resource flows to Ethiopia show some inconsistency even between the DRS (which appears to include portions of CMEA aid) and CRS databases. I rely primarily on CRS data. Alternative measures of aid will also be utilized as shown in Table 2.1. 8 Table 1.2 Ethiopia: Fiscal Trends Baseline: Revenue/Expenditure 1980-85 1986-90 1991-95 1996-97 1998-99* I. REVENUE AND EXPENDITURE (as % of GDP) Total Revenue 15.4 24.6 17.8 21.5 20.2 Domestic 15.2 21.0 15.0 17.9 16.7 External Grants 0.2 3.6 2.8 3.6 3.6 Total Expenditure 20.5 31.7 23.1 23.8 24.9 Recurrent .. 22.1 14.8 13.5 13.9 Capital .. 9.6 8.2 10.4 11.0 Fiscal Deficit (cash basis) Before Grants -5.3 -10.7 -8.1 -5.9 -7.8 After Grants -5.1 -7.1 -5.3 -2.2 -4.3 Average GDP (millions of current Birr) 10841 15790 29523 43335 45204 II.. ALLOCATION OF EXPENDITURE (% of total) General Administration 7.2 10.7 10.8 Defense 30.8 10.6 7.9 18.0 Economic Infrastructure 5.6 10.7 16.5 of which: Roads 1.9 5.0 10.3 Economic Development 24.4 23.1 22.0 of which: Agriculture 7.9 7.4 6.6 Social Services 15.9 23.2 25.1 of which: Education 9.5 14.1 15 Health 3.3 5.1 6.4 Others 16.1 21.6 17.9 of which: Interest & Charges 4.8 10.5 8.6 External Assistance 6.1 2.8 0.0 Total Expenditure (millions of Birr) 4937 6808 10471 11917 Sources: *World Bank, Ethiopia: Review of Public Finances, vol. 1 (1998a), tables 1. 2 and 3.1. IMF, International Financial Statistics 1998. GOE, 1999 Ethiopia Public Expenditure Review, Draft Aide Memoire. 9 Table 2.1 Ethiopia: Official Development Assistance, 1980-97 In millions of US$ Year Net ODA1: Net ODA ODA IDA IMF FDI Loan+Grant Loans Grants Credits Credits (CRS) (CRS) (CRS) (DRS) (DRS) (DRS) 1980 212 43 169 28 9 0 1981 245 46 199 28 73 0 1982 200 40 160 28 26 2 1983 339 126 213 42 0 -3 1984 361 75 286 41 0 5 1985 710 92 619 50 0 0 1986 629 67 563 38 41 -1 1987 627 134 493 86 0 -3 1988 963 188 775 75 0 2 1989 749 132 617 70 0 0 1990 1020 161 858 74 0 12 1991 1097 110 988 59 0 6 1992 1182 106 1076 112 20 6 1993 1094 361 733 230 30 7 1994 1076 311 764 150 20 3 1995 888 246 642 84 0 8 1996 850 227 623 142 21 5 1997 637 72 565 65 0 5 (continues) DRS Debtor Reporting System (World Bank data base) CRS Creditor Reporting System (OECD data base) 10 Table 2.1 (continued) Ethiopia: Official Development Assistance, 1980-97 In millions of US dollars Technical Food Net Net Debt ODA Year Assistance Aid ODA2ODA3 Service Deflator (CRS) (CRS) (DRS) (CRS) (DRS) (CRS) 1980 44 30 274 286 45 0.53 1981 64 54 1251 363 58 0.51 1982 53 38 1625 291 76 0.49 1983 64 41 1016 444 107 0.49 1984 81 91 929 533 130 0.47 1985 104 167 1296 981 159 0.48 1986 115 175 1108 919 224 0.60 1987 147 91 1000 865 248 0.69 1988 197 157 1263 1317 309 0.75 1989 234 97 902 1080 304 0.74 1990 254 195 1201 1469 236 0.83 1991 206 265 1565 1568 138 0.86 1992 190 222 1484 1594 109 0.91 1993 171 118 1136 1383 95 0.89 1994 126 48 958 1250 112 0.93 1995 167 125 837 1180 154 1.02 1996 202 70 657 1122 347 1.00 1997 168 75 723 880 99 0.92 Notes and Sources: World Bank, Global Development Finance (1999 CD-ROM). DRS data. OECD, Geographic Distribution of Financial Flows (1999 CD-ROM). CRS data. Net ODA Loans Loan Disbursements, net of amortization Net ODA1 (CRS) Net ODA Loans + Grants (= financial aid) Net ODA2 (DRS) Net Flows on Debt + Grants + Technical Assistance + Food Aid Net ODA3 (CRS) Net ODA Loans + Grants + Technical Assistance + Food Aid FDI Foreign Direct Investment (private) (1) ODA2 and ODA3 are the broadest measures of aid given the sometimes dubious assumption that food aid releases investable dollars that would have otherwise financed essential food imports. Food aid (93% of which comes in the form of grains) has recently been equivalent to a quarter of national grain sales (Clay, et al., 1998). (2) The above FDI figures are underestimates. During 1992-98, the Ethiopian Investment Authority approved 163 projects worth US$1.4 billion (mainly Saudi and European). Of these, only 21 projects are operational and another 49 projects are under implementation (UNCTAD/ICC, 1999, tables 5 and 6). 11 Ethiopia's ODA receipts exceeded US$17 billion in nominal terms (US$23 billion in real terms) during 1980-97. This comes out to US$1.0 billion (nominal) or US$1.3 (real) billion of annual inflow. In 1996 prices, the annual inflow of Net ODA3 (including loans, grants, technical assistance and food aid) to Ethiopia averaged US$1.2 billion per year in the 1980s. It rose to US$1.4 billion per year in 1991-96 before slowing down subsequently. In constant dollars, the pattern is M-shaped with peaks occurring in 1985 (famine effect) and 1992 (honeymoon effect). This pattern suggests that the ODA premium for the ambitious reform program of the 1990s has been a modest increase of 17 percent. In terms of aid intensity, the relevant figures are as follows. The ODA3-to-GNP ratio rose from 12 percent in the 1980s to 23 percent in the 1990s yielding an average of 16 percent for 1980-97. In per capita terms, receipts rose from US$15/$23 (nominal/real) in the 1980s to US$27/$30 in the 1990s for a period average of US$20/$26 during 1980-97. The table shows other interesting patterns of aid flow. First, the share of grants in net cash disbursements rose substantially to exceed 50 percent after 1985. Second, debt service levels reached their peak, in absolute terms and relative to exports, in the second half of the 1980s. Third, Ethiopia relies heavily on ODA from four multilateral sources (IDA, EC, AfDB and U.N.) and a handful of European bilateral donors (mainly the Nordic countries, Germany and Italy). Although it has managed to obtain a steady average annual flow of US$ 275 million in trade credit, Ethiopia remains distinctly unattractive to private investment from abroad. The level of FDI tripled to US$178 million between 1996 and 1997, but still falls far below East African levels even in absolute terms (UNCTAD/ICC, 1999). Data on EDA (Table 2.2) confirms much of the trend captured by the conventional measures. The "effective" grant equivalent (GEQ) of concessional ODA loans averages 30 percent. This proportion rose from about 21 percent in the 1980s to 46 percent in the 1990s, which compares unfavorably with an average grant element (AGE) of 70 percent for 1991-97 (see Figure 2.2). In real terms, the size of EDA loans was on average twice as high in the post- socialist period as in the socialist period. In terms of grant equivalency, bilateral loans declined to modest levels in the 1990s. Finally, the share of technical assistance in total aid was 20 percent--below the SSA average of 25 percent (see O'Connell and Soludo, 1998, Table 1). 2.2. Sources and Uses of Aid Being a non-colony, Ethiopia did not until recently attract the level of development assistance that it needed and could effectively absorb. It was, nonetheless, the largest recipient of the modest U.S. economic assistance to Sub-Saharan Africa during 1961-76. The Commission 12 Table 2.2 Ethiopia: Effective Development Assistance, 1980-96 In millions of US$ Year TEDA EDA2 EDA1 GEQ Grants TA GEQ: GEQ: TA/EDA2 Multilat. Bilater. 1980 239 209 165 40 125 44 26 14 21 1981 444 390 326 191 135 64 33 157 16 1982 269 231 178 71 107 53 32 39 23 1983 406 365 302 152 150 64 50 102 18 1984 553 362 281 76 205 81 42 34 22 1985 923 756 651 137 514 104 65 72 13 1986 855 680 564 117 447 115 49 69 17 1987 742 651 504 158 346 147 95 63 23 1988 1097 934 737 159 578 197 93 66 21 1989 848 751 517 134 383 234 83 51 31 1990 1175 980 726 122 604 254 97 25 26 1991 1347 1082 876 95 781 206 81 8 19 1992 1393 1171 981 95 886 190 119 8 16 1993 1098 980 809 247 562 171 246 0 17 1994 964 916 790 152 638 126 152 0 14 1995 898 773 606 131 475 167 125 0 22 1996 850 780 578 158 420 202 158 0 26 Notes and Sources: TEDA Total Effective Development Assistance (= GEQ + Grants + TA+ Food Aid) EDA1 Effective Development Assistance ( = GEQ + Grants) EDA2 Effective Development Assistance ( = EDA1 + TA) GEQ Grant Equivalent of an ODA loan TA Technical Assistance (=Technical Cooperation) . Based on data developed by Charles Chang, Eduardo Fernandez-Arias and Luis Serven, "Measuring Aid Flows: A New Approach," Policy Research Working Paper 2050 (World Bank, February 1999). 13 Figure 2.1 Ethiopia: Average Grace Period and Maturity of ODA Loans 45 40 35 30 s Yearfo 25 er 20 mbu N 15 10 5 0 1970 1971 1972 1973 197 1975 1976 197 1978 1979 198 1981 1982 198 1984 1985 198 1987 1988 198 1990 1991 199 1993 1994 199 1996 1997 4 7 0 3 6 9 2 5 AGP (YEARS) AM (YEARS) Figure 2.2 Ethiopia: Average Grant Element and Share of Multilateral Debt 80 70 60 50 e tag 40 cen Per 30 20 10 0 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 AGE (%) MD/EDT (%) 14 of the European Communities under Lomé was especially generous providing grants of US$2.5 billion in terms 1990 dollars during 1976-94 (Maxwell, 1998, p. 105). Other multilateral agencies (notably IDA and AfDB) and bilateral donors (especially Italy and Sweden) provided the bulk of the external assistance. The U.N. agencies (most notably UNDP and WFP) were also active. WFP alone provided US$600 in food assistance during 1990-93. Between 1980 and 1997, 60 percent of grants in cash or in kind came from Europe and another 20 percent from the United Nations agencies. Since the mid-1990s, there has been a gradual decline in commitments by the EU and Italy. This trend has been countered by increased support from IDA, Germany and the United States. Japan, the Netherlands and Canada also increased their aid in the 1990s. The Nordic countries as a group present an interesting model of stable and predictable aid provision to Ethiopia (Table 2.3). Nearly all the concessional loans come from two multilateral institutions: the International Development Association (of the World Bank Group) and the African Development Fund (of the African Development Bank). The former made its first African loan to Ethiopia in 1950. The latter began its lending program to Ethiopia in 1975 (as did the EC), and it has beefed up its commitments in the past five years (AfDB, 1996). The CRS data reveal other interesting patterns concerning DAC assistance to Ethiopia. The share of DAC bilateral aid in Net ODA1 fell from 60 percent in 1991-93 to 45 percent in 1994-97; ODA accounts for 98 percent of official development finance (ODF) from all sources; one-third of technical cooperation grants come from non-DAC sources; and there has been a noticeable decline of both net disbursements and commitments after 1995. The distribution of aid disbursement by sector provides useful information on the priorities of donors, creditors and the two Ethiopian governments. Table 2.4 presents data on bilateral commitments by purpose. For net loans and grants from all sources, the gap between commitments and actual disbursements is generally small, except during the second half of the 1980s when the surge in external support commenced on the heels of the 1984 famine. The table shows that bilateral ODA commitments are largely accounted for by expenditures on social infrastructure, program aid and food aid. Program assistance was especially strong (nearly half 15 Table 2.3 Ethiopia: Net ODA Loans and Grants by Creditor or Donor, 1970-97 In millions of US$ of which: ----------- of which: ----------------------------------------------------------------- Year Loan: ADF IDA Grants: EC UN Germ Italy USA Nordic Japan Nethld 1970 14 0 3 26 0 4 4 3 7 13 0 0 1975 63 0 19 71 8 16 6 3 9 18 1 3 1980 43 3 28 169 32 48 10 8 20 38 1 4 1981 46 11 28 199 42 67 11 15 4 30 3 4 1982 40 7 26 160 27 58 11 4 3 17 1 3 1983 126 12 41 213 47 15 10 14 8 32 3 3 1984 75 5 39 286 58 64 29 12 21 62 4 9 1985 92 17 48 619 103 108 27 63 146 236 8 8 1986 67 15 35 563 85 90 24 134 94 251 7 8 1987 134 23 82 493 93 103 27 98 8 133 16 7 1988 188 17 72 775 134 156 38 158 69 265 14 23 1989 132 28 66 617 45 192 41 119 26 185 13 11 1990 161 43 69 858 102 203 46 131 53 231 11 25 1991 110 41 54 988 159 367 48 77 88 212 18 26 1992 106 50 106 1076 292 239 204 41 68 313 11 29 1993 361 117 225 733 123 183 50 12 131 193 48 34 1994 311 61 156 764 182 76 89 41 122 252 44 33 1995 246 82 77 642 82 107 76 27 70 173 63 40 1996 227 79 128 623 53 117 81 40 56 178 50 60 1997 72 24 50 565 42 142 58 32 60 150 37 35 Notes and Sources: OECD, Geographic Distribution of Financial Flows 1999 (CD-ROM). Nordic here consists of Sweden, Norway and Finland. Two other less generous but still important DAC donors not listed here are Canada and the U.K. ADF African Development Fund (of the African Development Bank, AfDB) DAC Development Assistance Committee 16 Table 2.4 Ethiopia: Bilateral ODA Commitments by Purpose, 1973-97 In millions of US$ of which: ------------------------------- of which: ------- Year Total Infrastr+ Infrast- Infrastr: Infrastr: Program Food Prod- Commit. Program ructure Social Ecnomic Aid Aid ction 1973 32 21 20 13 7 2 0 2 1974 42 10 10 4 5 0 0 21 1975 67 51 38 29 9 14 8 16 1976 17 12 5 5 0 7 7 5 1977 31 17 7 6 1 9 9 13 1978 53 26 18 3 15 8 8 10 1979 42 23 5 5 0 18 18 7 1980 48 34 6 6 0 23 23 12 1981 80 57 8 8 0 41 36 25 1982 69 86 31 31 0 25 25 8 1983 76 57 10 6 4 36 36 16 1984 202 154 41 8 32 73 73 52 1985 308 268 61 24 37 145 145 45 1986 231 184 22 17 5 140 140 26 1987 361 170 43 31 12 84 84 62 1988 486 500 164 64 101 172 135 78 1989 278 193 48 25 24 96 90 82 1990 353 235 38 29 9 159 157 37 1991 210 185 16 7 8 154 150 20 1992 430 371 106 95 11 160 155 33 1993 368 312 79 67 12 154 111 64 1994 306 289 119 74 45 50 50 48 1995 566 420 147 104 43 126 85 58 1996 393 377 156 142 13 65 46 98 1997 295 284 103 82 21 77 65 46 Notes and Sources: OECD, Geographic Distribution of Financial Flows, 1999. 1. Program Aid includes food aid, and BOP and budget support. 2. Remainder categories include action relating to debt, emergency assistance and unspecified. 3. Reported Total Commitment does not always equal the sum of its components (e.g., 1988). 17 of bilateral resources) during the unstable years of 1985-93. Production sectors such as industry and agriculture received much less as is often the case in Sub-Saharan Africa. 2.3. External Debt: Stock and Service At the end of 1997, Ethiopia's external debt stock (EDT)--nearly all of which was public debt--was estimated at US$10.0 billion (see Table 2.5). The distribution of debt by ownership is as follows. Some 62 percent is owed to bilateral lenders (mainly to Russia for military purchases during the Soviet era). Multilateral creditors accounted for over one-third of the total. IDA, with a portfolio of $1.5 billion owns half of the multilateral debt. Private creditors claim the remaining 5 percent of the total debt stock. The data also show that short-term debt accounts for less than 5 percent of the total debt stock. Until recently, Ethiopia has had a record of remarkable diligence in servicing its debt. Arrears of both principal and capitalized interest increased substantially after 1990, reaching a whopping 53 percent of EDT by 1998. The real debt stock per capita in the 1990s was about six times as high as aid flow per capita (US$180 as against US$30). In NPV terms the servicing of Ethiopia's external debt burden is quite unsustainable (Ghani and Zang, 1995). In current value terms, debt service payments averaged 27 percent of GNP in the 1980s before falling to 23 percent in the 1990s. International reserves were, in fact, down to less than one month of imports in 1988-91. At the beginning of 1998, the debt inclusive of arrears to Russia represented 963 percent of exports (see Table 2.6) or 159 percent of GNP. The forecasts in PFP5 show that the debt-to-export ratio will fall to 265 percent in 2001 (GOE, 1998a). Ethiopia was granted debt relief on enhanced Toronto terms by the Paris Club in December 1992. The relief involved debt cancellation (US$ 101 million) and debt rescheduling (US$ 271 million). It also bought back some of its commercial debt (US$ 226 million) in 1995 at a rate of 8 cents per dollar, utilizing the World Bank's debt buy back facility. The preliminary review of eligibility for the Heavily Indebted Poor Countries (HIPC) has been completed for Ethiopia. Finalization of the debt relief package is pending settlement of the Ethio-Eritrean conflict. 18 Table 2.5 Ethiopia: Total Debt Stocks, 1980-97 In millions of US$ Total IDA: ARREARS: Export Year EDT LDOD LDOD SDOD IMF Int. + Prin. Credit 249 2 1980 824 688 57 79 .. 276 2 1981 1842 1638 59 145 .. 302 2 1982 3280 3051 68 161 120 342 2 1983 3845 3649 63 134 177 379 2 1984 4220 4054 68 98 279 437 1 1985 5206 5057 77 71 297 486 1 1986 6134 5967 83 84 323 601 55 1987 7364 7191 98 76 389 658 16 1988 7704 7515 134 55 372 718 69 1989 7842 7700 112 30 323 824 279 1990 8634 8483 145 6 311 883 1088 1991 9119 8843 276 0 292 964 1780 1992 9341 9003 319 19 239 1187 2105 1993 9703 9287 368 49 235 1373 3204 1994 10067 9570 424 72 234 1470 4067 1995 10309 9775 461 74 239 1555 4784 1996 10078 9484 502 92 300 1532 5298 1997 10079 9427 565 87 237 Sources: World Bank, Global Development Finance 1999. The above include Ruble-denominated defense loans from the former Soviet Union (FSU). The Ministry of Finance puts the external debt as of June 1998 much lower at US$7.8 billion, of which US$4.1 billion was owed to Russia. EDT Total Debt Stocks, including IMF and SDOD LDOD Long-term Debt Outstanding and Disbursed SDOD Short-term Debt Outstanding and Disbursed 19 Table 2.6 Ethiopia: Debt Stock and Debt Service Indicators, 1980-97 Percentages Year CL/EDT RES/MGS RES/TDS EDT/XGS TDS/XGS EDT/GNP (months) 1980 68 4 585 140 8 16 1981 78 5 643 326 10 36 1982 84 3 367 610 14 60 1983 86 3 192 666 19 64 1984 85 1 84 674 21 74 1985 86 2 136 932 28 78 1986 86 3 148 891 33 88 1987 87 2 90 1150 39 99 1988 86 1 48 1189 48 101 1989 87 1 40 1035 40 99 1990 87 1 23 1276 35 127 1991 84 1 77 1667 25 172 1992 84 3 247 2037 24 169 1993 84 4 527 1889 19 158 1994 84 6 526 1788 20 209 1995 85 7 529 1277 19 180 1996 86 5 211 1224 42 169 1997 86 3 507 963 10 159 Source: World Bank, Global Development Finance 1999. Includes Ruble-denominated defense loans from FSU. CL Concessional Loans EDT Total External Debt RES International Reserves XGS Exports of Goods and Service TDS Total Debt Service MGS Imports of Goods and Service 20 Ethiopia also stands to benefit from the more generous Cologne Summit Declaration on HIPC. The annual saving from debt relief, however, appears to be modest given the recent trends of resorting to arrears. Assuming a 70 percent reduction in debt service payments from debt relief, the money that would be available for additional social spending is about US$50-$100 million (i.e., given the wide range of estimated annual debt service payment of $150-$350 million in recent years). Ethiopia also continues to press for a major write-off of Russian debt.2 2.4. Timing and Stability of Aid Flows In a world of imperfect markets, the impact of aid on growth depends in part on its timing and degree of instability (predictable or not). In the case of Ethiopia, the critical policy focus is on the quantity (and quality) of investment. Foreign aid (which accounts for over one-fifth of GNP) finances nearly half of the capital expenditure in the federal budget. Furthermore, in the spirit of two-gap models, an important determinant of changes in Ethiopian capital-good imports is the time path of forex earnings from exports (Etherington and Yainshet, 1988; Lensink and Morrissey, 1999; World Bank, 1998a). An eyeballing of figures 2.3 and 2.4 shows some interesting patterns in this regard. First, the variation in the growth rate of Net ODA1 is much higher than that of exports (or GDP). Generally, multilateral aid is more stable than bilateral aid; and project aid is steadier than program aid. Second, the correlation between the growth rates of export revenues and ODA (grants and loans) is low and negative (Pearson r = -0.25). That means, the changes in the flow of aid do not have a systematically countervailing relationship with fluctuations in export earnings. This is due to a number of factors including the changing share of non-project aid, and the bunching of aid receipts relative to commitments which are driven largely by bureaucratic factors. It also means that ODA flows are not entirely geared at stabilizing the overall supply of foreign exchange. 2 The 1997 accord between Russia and the Paris Club calls for an upfront discount of 80 percent on the stock of debt owed and the application of Naples terms to the remainder (which means another 67 percent reduction in NPV terms of the remainder 20 percent). If these assumptions hold (i.e., a 90% Russian debt write-off) the debt stock will fall to US$2.8 billion and the debt-service to 184 percent of exports by 2001. Almost all of the Russian debt is military-related. 21 Figure 2.3 Ethiopia: Growth Rates of GDP, Exports and Net ODA (1996 USD) 120.0 100.0 GDP Growth 80.0 EXPORT Growth ) %( 60.0 Net ODA1 Growth tea R th 40.0 owr Gl nnua 20.0 A 0.0 Fi -20.0 Year -40.0 F igure 2.4 E thiopia: S ources and U ses of F oreign E x change (1996 U S D ) 3 0 0 0 R R E C E IP R X G S 2 5 0 0 F O R E X R M G S 2 0 0 0 llarsodfo 1 5 0 0 sn illio M 1 0 0 0 5 0 0 0 1970 72 19 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 Notes and Sources: 1. Net ODA1 refers to financial aid (grants and net loans) only to make it comparable to export earnings. It averaged US$500 million (or US$700 million in 1996 prices) during 1970-97. 2. Receipts (RECEIP) include all forex receipts from ODA and non-ODA sources, excluding exports. 3. FOREX refers to total foreign exchange inflows, including export earnings. 4. X (M) GS means exports (imports) of goods and services. IMF, IFS98 (for GDP), World Bank, GDF99 (for imports and exports), and OECD GDFF99 (for ODA, Receipts and Deflator). 22 Third, the level of total receipts (inclusive of ODA, OOF and private flows) does compensate for the volatility of export earnings (r = -0.57) so that total forex flows are reasonably stable. As shown in Figure 2.4, this produces the expected synchronization between total forex receipts and the import bill. Two inferences emerge from this cursory overview. Increases in the share of program aid, by stabilizing total forex inflows, would help to reduce growth-retarding uncertainty for investors (Lensink and Morrissey, 1999; Collier and Gunning, 1999). Furthermore, the rising share of non- project aid may bring more flexibility, but at the expense of aid predictability since the tap may be more easily turned on and off by disgruntled donors. III. Institutional Development and Policy Reform During the period under consideration, Ethiopia underwent two waves of radical economic reform: one revolutionary and anti-market in orientation, and the other more modest and mixed economy in orientation. The former reflects the pervasive influence of the left-wing student movement, and the latter a reluctant retreat from socialist economic ideas by the products of the same movement. The chronology of institutional and policy reforms introduced by the two regimes are highlighted in Table A3.1 (located at the end of the paper). 3.1. Institutional Transformation under Derg The period 1974-90 was one of revolutionary turmoil following the coup d'etat against the monarchy, civil wars, a war with Somalia, and an experiment with agrarian socialism. By the end of the 1970s, the broad outline of a quasi-socialist economy was in place: all land, extra houses and large as well as medium private enterprises were nationalized without compensation. Owner-cultivators and share-croppers were given use rights (usufruct) to land in lieu of their customary freehold or communal ownership rights. The rural population was then organized into peasant associations (PAs). These state-sponsored PAs were soon to serve as stepping stones for the establishment of producer cooperatives or quasi-collective farms. Much of inter-regional 23 wholesale and retail trade (in key staples and important industrial goods) was reserved for parastatals. The military regime's vision of state-led industrialization included an over-zealous "villagization" campaign (to facilitate political control as well as the delivery of social services), an ambitious but poorly planned program of "resettlement" of the drought-vulnerable population, and a program of import substitution for key industrial goods. By mid-1988 villagization encompassed 12 million people (one-third of the rural population) in over 12,000 large villages (Clapham, 1988, p. 251). Though briefly distracted by the devastating 1984-85 national famine (when GDP contracted by 13.2 percent), the Derg consolidated its power in the second half of the decade. It formulated a Ten-Year Perspective Economic Plan (1985-94) with an unrealistic expectation of receiving foreign aid to finance over half of the planned investment (World Bank, 1987; Griffin, 1992). The perspective plan nevertheless provided a more consistent framework for medium-term and annual plans than the previous disparate-projects approach. The economy rebounded briefly but spectacularly at the rate of 9.3 percent and 13.8 percent in 1986 and 1987, respectively. To its credit, the Government also expanded elementary and secondary education and rural health clinics at unprecedented rates. The ambitious literacy campaign was awarded a UNESCO prize. However, two-thirds of the adult population still remains illiterate today. The military regime also belatedly established a socialist party, the Workers' Party of Ethiopia (WPE), with military officers and the upper echelon of the civil service at its core. Ethiopia finally joined the "peoples democracies" of the Soviet bloc in 1987 with the declaration of the Peoples Democratic Republic of Ethiopia (PDRE) under a constitution ratified by a national referendum. The Derg intensified the administrative centralization program of the monarchy under a unitary state, and implemented an unprecedented degree of state ownership and control of the modern sector of the economy. The country's 15 regions were reorganized into thirty administrative units. Imitating the nomenklatura systems of the socialist states, the regime established parallel networks of government and party organs down to the level of the kebelle (neighborhood)--all controlled by WPE functionaries. The Derg also introduced a still enduring 24 cadre system of party control of the economic, military and administrative sectors of the state bureaucracy. A number of developments accentuated the "normal" level of crisis in the waning days of war socialism, which eventually led to policy reforms. Firstly, the burden of financing a huge army was too high for the feeble domestic economic base to support, especially since the Soviet Union was itself in the midst of a profound crisis after 1988. Secondly, the idea of socialism was discredited internationally and domestically. Thirdly, the economic consequences of the campaigns of villagization, resettlement, and producer cooperatives proved disastrous for an already impoverished population. The economy cumulatively shrank by 7.0 percent during and immediately after the height of the civil war, i.e., 1988-92. While most captive civil servants could do little to arrest the double squeeze on salaries (pay freezes and double-digit inflation), small producers increasingly resorted to the age-old but inefficient methods of survival-- informal exchange and subsistence production. The IDA maintained a modest portfolio of loans (US$35 million annually) during 1976- 86. It confined its funding to economic and sector work (technical assistance and country economic memoranda), and financing of selected social services, transportation and communication, and the important coffee sector. Loan disbursement rose to US$52 per year during 1989-90 with an agreement to support an agricultural extension and credit program called PADEP (more on this below). As shown in Table 2.4, the Government managed to obtain a respectable level of grant money in the 1980s (on a per year basis): EU-EC (US$70 million), Italy (US$68 million), Scandinavian countries (US$135 million), Germany (US$25 million) and the U.S. (US41 million, mostly food aid). Data on CMEA economic aid are sketchy but it was rather modest (US$30 million in 1983) and came primarily in the form of project aid (Clapham, 1988).3 Liberalization was seriously considered by the Derg after 1988 although the implementing legislation was not in place until the WPE Congress of March 1990 (Haile Mariam, 1990)--a year before the regime's demise. The impetus appears to have come from the pressure created by the steady gains of the northern rebel movements, and the disappointing 3 The projects included Komolcha Textile Mill, Mugher Center Works, Nazret Tractor Assembly, and Melka Wakena hydroelectric project. 25 economic results of agricultural cooperatives and parastatals in the distributive sector. The reforms were in the nature of "perfection of control" rather than significant concessions to the market. Their significance lies in the dynamic they created for further reform. The most notable measures were the following. Inheritable use rights were granted on nationalized land, membership in producer cooperatives became voluntary (which led to their spontaneous dismantling by jubilant cooperators), many prices were decontrolled, and the draconian restrictions on private trade were grudgingly lifted. The liberalization of the grain market, the most important market in Ethiopia, paved the way for a steady (but still incomplete) integration of inter-regional markets. Most foreign owners of nationalized enterprises did, under pressure from the IFIs and bilateral donors, eventually receive compensation. Most citizens whose property was confiscated are still awaiting restitution. As we will see below, these issues keep coming up in the policy dialog with multilateral donors and in the lingering complaints of the business community. 3.2. Institutional Reform under the EPRDF The year 1991 was a watershed for reform: a coalition led by two liberation Fronts ended the civil war victoriously. The province of Eritrea became an independent state in 19934. A 4 The terms under which the province of Eritrea (3.5 million people) formally seceded from Ethiopia in 1993 turned out to be of strategic importance for both countries in economic terms. Notable issues such as border demarcation, access to seaports, division of assets and the national debt, and equitable treatment of each others' citizens were either postponed or handled in a manner that was widely perceived as prejudicial to Ethiopia's interests. These inequities are generally believed to be at the root of the 1998-1999 border war between the two countries. According to the Ethio-Eritrean agreements on economic and political cooperation, Assab and Massawa were to be free ports; free movement of goods, capital and labor was to be implemented; and the Ethiopian Birr was to serve as a common currency. However, the two countries had contrasting strategies of development, harmonization of macro policy was slow, and some 150,000 resident Eritrean nationals were accorded equal treatment with Ethiopians without full reciprocity by Eritrea. These loopholes produced an estimated 1 billion birr of debt incurred by the Eritrean monetary authorities to their Ethiopian counterparts, a substantial forex leakage due to reexports by Eritrea of birr-purchased Ethiopian export products and participation of Eritrean agents in forex auctions in Addis, rising charges for port services, and the like (Addis Tribune, 1998d). The arrangements were de facto terminated in 1998 with Ethiopia switching its major port to Djibouti and requiring trade in hard currency upon the introduction of the Eritrean Nakfa in 1997. The immediate consequences of the war have included deportation of each others' citizens on grounds of national security, tens of thousands of casualties on both sides, displacement of half a million people, and diversion of development funds (from investment and debt service funds in the case of Ethiopia) for the acquisition of sophisticated weaponry ­ not to mention an estimated daily expenditure of US$1 million by each to prosecute the war (Economist, 1999; GOE, 1999). 26 transitional government of Ethiopia (TGE) was formed in Addis Ababa by a loosely affiliated assortment of affiliated liberation fronts (known as the Ethiopian Peoples Revolutionary Front-- EPRDF). The Tigray Peoples Liberation Front (TPLF), which serves as the vanguard of the coalition, accounts for most of the senior government leadership. This was a rather unusual conclusion for a civil war. Groups at the periphery in effect gained exclusive control of the state machinery. The country still contends with all the contradictions entailed by such a dispensation for both victor and loser (Bevan, 1994; Chole, 1993). The large national army of 500,000 was disbanded and replaced by EPRDF's own army (Dercon and Ayalew, 1998). The civil service (presently numbering 250,000 permanent members) was trimmed by about a quarter--many of whom were fired for political reasons. The TGE proceeded to decree a number of important institutional changes in 1991-94, prior to the national election. The administrative structure of the country was revamped once again, this time into 9 federal units (Killils) and special administrations in the two richest cities (Addis Ababa and Dire Dawa). The regional states were further divided into 66 zones comprising 556 woredas or units of local government with about 100,000 people each. The 1994 Ethiopian Constitution enshrined an ethnic-based federalism. After the 1995 "administratively fair, yet uncompetitive" elections (World Bank, 1998c), a bicameral parliament dominated by the EPRDF assumed power under the newly created Federal Democratic Republic of Ethiopia (FDRE). The federal constitution is Sovieteque in granting to any ethnic group the right to self-determination `including and up to secession.' While the idea of a decentralized governmental structure finds broad support, the wisdom of gerrymandering districts to create ethnic fiefdoms remains a highly contentious issue in Ethiopian politics. Ethiopia today employs an unusual form of fiscal federalism. The large vertical imbalance between the revenue raising power of the Federal Government (which collects some 80 percent of revenues) and the responsibilities of the regions is bridged by unrestricted block grants. These subsidies are based on a formula that purports to reflect a region's population, development level and local revenue generation. The advantages of responsiveness to local needs which would normally result from decentralization are, therefore, diminished by the lack of performance-based reciprocity, inter-regional gaps in technocratic capacity, and the relative neglect of the special needs of municipalities (World Bank, 1999). 3.3. Policy Reform under the EPRDF A few months after assuming state power, the EPRDF issued its first economic program (GOE, 1991). The programs proposed to deepen the price decontrol and domestic-trade liberalization introduced by the previous government. It advocated continuation of state 27 ownership and control of industry, finance, modern services and land. The leadership's vision of the post-socialist economy, though amended after 1993, remains one of a "state-led market economy" with mixed ownership and active regulation to stem market failures. The macroeconomic and structural reforms the Government has undertaken or plans to undertake are summarized in Table A3.3. The first PFP (1992-93) marked the formal start of an extensive stabilization and structural adjustment program for poverty reduction and transition to a market economy. Key elements of the various rounds of the first-stage reforms included: ˇ prudent fiscal policies (mainly via retrenchment of the army and civil servants, tax reform and lower defense spending) and monetary policies (limiting monetization of the fiscal deficit), ˇ a substantial correction of the overvalued nominal exchange rate, ˇ decontrol of many prices, ˇ liberalization of the foreign trade and forex regime, ˇ SOE autonomy and small and medium enterprise (SME) privatization, ˇ financial market reform, including the licensing of six local private banks, and seven insurance companies so far, and interest-rate (except for deposit rate) liberalization, and ˇ a liberal investment code with an investment authority that is organized as a "one- stop shop." The last two PFPs have paid greater attention to agriculture/rural development, social sectors, infrastructure, energy, export and private sector development, and the financial sector. A civil service reform program is also underway. It encompasses five areas including financial management, judicial reform, human resource management, top management systems, human resource management and service delivery. The Government adopted a three-year adjustment program in mid-1996 under IMF's Enhanced Structural Adjustment Facility (ESAF). In addition to the stabilization program (also supported by a SAC I loan from IDA), it introduced structural reform measures aimed at liberalizing the economy. All retail prices, except petroleum, have been decontrolled. Export taxes were eliminated except on coffee, in the case of which license fees for exporters were reduced. Measures to revive the private sector included allowing domestic private participation in freight forwarding and clearing, and in banking and insurance activities. The parastatal monopoly in coffee marketing was effectively abolished by allowing the private sector into coffee trading, marketing and exports. 28 Maximum import duties were lowered from 230 percent to 40 percent, and to a weighted average of 21.5 percent by 1998. The negative list used to determine eligibility for imports through the foreign exchange auction, has been reduced. Exporters of goods and services are now allowed to retain 10 percent of foreign exchange earnings. The Government also issued a new labor code that incorporates much international practice in the areas of employment conditions and compensation, eliminated the monopoly power of some official marketing and trade corporations, and completed the decontrol of agricultural prices. The investment code was revised four times in as many years and presently allows greater private-sector participation, including in public utilities where the state enjoys de facto monopoly. In addition, a number of modest measures have been taken to promote exports. Ethiopian public enterprises have historically been commercially (if not always economically) profitable, and their privatization does have important fiscal implications. For example, they accounted for 90 percent of federally collected revenues in 1996--a full 40 percent coming from just 10 of the largest SOEs (World Bank, 1998a, pp. 46-47). Nonetheless, the Ethiopian Privatization Agency has recently stated that, during 1993-98, it had undertaken the divestiture of 176 (three-quarters of which are retail shops and hotels) firms at a total price of 2.5 billion birr (US$ 330 million). Financing and the inefficient land-lease policy are the major constraints on the planned case-by-case privatization of 120 large enterprises in the next 3 years Some 16 of the enterprises were sold to foreign capital. The largest investor, the Saudi conglomerate of MIDROC Ethiopia, which accounted for 64 percent of privatization receipts, paid US$172 million for the Lega Dembi Gold Mines and another US$90 million for ten agro- processing enterprises. It also owns the new Sheraton Addis and a network of construction, service, and financial establishments (UNCTAD/ICC, 1999). The Birr has been allowed to float as a result of which the ETB/US$ exchange rate rose from about 2 in mid-1992 to about 8 in mid-1999. This flexibility has resulted in the convergence of the official, auction and parallel market exchange rates. The real exchange rate depreciated significantly as a result of the nominal devaluation and disinflation. The growth rate of the GDP deflator declined from 21 percent to 3 percent between 1991 and 1997. With additional support for improvements on the supply side, these reforms will result in a marked improvement in the competitiveness of Ethiopian exports. A number of tax policy reforms and other government revenue enhancement measures were also introduced. The number of income tax brackets was reduced and the top marginal tax rates lowered from 85 percent to 40 percent. Revenue mobilization, at less than 15 percent of GDP, still remains inadequate. Tax arrears and corruption (notably in Customs, parts of Inland Revenue, and at the regional level) continue to be major problems as is the 1960-vintage commercial code (World Bank, 1998a; 1998c). 29 3.4. Reform Dialog: Government and Stakeholders A striking feature of the landscape of reform-making in Ethiopia is the absence of an institutionalized mechanism for discussing or debating economic policy proposals between the Government and the various business and civic organizations. The dearth of public information and limited public participation in policy deliberation makes Ethiopia an odd man out among such progressive reformers as Uganda and Ghana. The marginalization of important stakeholders (namely, opposition parties, various chambers of commerce and sectoral associations, farmers groups, the independent press, professional associations, trade unions, local NGOs, the academic community, and the like) reflects the autocratic legacies of the country's political systems. Successive governments have prevented, coopted or marginalized nascent civil society institutions. This has had several unsavory consequences: the business community is impelled to resort to corruption; the "watch dog" role of an independent, informed press is weakened; the sense of ownership of the reforms by the public is foregone; effective oversight by Parliament is undermined; and issues of distributional fairness across regions or sectors become overly politicized. Recent research on aid effectiveness and sustainability of reforms points to a number of conclusions. First, equitable growth is predicated on sound macroeconomic fundamentals, good budget management, and institutional reforms that foster the mobilization of underutilized resources. Second, the sustainability of good reforms, in turn, depends on domestic ownership as manifested by the commitments of the leadership and the citizenry to the reforms along with the capacity to implement them. Third, good governance ultimately undergirds the above in the form of an accountable executive, a professional bureaucracy, the rule of law, a stable and transparent policy-making process and an engaged civil society (World Bank, 1998a; Leonardo, et al., 1999). The formal decision-making framework under the current EPRDF government is clear enough: (1) The overall development strategy and policy direction are apparently scrupulously debated within the central of committee of the EPRDF, (2) The Prime Minister's Office is the focal point of coordination for major policy decisions between the legislative branch (EPRDF congresses and the ruling-party controlled Parliament) and the executive branch, and (3) The Ministry of Economic Development and Cooperation (MEDAC) is the main gateway for donors. 30 Line ministries and regional planning bureaus also have varying levels of input. Any loan secured has to be signed off by the Council of Ministers. Donors focusing on program support and sector programs have found this arrangement slow but expedient since it lends itself to efficient technocratic relations with unambiguous lines of authority. However, the Ethiopian governance system, by short-circuiting broad participation, undermines the sustainability of reforms in the long-run. The easy phase of reform (involving rehabilitation, liberalization and stabilization) seems to have enjoyed the tacit support of most sectors of the population, consulted or not. Second-generation reforms involving ownership of land, large-scale state enterprises, Party-owned business groups, and competition policy are, however, too contentious for the viability of a model based on dictat. The angst of the business community and other stakeholders is eloquently expressed by the well-informed, if sometimes overly critical, Economics Correspondent of the respected weekly, Addis Tribune (1998e). The assessment is worth quoting at length since it aptly captures the agenda for future policy debate: It is no wonder that in Ethiopia today several risible economic policies and practices are in effect: Individuals cannot own land; houses and other properties confiscated by the predatory Derg regime continue to be in the possession of the present government; the official policy of ethnic bias prompts leaders to allocate disproportionate amounts of public resources to favoured regions; the mixing of politics with business is flagrantly practised; privatization of nationalized enterprises is going on without any regard to the issue of compensation; back-breaking increases have been made in rents on government-owned (but previously privately-owned) commercial premises; unprecedently high taxes are levied on rental income and capital gains; bona fide franco valuta imports are prohibited; private foreign exchange bureaus are not allowed; foreign participation in the banking and insurance sector is banned; directed credit is commonplace; top- notch university lecturers and experienced civil servants have been dismissed or forced into early retirement; peasants critical of official economic policy and political thinking have been disposed of their plots of land; commercial farming is eyed with suspicion; and peasant agriculture is being saturated with fertilizer to the neglect of irrigation. Similar deficiencies are easily observed in the political arena as well, but there is no need to go into them here as they are quite well- known. Donors, including usually outspoken bilaterals, have chosen to bifurcate economic conditionality and political conditionality. Their complaints appear to have been largely confined to the more programmatic issues of public sector efficiency, transparency, and a blanket appeal for stakeholder participation and respect for basic human rights. A number of donors, for example, suspended program aid in 1999 for fear that fungible resources might be diverted to 31 finance the Ethio-Eritrean war which claimed 20 percent of the 1999 budget (GOE, 1999; Economist, 1999). Given the current outpouring of pan-Ethiopian patriotism, many hope that the EPRDF administration will take advantage of this opportunity to establish a more inclusive governance system.5 3.5. Donor Assistance Strategy Given the importance of multi-donor participation--in terms of both ideas and money-- to the post-Derg reform programs, it would be useful to take a brief look at the goals and strategies of the donor community. I draw on the World Bank's country assistance strategy heavily since it is well documented. The highlights of reform measures pursued by the multilateral institutions and multi- donor conferences are synthesized in Table A3.2. The history of reform in the past three decades and the ODA data reviewed earlier both suggest that, with the exception of the IFIs, donors have on the whole been unable or unwilling (perhaps for fear of punishing the poor) to effect a close linkage between aid and reform. Ethiopia enjoys a good relationship with the community of multilateral and bilateral donors. IDA assistance was devoted in the early phase of the TGE reform program on economic recovery, structural adjustment, and the preparation of policy framework papers (PFPs) as well as technical documents. It then switched emphasis to 5-10-year, primarily externally funded, sector investment programs (SIPs) in 1997. Other multilateral donors active in these programs include the African Development Bank, the European Union and the United Nations agencies. Several bilateral donors also provide active support for these programs. Overall, donor assistance strategy for the utilization of the aid to Ethiopia has been directed at four policy clusters. The first cluster focuses on policy dialog, and economic and sector work designed to emphasize the importance of macroeconomic stability, private sector development and improvements in capacity for implementing projects or policies. A number of studies, including five public expenditure reviews and an equal number of PFPs, have provided a coherent framework for linking project funding with sector and program aid in a forward-looking manner (World Bank, 1997a). A second cluster is devoted to infrastructural development with a special emphasis on expanding the road network and the supply of electricity. The third cluster concentrates on 5 According to the 1999 reports of Amnesty International, Human Rights Watch and the Ethiopian Human Rights Council, over 10,000 political prisoners in Ethiopia, most without formal charge or trial. The second national parliamentary elections are scheduled for the year 2000. 32 growth-enhancing support for agriculture (mainly improved seeds and fertilizer), exports (designed to diversify the export basket and reduce dependence on large flows of aid), and small- scale enterprises. A fourth cluster of reforms is aimed at the alleviation of poverty with the help of ambitious sector investment programs in education and health, and modest projects in the areas of population, gender equality, food security and water supply. SIPs have a number of defining characteristics: they are sector-wide in scope; they provide a coherent framework for sector policy; they put local stakeholders in the driver's seat; they let all donors to sign on; they enjoy common implementation arrangements with a multi-year horizon and transparency; and they require minimal long-term foreign technical assistance (Rose, 1985; Lister, 1998). Presently, there are three SIPs in priority sectors: Transport (Roads subsector), Education and in Health. They are integrated into the government's budget and implemented by government agencies in collaboration with the private sector (especially health and road building).6 Similar initiatives in the areas of energy and food security are in the making. The formal mechanisms for aid coordination in Ethiopia are strong, both at the external and country levels. At the external level, aid coordination is carried out through Consultative Group Meetings, four of which have been held biennially since 1992. The multi-donor special program of assistance for Africa (SPA) meetings, also held biennially since 1992, provide a useful forum for aid coordination and review of country performance. At the country level, MEDAC is the overall coordinator of all external assistance delivery to Ethiopia. Other key participants include the Prime Minister's Office, the Ministry of Finance, the National Bank of Ethiopia and the Council of Ministers. Frequent consultation between local representatives of donors and the Government is now a routine affair. Over fifteen donor countries are active in the country. Their areas of emphasis include relief and rehabilitation, food aid, social sector development, infrastructural development and adjustment support in collaboration with the multilateral development institutions. Germany, Japan, Italy, USA, the Nordic countries, the Netherlands and Canada are the leading bilateral donors (Table 2.4). Some of the less generous bilateral donors include the United Kingdom, France, Switzerland, Austria, Australia, China and S. Korea.7 6 The donors for the transport (road sector) SIP include EU, DfID (UK), GTZ, KfW, NDF, and JICA. The education sector SIP was a cooperative effort of AfDB, Finland, Germany, Ireland, Italy, Sweden, USA, UNESCO, UNICEF, and UNDP. Finally, the health sector SIP was in part funded by AfDB, EU, Finland, Ireland, Italy, Netherlands, Norway, Sweden, Austrian Development Cooperation, UNDP, UNICEF, UNFPA, USA, and WHO. 7 There are "national-flag" patterns here as in other African countries. The U.S., for example, limited its cumulative aid of over US$1 billion in 1984-94 to emergency food assistance and rehabilitation. It has since expanded its menu to include development assistance in the areas of education, health, food security and democratization (USAID, 1999). The German aid program, channeled mainly through GTZ, emphasizes skill transfer, poverty alleviation, food security, resource management and selected social development programs. 33 Many NGOs are also active in Ethiopia although their number is low by the standards of other African countries. According to the Federal Disaster Prevention and Preparedness Commission, there are 310 registered foreign NGOs (one-third of which are foreign) operating in the country. They are active in the areas of relief and rehabilitation as well as in development programs geared toward education, health, water supply, agriculture (most notably the Sasakawa/ Global 2000 Program), and off-farm income generating activities for poor households. The combined annual budget of NGOs in Ethiopia is variously estimated in the range of US$150 million. A new "Code of Conduct for NGOs in Ethiopia" was signed by 165 NGOs in March of 1999. This was preceded by a period of friction with the Government that is accused, among other things, of favoring Party-affiliated NGOs. 3.6. Donor Assessment of Reform Record A World Bank mission (World Bank, 1987, p. 3) to socialist Ethiopia recalls an oft- repeated observation regarding the perplexing gap between the potential and the actual performance of the Ethiopian economy: Ethiopia has some comparative advantages stemming from its size, natural resources, location, and its competent civil service... Ethiopia has generally avoided heavily bank-financed deficit spending and consequent inflation. The country's external debt burden has so far been kept within manageable bounds, and in (favorable) contrast to the experience of many African countries, its rates of capacity utilization in industry have been quite high. Its record of project implementation has been relatively good, and its technocrats and civil servants have a reputation for honesty, dedication and competence. These are indeed important strengths and successes of economic management to build upon. It then goes on to explain this paradox of high potential and low achievement in terms of "initial conditions" of poverty and underdevelopment, vulnerability to numerous internal and external shocks, and the misguided economic policies of the Derg Government. Carefully noting that it had no quarrels with socialism per se, the Mission made a strong plea for policy reform in the name of economic efficiency. The first full-scale single country evaluation of European Community aid was conducted on Ethiopia in 1995 covering the previous twenty years (IDS/IDR, 1996; Maxwell, 1998). It arrives at several conclusions, inter alia: (1) the effectiveness of EU aid was undermined by the unfavorable policy environment and the numerous shocks to the economy; (2) there was a mixed 34 pattern of success (infrastructure, water, energy, Shoa PADEP, sectoral import program for agriculture) and failure (rural water supply, soil conservation, food aid, Stabex); (3) a gradual `learning by doing' by the EU was evident since the agency was inadequately prepared but also contractually constrained to engage in policy dialog especially under Lomé I and Lomé II; and (4) the bias of the Ethiopian governments toward over-centralization and over-standardization was not vigilantly countered. The review recommends a unified Ethiopian program (instead of a multiplicity of instruments and a long chain of decision-making back to Brussels) and greater emphasis on policy dialog with the Government. What one gleans from recent World Bank reports regarding the first-generation or Washington-Consensus reforms under the new Government (i.e., reconstruction, stabilization, liberalization, and some privatization) is a much more upbeat assessment. First, the leadership is described as knowledgeable and committed to serious economic reform.8 Second, policy makers are said to exhibit a strong sense of `ownership' of the reform programs--a point proudly underscored by the Prime Minister himself (Zenawi, 1999). Ownership here means irreversibility of commitment to the reforms, and the capacity to implement them. Third, the economic record, in terms of growth and stability, has so far been impressive. Some concerns are also expressed in these reports. Loan portfolio performance, despite significant improvements over the 1980s, still receive an only "satisfactory" rating. Although the Government has shown a steep learning curve in implementing promised reforms, it continues to confront a number of problems: tax collections (especially income and domestic sales) are narrowly based and backlogged even in relatively developed areas like Addis Ababa, the overregulated government system breeds corruption although not at a grand scale comparable to many other African countries, the civil service is underpaid and initiatives from below undervalued, and the judicial system remains woefully inadequate (World Bank, 1998a; 1998c). The net effect is that absorptive capacity (defined narrowly as the current rate of disbursement) is increasingly becoming a binding constraint on the full utilization of pledged aid. That is not to say that donors are always efficient. A recent assessment of the three SIPs laments the fall in external financing to less than half of what was promised (GOE, 1999). Concerns about the unreliability of ODA disbursements are also reflected in the strong 8 The most recent Country Assistance Strategy (World Bank, 1997, p. 2) points out that, notwithstanding the unfortunate ethnic flavor to politics, the developmentalist nature of the Government is reflected in "an exceptionally strong sense of `ownership' of economic policies and to probity in their implementation. Policymakers are given to very serious consideration of policy options, do not make policy commitments lightly and have a solid track record of delivering on those they do. The Prime Minister [Meles Zenawi] himself is very keenly involved in economic management and has an impressive grasp of economic issues to an extent that the label of an outstanding economic technocrat could be applied to him without much exaggeration." 35 preference expressed by regional governments for predictable federal subsidies over earmarked foreign aid of similar amounts (Personal Interview). The effectiveness of EU aid is also partly attributed to its weakness in policy dialog and the callous neglect by the EU bureaucracy of the importance of good local administrative resources to oversee the implementation process (Maxwell, 1998). Poor food aid targeting has meant that only half of the food insecure households benefit from grain distribution (Clay, Molla and Habteworld, 1998). The leadership is seen to be especially attracted to the East Asian model of development, and the dual-track and gradualist Chinese approach to post-socialist reform. Ethiopia, however, lacks many of the ingredients of East Asian success: a government committed more to nurturing than controlling weak markets, a highly meritocratic and politically insulated technocracy, high rates of saving and investment, a reservoir of off-farm production capability comparable to China's dynamic township and village enterprises, a neighborhood with lots of positive spillovers, and a much less fractious polity (Wade, 1990; World Bank, 1993; Woo, 1998). The Government has now begun to confront, albeit less decisively, the second generation of reforms in the areas of financial deepening, competition, managerial capacity, transfer of technology, diversification of exports, and labor-intensive private SMEs (Stiglitz, 1998). The World Bank (1997a, p. 10) laments the reluctance to go all out for a radical transformation into a market economy in these terms: The careful consideration of policy options and the `ownership' that it reflects are welcome but, in general, the Government tends to be over-cautious in undertaking the second generation of reforms that are needed now. The boldness and decisiveness of the initial phase of reforms has given way to an all-round cautiousness on virtually all fronts in the next phase. Also weaknesses in implementation capacity need to be addressed if the absorption of concessional resources is to rise substantially and the economy is to move to the `frontier' of the balance between stabilization and growth. Some Ethiopian economists have suggested that financial stability has been attained largely through budgetary restraint and significant increases in aid flows (Chole, 1993; 1994; Addis Tribune, 1998c). Achieving sustained and broad-based growth requires greater emphasis on a positive supply response. It is also worth noting here that the relevant standard for judging the efficacy of current policy stance is no longer the dismal record of the Derg regime. Domestically, the relevant frame of reference ought to be the goal of doubling real per capita income every 10-15 years. Externally, the benchmark should be the incentives provided to foreigners relative to those of countries which compete most directly with Ethiopia for investment or export markets. 36 One challenge is to raise the rather dismal domestic saving rate (6%) and the modest investment rate from 18% to at least 25% of GDP. This, in turn, cannot be accomplished with increased tax effort alone. Accelerating reforms in such areas as land use, banking and transparency would improve Ethiopia's competitiveness by providing an unambiguous signal to the overcautious private sector that the Government is committed to an open, rule-based market economy (UNCTAD/ICC, 1999; World Economic Forum, 1998). The upshot is that Ethiopia is at the cross-roads between the easy phase and hard phase of reform. It is by no means clear that the remarkable macroeconomic successes can be easily replicated at the microeconomic level if the reform momentum is not accelerated. IV. Relationship Between Aid and Reform The relationship between aid and reform, especially when it involves institutional transformation, is profoundly political. It involves redistribution of power within society and, therefore, calls for both positive and normative analyses. The comments below are based on the history of the Government's preferred policy posture and the negotiated reform measures actually undertaken, and what I was able to glean from various donor reports or gather from interviews primarily with government officials in March 1999. A more nuanced treatment of the political economy of Ethiopian reform calls for an intimate knowledge of policy debates within the ruling party and the nature of negotiations between the donors and the Government. Unlike the case of other African reformers, the identification of pro-reform and anti-reform policy elites is made more difficult (and perhaps moot) by the opaque nature of the decision-making process under the collective model of party governance that currently prevails in Ethiopia. The Aid and Reform in Africa project, of which this paper is a part, proposes to test three hypotheses: (1) Governments choose to reform, or regress, independent of the aid relationship; (2) Non-financial aid has a better impact than financial aid on the generation of policy reforms; and (3) Financial aid works when policy reforms and institution-building are already underway. The overall conclusion is that the aid-reform linkage in Ethiopia is rather complex: aid appears to have induced or cemented some reforms, but not others; and the optimal sequencing of the forms 37 of assistance (money or ideas) or the efficient timing of aid is often hard to establish in the absence of a clear counterfactual. Selected illustrative examples will be used to elucidate these issues. 4.1. Where Aid Succeeds in Inducing Reform In countries like Ethiopia, aid is likely to induce and lock-in new reform or to broaden existing reform when the foreign exchange constraint is binding and (a) the payoffs to both the government and citizens are demonstrably high, or (b) the payoffs to the rulers are high even though the `masses' get little, or (c) successful reforms are introduced progressively and delicately so that even a reluctant government can conveniently claim ownership. The payoff to the ruling elite may take the form of directing externally-funded projects and off-budget expenditures to favored regions or sectors, shifting fungible resources to politically favored but low return activities, and misappropriation of public funds where corruption is rampant. It also depends on the presence of a crisis situation (thereby generating intense pressure in favor of high- stake reforms) at the time reforms are seriously contemplated (Grindle and Thomas, 1991). In the case of Ethiopia, the willingness to engage in meaningful reform seems to have prevailed since 1988. The Derg approached Western donors with hat in hand once it became clear that Soviet bloc aid was about to end and that the socialist experiment had failed. The treadmill of modest reforms did not find a responsive chord, however.9 Major bilateral donors such as Italy displayed little interest in pushing form reform, and more reform-minded ones like Sweden limited their efforts to projects that skirted policy barriers to reach the poor directly. The IMF and the World Bank were active in the preparation of PFP1 under the Derg, but it was not implemented because the Government balked particularly against the requirement of a substantial devaluation of the Birr (Chole, 1994). 9 The conflict over the Peasant Agricultural Development Extension Project (PADEP) provides a good illustration. Drawing on the experience of the minimum package extension programs of the previous government and the successful training and visit system in India, IDA offered as much as US$400 million in the early 1980s to provide modern inputs and advice in potentially surplus districts. This was conditional on Government action to raise farmgate prices and to remove draconian restrictions on private trade. The Derg refused on account of different priorities and ideological commitments. These terms, however, were in principle accepted by the same government at the end of 1987 and limited funding became available once the crisis of agrarian socialism became undeniable (Clapham, 1988). In 1989, IDA disbursed US$77 for PADEP I. 38 In the post-socialist period, quick-disbursing and highly fungible aid initially followed the restoration of peace and the enactment of policies focused on stabilization and rehabilitation. Subsequent aid financed jointly-designed sector programs, financial reform and improvements in public-sector management. The EPRDF government was quite keen to obtain funds for essential imports, emergency recovery and rehabilitation (Bevan, 1994; World Bank, 1997b; Maxwell, 1998). Despite an initially poor loan portfolio performance, there was mutual eagerness among donors and the Government to push for significant increases in quick-disbursing loans and grants during 1991- 93. While calling for a convincing strategy for marketization, a consortium of donors orchestrated by the World Bank provided US$680 million for the Economic Recovery and Reconstruction Program (ERRP) followed by a US$250 million Structural Adjustment Credit (World Bank, 1997b). The significant evolution in official thinking (or calculation) becomes most evident when one contrasts the EPRDF Government's initial apprehensiveness about a rapid marketization of the economy as expressed in Ethiopia's Economic Policy During the Transition Period (GOE, 1991) with the latest reform measures agreed to in PFP5, Ethiopia: Policy Framework Paper 1998/89-2000/01 (GOE, 1998a). The 1991 policy paper, acknowledging the failures of the Derg regime, committed the TGE to deepening the liberalization initiated earlier by the Derg in the areas of road transport, domestic retail (but not wholesale) trade and rural labor markets. It also put great emphasis on reconstruction of war-torn regions and rehabilitation of infrastructure. Critical decisions concerning external trade and forex system reform, financial reform, privatization of public enterprises and land and civil service reform were postponed until after the 1995 national election. Many of these issues were side-stepped even after the end of the transitional period, as we will see below. The potential leverage at the disposal of the major donors (notably the World Bank, AfDB, EU, Germany, Italy, Sweden and the U.S.) is indicated in part by the level of aid intensity. In 1996, Ethiopia received 5 percent of the Net ODA of some US$17 billion to Sub-Saharan Africa. Among countries of its size, this comes out one of the lowest aid per head (US$16 ) but 39 above the SSA average as a percentage of GNP (16 percent).10 In the 1990s, the aid share in GNP averaged 23 percent. Since total government expenditure is a quarter of GNP, these figures imply that ODA flows were equivalent to 90 percent of budgetary expenditure. Aid currently finances as much as half of the capital budget and one-sixth of federal government revenue (see Table 1.2). During 1992-94, the World Bank made it clear that the coordinated support for recovery and stabilization was being provided with the expectation that it will be followed with an expanded program of assistance tightly leashed to a number of policy "triggers." It is fair to say that many of the structural reforms would not have been considered seriously much less fully implemented without the infusion of substantial policy loans and grants (Abegaz, 1994; Chole, 1994; World Bank, 1997a). The contents of successive PFPs clearly show (see Table A3.3) that a number of policy measures are conspicuous by their recurrence. With donor insistence, the investment code was revised four times in as many years to bring it closer to best practice among African reformers. The forex system was progressively pushed beyond episodic devaluations to include auctions and an inter-bank market. The IFIs are currently pushing for the phasing out of the auction system and full current account convertibility (more on this in 4.2 below). The Government has reluctantly committed itself to the privatization of large-scale SOEs although PFP5 still calls for improved modality that includes transparency in the utilization of the proceeds. The persistent call for rationalizing the decentralized budgetary system, including the introduction of rolling public investment plans, is beginning to be heeded. Finally, the call for a comprehensive food security plan is being taken seriously and might result in a full-fledged sector program. While most of these reforms were initiated by the Government, the prima facie evidence suggests that donor pressure has made a critical difference in transforming many of them into more than piecemeal measures. How far the aid leverage could have been deployed by a "united front" of donors to buy even more ambitious reforms is nonetheless hard to establish. 10 The comparable DAC figures for its peer group are as follows: Ghana ($38/11%), Kenya ($22/7%), Uganda ($25/13%), and Zambia ($40/11%). The top figure is aid per capita; the bottom is aid as % of GNP--all for 1996. For other comparative figures (early 1990-95) which make Ethiopia's aid intensity even lower, see also O'Connell and Soludo (1998, Table 4). 40 4.2. Where Aid Fails to Generate Reform Cases where aid, for various reasons, failed to produce non-marginal reforms are more readily identifiable. Consistent with the experiences of many countries, it is hard to identify cases where conditionalities overcame stiff government resistance (Collier, et al., 1997; Leonardo, et al., 1999). The following illustrative examples pertain to disagreements that involve national security issues, or the regime's power base, or its view of the optimal mix between the public sector and the private sector (as in the case of certain conditionalities of the IMF). The threat of reduction or withdrawal of aid appears on the whole ineffective whenever Ethiopia is in the midst of a major national crisis. When the Derg declared Ethiopian socialism, some multilateral donors (notably the World Bank) and important bilateral donors (USA) withheld much of the already meager assistance to the country thereby losing potential leverage in the process. The U.S. went further and refused delivery of weapons already paid for. Facing secessionist rebellions at home and an invasion by Soviet-armed Somalia, the regime sought and received massive (mostly military) assistance from the Soviet Union itself. In the spirit of the Cold War, the two superpowers ended up switching sides. Many informed Ethiopians blame short-sighted Carter Administration policy for unwittingly aiding in the transformation of an otherwise reformist committee of non-commissioned officers into one controlled by a dictator (Mengistu Haile Mariam) with a grandiose vision of garrison socialism. Similar threats in 1999 by some donors to suspend program aid has been bitterly resisted by the EPRDF government.. Aid does not also seem to buy good reform when the conditionalities are perceived, rightly or wrongly, to undermine the core of the regime's ideology or its power base. Although the World Bank and other major donors understandably regret the lingering mindset of control over facilitation in the bureaucracy, slippage in economic management, and inefficiencies in the land lease system, they have also chosen to sidestep or soft-pedal some of the more fundamental institutional or policy reforms needed to solve them. That is probably because they know that these and other reforms are unpalatable to the Government (World Bank, 1997b). A notable policy issue that is conspicuous by its absence from the economic reform agenda is the question of private ownership of agricultural and urban lands. Ethiopia had a long history of private ownership of land in a multitude of tenure forms (freehold, kinship, village, 41 parish, clan, crown). The common denominator to all these institutional forms is the fact that wealth was based less on ownership title than on entitlements (customary or legal) to the produce from the land. European-type feudal estates or Asian-type widespread landlessness were rare. This explains in large part why the nationalization of land, which transferred use rights to producers, had such a negligible impact on output or productivity. Officials put the case against the granting of freehold tenure in terms of the need for access to land for all in order to ensure economic security for the rural poor (Zenawi, 1999). The Government's longstanding promise to hold a national referendum on land ownership has yet to materialize (GOE, 1991). Its current strategy of agricultural development-led industrialization (ADLI) focuses instead on long-term leases and such short-cuts as supplying fertilizer and improved seeds. The experience of Ethiopia and other countries, however, demonstrates that the economic case against reprivatization is unconvincing (Deininger and Feder, 1998). The linking of access to land to residency, the inability to use land as collateral, and the insecurity that arises from loss of land as a result of periodic redistributions (due to demographic pressure or as political punishment) have stifled agricultural development. Labor mobility has been reduced or frozen. Optimal farm sizes fail to emerge as the efficient farmers who produce much of the marketable surplus cannot increase their farm size in a country where 20 percent of farmers account for 80 percent grain sales. More alarmingly, investment in permanent improvements of the land is discouraged. Finally, off-farm activities which would absorb those who choose to mortgage or sell their subsistence plots (some 60 percent cultivate less than 1 hectare) are stymied. A secure private land title (individual and communal) is, of course, a necessary but not sufficient condition for a green revolution to occur. This is far from the case of a "dry prairie" awaiting the drenching rains of private incentives to bloom. Ownership reforms will have to be complemented with a more efficient marketing system, and enhanced access to modern inputs. Support in the form of public infrastructural investment, credit and extension services remains minimal. The inescapable fact is that the most effective anti-poverty program is one that reinvigorates rural development by harnessing the willingness of Ethiopian farmers to embrace demonstrably profitable innovations (Howard, et al., 1998). 42 In the urban areas, some half a million poorly maintained housing units remain in government hands. The urban land leasehold system, whose inefficiency and corruption are the substance of perennial complaints by the business community, has contributed to a severe shortage of commercial space and housing. Addis Ababa, whose municipal limit boasts plenty of land, continues to suffer an unacceptably high rate of open and disguised unemployment rate-- many with secondary and tertiary education. The binding constraint here is one of finance or institutional infrastructure rather than a shortage of disciplined and easily trainable labor (UNCTAD/ICC, 1999; Middlebrook and Corato, 1999; Addis Tribune, 1988a). A second area of policy contention that has received little attention in the dialog between the Government and the various donors is the concern with the growing number of enterprises which are surreptitiously owned by the ruling party. While the use of political power by big men to accumulate personal wealth and to grease a patrimonial system is all too common throughout Africa, the phenomenon of an extensive network of Party-controlled enterprises is a rather new development in Ethiopia.11 According to several unpublished reports and the private press, whose reliability cannot be fully ascertained, a large number of wholly or partially EPRDF-owned enterprises are being 11 This phenomenon of cadre capitalism, familiar from the early years of the KMT in Taiwan, is also ubiquitous in the Sudan and Eritrea of the 1990s. Although the 1994 Ethiopian Constitutions prohibits political organizations from engaging in profit-making activities, the public is keenly aware of the prominence of these corporations. For example, a recent World Bank mission on corruption observes (World Bank, 1998c, pp. 7): "Private sector and civil society representatives told the mission that political parties belonging the EPRDF, especially the Tigrayan People's Liberation Front (TPLF), have established companies which can include banks and insurance companies. There is a common perception among some in the business community that these companies have preferential treatment over private sector companies with respect to government contracts, government- controlled credit facilities, import and export licenses, and customs clearances. The mission was not able independently to confirm these allegations. Government interlocutors explained that a number of organizations developed during the seventeen years of struggle against the Dergue, acquiring assets and property. Once the war was won, the winning side had to decide what to do with them. Since holding companies are not provided for in Ethiopian law, the EPRDF created foundations under which the party-affiliated companies (and NGOs) are located. The intention of such companies and NGOs is to provide benefits to the survivors of war--wounded, widows, orphans, handicapped, etc. Several of the party-affiliated organizations are headed by party members. These organizations are not allowed to acquire institutions which have been or are being privatised." Senior Government officials insist that the firms do not receive any special favors from the government (Zenawi, 1999). Another argument advanced in their favor is that, being large, these firms compete solely with foreign enterprises. They, therefore, play little role in displacing the domestic private sector (Personal Interview). 43 organized into interlocking financial-and-industrial groups (FIGs). Their assets are estimated in the range of 1-2 billion Birr. They are said to be registered under various individuals, NGOs or regional governments. Most of the TPLF business concerns are reportedly under the supervisory umbrella of the Endowment Fund for the Rehabilitation of Tigray (EFFORT) whose major divisions are headed by senior party members (Aynekulu, 1996a; 1996b; Anonymous, 1999). Some of the largest EPRDF companies or their affiliates include: Agriculture (Hiwot Agricultural Mechanization, Zeleke Agricultural Mechanization, Tesfa Livestock, Rahwa Goat and Sheep Export), Finance and Trade (Wegagen Bank, Africa Insurance, Guna Trading House, Ambassel Trading House, Dinsho Plc, Wondo Trading, Dedebit Credit and Savings Institution), Industry and Mining (Meskerem Investment, Ezana Mining, Sheba Tannery, Almeda Textiles and Garment, Addis Pharmaceuticals, Mesfin Industrial Engineering, Beruh Chemical, Dashen Brewery), Construction (Addis Construction, Sur Construction, Mesebo Building Materials), Transport (Trans-Ethiopia, Blue Nile Transport, Express Transit), and Service (Mega-Net Corporation, Experience Ethiopia Travel).12 The concern about the role of these holdings is multifaceted. First, such firms in effect create two distinct private sectors--one politically connected, and another that is not. The former can exploit political strings (through party cadres in the bureaucracy) to practice unfair competition for government business, and to engage in rent-seeking in the form of tailored state subsidies or protection, and insider privatization. They can also potentially be used to undermine the economic position of the party's political opponents--all without violating the letter of the law. Second, the Government's near monopoly over the banking system lends itself to directing credit to favored enterprises on less than commercial terms. Third, one can only wonder how the existence of these organizational economic assets have shaped the government's attitude toward future reforms especially in banking, the commercial code and the judiciary. If Party and bank organized FIGs are here to stay, damage from rent-seeking and stripping of public assets may be minimized with an introduction of enforceable anti-trust and fair competition legislation along the lines of Russia or S. Korea (Johnson, 1997). 12 Other TPLF-affiliated NGOs that are active investors in the front companies include the Tigray Development Association (TDA) and the Relief Society of Tigray (REST). The rest are owned or co-owned by the other allied parties of the EPRDF coalition: Amhara National Democratic Movement (ANDM), Oromo People Democratic Organization (OPDO), and South Ethiopia Peoples Democratic Movement (SEPDM). 44 A third illustration of the limits of aid as a leverage in the cause of reform comes from the row between the Government and the IMF. The mid-term review of the three-year ESAF was allowed to lapse at the end of 1997 over a number of disagreements with the IMF (Addis Tribune, 1998b). The unfulfilled conditionalities reportedly flagged by the IMF mission included the need to reduce the current surrender rate of export proceeds from 90 percent 50 percent, the need to allow private operators to open forex bureaus, the need to split the near-monopolist Commercial Bank of Ethiopia into three or four functionally-organized banks in order to foster competition, the necessity of accelerating the privatization program along with a full disclosure of the disposition of sales proceeds, and the imperative of opening up the banking and insurance sectors to foreign investors. Judging from the content of PFP5 (see Table A3.3.), the IMF released one semi-annual disbursement a year later in November 1998 in exchange for minor adjustments to the forex surrender system and the permitting of an inter-bank market (IMF, 1998). The ability of the banking system to meet the needs of the growing private sector remains hobbled by the lack of competition in the banking sector (the CBE controls 90 percent of deposits), directives that hobble the newly-established private banks (such as the recently issued directive that raised the minimum initial paid-up capital of from 10 to 75 million birr, and the 15-perecent single- borrower limit directive), and the prohibition of foreign participation in banking (Harvey, 1996; Demeksa, 1999). The Ethiopian experience suggests that aid followed rather than led the initial wave of reforms. Subsequently, a good mix of technical and financial assistance has had the effect of coaxing the government to get closer to the limits allowed by its own ideological or narrow organizational interests. The sequencing of Ethiopian reform is the traditional one of progression from the easy and immediately necessary to the hard. The reform program began with emergency recovery and stabilization and moved on to marketization and modest privatization. While other post-socialist reformers (Tanzania and Mali, for example) have made some headway in the transition between first-generation and second-generation reforms, Ethiopia continues to hesitatingly ponder the difficult jump into a full-fledged market economy . 45 4.3. The Coupling of Money and Ideas Non-financial assistance, in the form of policy dialog, analytical work (country memoranda, specialized sector studies, policy framework papers, public expenditure surveys, public finance management, and statistical work), and the provision of equipment and training can make a critical difference when it is focused on upgrading domestic capability. Ideas concerning the virtuous circle of openness, stability and growth or the importance of a medium- term expenditure framework (MTEF) for rationalizing overall policy are certainly being internalized in Ethiopia. Some Ethiopian policymakers, however, express mixed feelings about some conventional aspects of technical assistance. They are supportive of technical aid where knowledge transfer is programmed, the advice is grounded in Ethiopian conditions, and the country is given some freedom to buy expertise in the free market as needed. However, they tend to look askance especially at "parachuted-in" consultants. These complaints are also familiar in other African countries (Berg, 1993). Technical cooperation can certainly be useful even in non-reforming economies since it is likely to lead to the minimization of damage from distortionary policies. The small non-financial assistance provided during the Derg years (about 15 percent of net ODA) was indeed helpful in upgrading technocratic skills. It was, however, too modest in size (and not tied to the prospect of a big aid inflow) to tip the balance in favor of what amounted to a regime-undermining systemic reform. The environment was such that open criticism by reform-minded civil servants would have been reckless (Griffin, 1992; World Bank, 1987). Anecdotal evidence suggests, however, that various segments of the bureaucracy strove to minimize economic damage by selective and slow implementation of misguided policies. Recent multilateral assistance to Ethiopia has had a high policy content and a good analytical foundation. Technical assistance has recently comprised some 20 percent of ODA-- about two-thirds of the level for developing countries (O'Connell and Soludo, 1998, Table 1). Aside from the expected problems of inexperience and staff turnovers faced by the new administration, it is praised for fulfilling the numerous conditionalities of the structural adjustment program faithfully. Ethiopia is nonetheless far from the policy frontier compared to 46 its competitors, and analytical work and policy dialog are judged to be at the formative stages (World Bank, 1997a, 1997b).13 4.4. Reforming the Donor-Beneficiary Relationship The currently fashionable rhetoric of full partnership notwithstanding, both the donor- recipient (for grants) and donor-client (for soft loans) relationships are fraught with inequality in Africa (van de Walle and Johnson, 1996). Donors typically come armed with money, superior analytical capability (but inferior knowledge of local conditions), and the mentality of a generous but calculating patron. Aid bureaucracies are partly motivated by narrower incentives such as maximizing disbursement rates even at the cost of quality. Anecdotal evidence suggests that the behavior of some officials from bilateral aid agencies betrays an unwillingness to maximize the pass-through of funds while insisting on micromanaging the aid relationship.14 For policymakers in countries like Ethiopia, where a strong sense of confidence and ownership prevails, some policy conditions and excessive documentation requirements are viewed as wasteful diversions of scarce administrative resources. Many express a clear preference for greater multilateralization of aid, guaranteed flow of aid under a negotiated framework, and disbursement in the form of block grants (earmarked or unrestricted budget support known as channel 1 funds) to the extent possible (Maxwell, 1998; Personal Interviews). A negotiated framework, they argue, provides flexibility of implementation while permitting donors to concentrate on the "rules of the game" and on "evaluating outcomes." Evaluation is, however, frustrated by the multiplicity of often conflicting output indicators akin to the "success indicator problem" of Soviet planning. 13 Ethiopia's 1998 overall CPIA score (see Table 1.1) of 3.7 is on par with Ghana (3.6) but behind such reform leaders as Botswana (4.4) and Uganda (4.2). The 1998 Africa Competitiveness Report ranks Ethiopia low in overall competitiveness (17 out of 23), but high for the optimism of the business community (5 out of 20) and for its FDI protection regime (World Economic Forum, 1999). 14 A senior policy maker in a line ministry recently told me of an encounter with a representatives of a bilateral agency who suggested that funds be deposited in the name the Minister. When told that such a practice is not permissible in Ethiopia, the response was that it is indeed common elsewhere in Africa. An ECA director calls this phenomenon "a race to the bottom" as such donors seek out the least demanding recipients to the detriment of aid effectiveness for which only Africans are almost invariably blamed. From the donor perspective, the challenge of managing an aid program in the complex environment of countries like Ethiopia is aptly captured by the metaphor: `hitting a moving target from a moving platform' (Maxwell, 1998, p. 107). 47 Finally, while there are a multiplicity of donor conferences, there is no Africa-wide forum for channeling the collective views or reform proposals of recipient countries. In addition to delinking development aid from narrow foreign-policy or commercial objectives, there are a number of issues to be addressed including recipient input concerning the amount and form of technical assistance and the uneasy roles of mushrooming NGOs. V. Concluding Thoughts Post-socialist economic reform in Ethiopia has so far been undertaken in two stages: rehabilitation, stabilization and basic liberalization during 1990-95; and infrastructure-oriented SIPs, financial liberalization and improvements in public sector management during 1996-98. Aid, flowing at an annual average of US$1.3 billion (in constant dollars) followed the first phase of home-grown reforms as an ex post reward for good behavior. Well-coordinated aid in the form of persuasion (policy dialog, analytical work) and money also played an important role in jointly forging the second phase of reform this time as much for persuasion as for rewarding. Despite the significant accomplishments of the past six years, there is a substantial opportunity to further revamp the still regimented economy. These include restoration of private ownership of most land, banking reform (breaking up CBE and autonomy for NBE), competition policy (large-scale SOEs and Party-owned businesses), and reform of the governance system (both the political and bureaucratic aspects). Many of these reforms have frontloaded costs and backloaded benefits which render them unappealing to myopic policy makers. They may also entail concentrated opportunity costs to the ruling party but diffused benefits to the electorate. Multinational agencies often pose the basic question of reform in Africa in terms of rescuing nascent markets from a bundle of contradictions that is the post-colonial state: "authoritarian and yet fragile; highly centralised yet dispersed; overdeveloped in size yet underdeveloped in function; trumpeting national unity yet controlled by narrow personal cliques; nationalistically resistant to outside political pressures but weakly subordinate to economic ones" (Africa Confidential, 1999, p. 5). While the Ethiopian state is old and resilient, it does share the African pattern of a center that is too weak to centrally plan a socialist economy or to "govern" 48 the market effectively. It is nevertheless strong enough for capricious interventions which, among other things, raise transaction costs substantially and stimulate growth-retarding informality. The Achilles heel of Ethiopian development (and reform) is its inability to nurture an equitable political system that secures lasting peace at home and stability in the Horn of Africa. In March 1978, Ethiopia won a bloody border war with Somalia. March 1990 inaugurated the country's return to a market economy in the midst of a raging civil war between an alliance of two liberation fronts and the Derg Government. A decade later, March 1999 witnessed the reigniting of an even bloodier war between the two erstwhile allies. The enduring legacy of over US$6 billion worth of Soviet bloc and Western arms supplies to the impoverished Horn of Africa region is one of chronic violence and mutual destabilization. The region has witnessed two of Africa's most mechanized wars in just two decades. The resumption of a war economy is once again accompanied by a major famine in the northern provinces affecting 5-6 million people. The economic distress is compounded by an intensifying AIDS pandemic. External destabilization and incessant domestic conflicts produce myopia born of a heightened sense of insecurity, destruction of capital (physical and social) and information, militarization of life, and mutual distrust. The two contradictory faces of external aid, one destructive and the other constructive, have yet to be explicitly coupled in the assessment of aid to Ethiopia. It is no wonder, then, that Ethiopia presents a classic case of an economy of great potential struggling to overcome a "low-level equilibrium trap." The hopeful signs of periodic spurts of growth, driven mainly by positive shocks (favorable weather, or higher coffee prices and aid flows) prove fleeting time and again. One manifestation of this is the desire of both the educated and capital to emigrate. Another is the strong preference by risk-averse capital for myopic investments that produce quick profits. Such an environment is invariably inimical to long-term investment and growth (Collier and Gunning, 1999; Bevan, 1994; Chole, 1993). All is not lost, however, and talk of aid triage (a severe form of selectivity) by some donors is rather misplaced. The past six years have shown that, with peace and modest economic reforms, the economy is capable of expanding at more than twice the rate of population growth. This growth record has been facilitated by a remarkable degree of macroeconomic stability. 49 One lesson from successful open economies is that, while sustained stability is necessary for investment, removing the remaining barriers of growth on the supply side is just as essential. The domestic capability that is being built up in Ethiopia with the help of external assistance will have to be aimed at exploiting the untapped opportunities at home and abroad (especially the underutilized Middle-Eastern, Lomé and GSP market outlets). The bridge between domestic capability and opportunity is supplied by sufficiently deep economic reform. Successful reform is, in turn, undergirded by a strong commitment to good governance without which the long battered confidence of citizens and the growing Diaspora would remain tentative. In the final analysis, political reform is a domestic affair. As the disappointing response to the highly liberal investment code amply demonstrates, a country that treats its citizens fairly is a country that is also attractive to foreign capital. If the Government manages to undertake the remaining reforms along with building up its implementational capacity, Ethiopia stands to benefit doubly from an increasingly selective international aid regime and a reinvigorated private sector. Donors can aid this process with a better calibration of aid and reform. 666 50 VI. References Abegaz, Berhanu, 1994, "Ethiopian Economic Reform," in Berhanu Abegaz (ed.), Essays on Ethiopian Economic Development, Aldershot: Avebury. Addis Tribune, (by Our Economic Correspondent/Commentator): 1998a, "Yet Another 5-Year Action Plan--This Time Around for Addis Ababa," February 20. 1998b, "Latest Talks Between Ethiopia and IMF Fail to Lift ESAF Suspension," March 13. 1998c, "Foreign Aid is Propping up the Ethiopian Economy," April 10. 1998d, "Unfair Economic Advantages Eritrea Enjoyed from its Hegemonic Relationship with Ethiopia," July 31. 1998e, "Ethiopia's Alarming Brain Drain," December 12. Africa Confidential, 1999, "Africa Scrambles for Africa," 40(1), pp. 2-5. African Development Bank, 1996, Ethiopia: 1996-98 Country Strategy Paper, Abidjan. Anonymous, 1999, Ethiopian Non-Governmental Business: Companies Controlled by or Associated with EPRDF-Member Organizations, processed, 87pp. Aynekulu, Awualom: 1996a, "The Emerging Monopolies of the TPLF," Ethiopian Register, July, pp. 20-32. 1996b, "The Emerging Monopolies of the TPLF, Part II," Ethiopian Register, August, pp. 14-22. Berg, Elliot, 1993, Rethinking Technical Assistance and Capacity Building in Africa, New York: UNDP. Bevan, David. 1994, "Economic Aspects of Ethiopian Transition to Peace," in Azam, Jean-Paul, et al., Some Economic Consequences of the Transition from Civil War To Peace, Policy Research Working Paper 1392, World Bank. Cassen, Robert and Associates, 1994, Does Aid Work? Oxford: Clarendon Press. Chang, Charles, Eduardo Fernandez-Arias and Luis Serven, 1999, Measuring Aid Flows: A New Approach, Policy Research Working Paper 2050, World Bank. Chole, Eshetu, 1993, "The Dismal Economy: Current Issues of Economic Reform and Development in Ethiopia," Ethiopian Journal of Economics, 2(1), pp. 37-72. 51 Chole, Eshetu, 1994, "A Preliminary Appraisal of Ethiopia's Economic Reforms, 1991-93," in Harold Marcus (ed.), New Trends in Ethiopian Studies, Trenton, N.J.: The Red Sea Press, vol. 2, pp. 311-33. Clapham, Christopher, 1990, Transformation and Continuity in Revolutionary Ethiopia, Cambridge: Cambridge University Press. Clay, Daniel, Daniel Molla and Debebe Habtewold, 1998, Food Aid Targeting in Ethiopia," Grain Market Research Program Working Paper 12, Addis Ababa. Collier, Paul and Jan Willem Gunning, 1999, "Explaining African Economic Performance," Journal of Economic Literature, 37 (March), pp. 64-111. Collier, Paul, et al., 1997, "Redesigning Conditionality," World Development, vol. 25, No. 99, pp. 1399-1407. Deininger, Klaus and Gershon Feder, 1998, "Land Institutions and Land Markets," Working Paper Series No. 2014, World Bank. Demeksa, Bulcha, 1999, "Revising the Banking Sector," Addis Tribune, July 30. Dercon, Stefan and Daniel Ayalew, 1998, "Where Have All the Soldiers Gone: Demobilization and Reintegration in Ethiopia," World Development, 26 (9), pp. 1661-75. Economist, The, 1999, "Africa's Forgotten War," May 15. Etherington, Dan and Alasebu Yainshet, 1988, `The Impact of Income Terms of Trade for Coffee on Capital Goods Imports and Investment in Ethiopia," Eastern Africa Economic Review, 4 (1), pp. 48-52. Ghani, Ejaz and Hyoungsoo Zang, 1995, "Is Ethiopia's Debt Sustainable?" Policy Research Working Paper 1525, World Bank, October. Government of Ethiopia (GOE): 1991, Economic Policy of Ethiopia during the Transitional Period, Addis Ababa. 1998a, 1996; 1994; 1993; 1992, Ethiopia: Policy Framework Paper, Addis Ababa. 1998b, The 1994 Population and Housing Census of Ethiopia, Addis Ababa. 1999, Aide Memoire: Ethiopia Public Expenditure Review, Draft Report of the 1999 PER Mission, May. Griffin, Keith (ed.), 1992, The Economy of Ethiopia, London: Macmillan. Grindle, Merilee and John Thomas, 1991, Public Choices and Policy Change: The Political Economy of Reform in Developing Countries, Baltimore: Johns Hopkins University. 52 Haile Mariam, Mengistu, 1990, Report by Mengistu Haile Mariam: Resolutions Adopted by the 11th Plenum of WPE, Addis Ababa: WPE. Harvey, Charles, 1996, "Banking Reform in Ethiopia," IDS Working Paper 37, Sussex. Howard, Julie, et al., 1998, "Can the Momentum be Sustained? An Economic Analysis of the MOA/SG2000's Experiment with Improved Cereals Technology in Ethiopia," Addis Ababa, September. IDS and IDR, 1996, An Evaluation of Development Cooperation Between the European Union and Ethiopia, 1976-1994, vols. 1-7, June. International Monetary Fund, 1988, "IMF Approves Second Annual ESAF Loan for Ethiopia," IMF Press Release No. 98/51, October. Johnson, Juliet, 1997, "Russia's Emerging Financial-Industrial Groups," Post-Soviet Affairs, 13 (4), pp. 333-65. Lensink, Robert and Oliver Morrissey, 1999, "Aid Instability as a Measure of Uncertainty and the Positive Impact of Aid on Growth," Processed. Leonardo, Jose, et al, 1999, "Towards a More Effective Conditionality: An Operational Framework," World Development, 27 (2), pp. 285-99. Lister, Stephen, 1998, Implementing Sector Development Programs in Ethiopia, Processed, December. Maxwell, Simon, 1998, "Effectiveness of European Aid to Ethiopia," in Marjorie Lister (ed), European Union Development Policy, New York: St. Martin's Press. Middlebrook, P. and F. Corzato, 1999, "Displaced, Disempowered and Distressed: A Focus on Poverty in Addis Ababa," World Bank, February. O'Connell, Stephen and Charles. Soludo, 1998, "Aid Intensity in Africa," Working Paper Series 3-99, Centre for African Economies, Oxford University. Ramachandran, Mahadevan, 1997, Food Security: Economics of Famine, Food Aid and Market Integration in Ethiopia, Ann Arbor: University Microfilms International. Rose, Tore, 1985, "Aid Modalities: Sector Aid as an Instrument in Sub-Saharan Africa," in T. Rose (ed.), Crisis and Recovery in Sub-Saharan Africa, Paris: OECD. Stiglitz, Joseph, 1998, "More Instruments and Broader Goals: Moving Toward the Post- Washington Consensus," The 1998 WIDER Annual Lecture, Helsinki: Finland. 53 UNCTAD/ICC, 1999, An Investment Guide to Ethiopia: Opportunities and Conditions, New York: United Nations, June (advance copy). USAID, 1999, Ethiopia: Congressional Presentation, Washington, D.C. van de Walle, Nicholas and Timothy Johnson, 1996, Improving Aid to Africa, Policy Essay No. 21, Washington, D. C.: Overseas Development Council. Wade, Robert, 1990, Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization, Princeton: Princeton University Press. Woo, Wing Thye, 1998, "Why China Grew," Peter Boone, et al. (eds), Emerging from Communism: Lessons from Russia, China, and Eastern Europe, Cambridge: MIT Press, pp. 153-82. World Bank (Washington, D.C.): 1987, Ethiopia: Recent Economic Developments and Prospects for Recovery and Growth. 1993, The East Asian Miracle: Economic Growth and Public Policy, New York: OUP. 1994, Adjustment in Africa: Reforms, Results, and the Road Ahead. 1997a, 1995; 1992, Ethiopia: Country Assistance Strategy. 1997b, Ethiopia: Implementation Report on Structural Adjustment Credit (SAC I). 1998a, Ethiopia: Review of Public Finances, vols. 1-2. 1998b, Assessing Aid: What Works, What Doesn't, and Why, Oxford: OUP. 1998c, Ethiopia: Anti-Corruption Report, October. 1999, Ethiopia: Regionalization Study. World Economic Forum, 1998, Africa Competitiveness Report 1998, Geneva. Zenawi, Meles (P.M.), 1998, "Interview," The Financial Times, March 2. 54 Table A3.1 Ethiopia: Chronology of Recent Economic Reforms, 1974-98 ˇ Workers Party of ˇ Stabilization ˇ Trade Liberalization Ethiopia (WPE) ˇ ER Liberalization ˇ Further ER Liberalization ˇ PDRE Declared ˇ Decontrol of Prices ˇ Privatization: medium The February 1974 ˇ Resettlement Project ˇ New Investment Code ˇ Ethio-Eritrean War Ethiopian Revolution ˇ Ten-year Perspective ˇ Birr Currency Area & ˇ Civil Service Reform Economic Plan, 1985-94 FTA with Eritrea 1974-78 1979-83 1985-89 1990-91 1992-94 1995-97 1998-99 ˇ Derg Nationalization: ˇ Inheritable Land Use ˇ Tariff Reform - All Land ˇ Producer Cooperatives ˇ Price Decontrols ˇ Privatization: small - Extra Houses ˇ Trade Monopolies ˇ Decollectivization of PCs ˇ Fiscal Federalism - Enterprises ˇ Industrial SOEs ˇ SOE Autonomy ˇ Peasant Associations ˇ Central Planning ˇ Inter-regional Trade Freed ˇ Forex Auctions ˇ Urban Kebelles ˇ Zemetcha Campaigns ˇ Demobilization of military ˇ EPLF in Asmara, 1991 ˇ FDRE Declared ˇ EPRDF in Addis Ababa CBE Commercial Bank of Ethiopia CONTENTIOUS AREAS OF SOE State-Owned Enterprises FUTURE REFORM: PC Producer Cooperative ˇ Reprivatization of Nationalized EPLF Eritrean People Liberation Front Land, Houses and Firms TPLF Tigray People Liberation Front EPRDF Ethiopian Peoples Revolutionary Democratic Front ˇ Privatization: LS SOEs & CBE ˇ EPRDF-owned Enterprises 1974-91 Derg Government (Peoples Democratic Republic of Ethiopia) ˇ Federal Killils: Form and Role 1991-- EPRDF Government (Federal Democratic Republic of Ethiopia) ˇ Ethio-Eritrean Economic Relations ˇ Governance: Level Playing Field 55 Table A3.2 Ethiopia: Highlights of Donor Assistance Strategy, 1980-2001 Policy & Capacity Adjustment Physical Poverty & Agriculture Development & Growth Infrastructure Human Svcs & Environ. Tech Assistance: Coffee Processing Energy Education V-VII Irrigation 1980-91 planning agency IMF Standby (1981-82) Transport/Telecom Family Health PADEP (Derg) CEM Towns/Urban CG1-2 (1992, 1994) Economic Recovery and 1992-94 SPA Reconstruction Program Road Rehab Food Security Improved Seeds PFP1-PFP3 SAC I & SAF Calub Gas Fertilizer (TGE) PER1 Priv. Sector Development Gilgel Gibe CG3-4 (1996,98) Power Distribut. Family Planning Sale of State Retail 1995-2001 PFP4-PFP5 ESAF (1996-99) Water Supply Social Rehab Support Services SIP (or SDP) Export Promotion Road SIP Education SIP Research & Training (EPRDF) MTEF/PIP SME development Energy Health SIP Food Security PER2-PER6 Rail Rehabilitation Women's Rights Conservation Sources: World Bank, Ethiopia: Country Assistance Strategy, 1992, 1995, 1997; USAID (1999); Maxwell (1998); IMF (1998). CG Consultative Group (multi-donor conferences) SIP Sector Investment Program SAC Structural Adjustment Credit ESAF Extended Structural Adjustment Facility SME Small and Medium-scale Enterprises CEM Country Economic Memorandum PFP Policy Framework Paper PER Public Expenditure Review MTEF Medium-term Economic Framework PIP Public Investment Plan 56 .Table A3.3 Ethiopia: Highlights of the Policy Reform Matrices of PFP1-PFP5, 1992-2001 Policy Area PFP1 (1992-93) PFP2-3 (1994-95) PFP4 (1996-97) PFP5 (1998-2001) I. Macroeconomic Policies A. Fiscal and Public Sector 1. Rationalize regional 1. Restructure IRA 1. Strengthen FIRA 1. Decentralization of adm. budgetary system 2. Broaden tax base 2. ditto 2. Consolidate Federal Budget 2. Tariff reform: customs, 3. Agricultural income tax 3. Recurr. cost wrt SIPs 3. Tax reform sales and business income 4. Raise share of PI 4. Ditto 4. Reduce security and 3. Less for security; more for 5. 3-year rolling PIPs 5. Introduce VAT subsidy expenditures econ. devel. and M & O 6. Integrate extra-budget 5. Introduce PIP 4. Initiate PER and PIP & counterpart funds 6. Enhance transparency 5. Eliminate monetizing BD 7. Pension reform 7. Reduce fiscal deficit 6. Centralized budgeting of grants & counterpart funds B. Monetary and Exchange Rate 1. Restrain domestic credit 1. Indirect policy instruments 1. Weekly forex auctions 1. Interbank money mkt. 2. Positive real interest rates 2. Market-determined interest rates 2. Improve T-bill auctions 2. TB secondary market 3. Eliminate ETB overvaluation 3. Policy coordination w/ Eritrea 3. Redemption of Birr 3. Phase out forex auction 4. Retention of 10% of forex 4. Interest-rate liberaliz. C. Financial Sector 1. Autonomy of central bank (NBE) 1. Ditto 1. Enhance NBE's capacity 1. Establish bond market 2. Bank supervision/capitalization 2. Ditto 2. Recapitalize CBE 2. Point of sale for CBB 3. Entry of private fin. institutions 3. Ditto 3. Revamp CBE auditing D. External Sector 1. Reduce discrim. ER rationing 1. Ditto 1. Reduce max duty to 50% 1. Full CA convertibility 2. Eliminate export/import 2. Import duty drawback system 2. Improve debt management 2. Reduce tariffs further subsidies & surcharges 3. Remove impediments: SAC I in MOF 3. Debt sustainability 3. Eliminate all export taxes, 4. Narrow negative list 4. Active export promo except on coffee 5. Promote tourism 4. Revise tariff structure 6. Liberalization of coffee sector 5. Debt relief, monitoring and 7. Relief from Paris Club creditors reduction 8. Capability for aid absorption (Table A.3.3: 1 of 3) 57 Policy Area PFP1 (1992-93) PFP2-3 (1994-95) PFP4 (1996-97) PFP5 (1998-2001) II. Structural and Institutional Policies A. Liberalization and Private-Sector Development 1. Eliminate legal discrimination 1. Simplify licensing and 1. Deregulate fertilizer & 1. Admit Foreigners in against private sector land acquisition pharmaceut. prices power sector 2. One-window FDI approval 2. Ditto 2. Revise investment code 2. Remove investment 3. Decontrol/rationalize prices 3. Decontrol all retail prices limits 4. Restructure AMC and EDDC 4. Sell AISCO retail outlets 5. New Labor Code 5. Other SOE prices to 6. No automatic Govt. hiring reflect import parity B. Privatization and Public Enterprises 1. Action plan for PE divestiture, 1. Retail shops and 29 hotels 1. Improve privatiz. modality 1. Privatiz. Action Plan liquidation or leasing 2. Public Ent. Proclamation 2. Land auction & compensation 2. Sale: 92 of 124 SOEs 2. Restructure and settle arrears 3. Harden budget constraint 3. ditto C. Agricultural Development 1. Decontrol cash-crop prices 1. Greater access to 1. Sell all rural retail outlets 1. Reform land lease 2. Access to credit and inputs to fert. and seeds 2. Microcredit and crafts 3. Viability analysis of state farms 3. Road sector program 4. Forestry sector action plan 2. Research and roads D. Transp-Comm-Utilities 1. Revise mining code 1. Commercialize ETA 1. National Energy Strat. 2. National energy policy 2. Ditto 2. Revisit rail system 3. Deregulate road-railway tariffs 4. Maintenance of roads 3. Rural roads strategy 5. Privatize Govt houses and 4. Ditto simplify zoning/permits 5. Plan for SIP: roads E. Capacity Building 1. Civil service review 1. Ditto 1. Better budget control 2. Raise policy formulation & 2. Coordin. macropolicy 2. Portfolio performance implementation capacity 3. Strengthen MOPED review for loans (Table A3.3: 2 of 3) 58 Policy Area PFP1 (1992-93) PFP2-3 (1994-95) PFP4 (1996-97) PFP5 (1998-2001) III. Poverty Alleviation and Social Sectors A. Population and Health 1. Rural basic health services 1. Primary health care 1. Ditto 1. By 2002: 2. Access to basic drugs 2. Equality of women 2. Cost recovery * 55% primary health. 3. Family planning and MCH 3. New population policy * 80% immunization B. Education 1. Access to primary schools 1. Reconstruct infrastrucure 1. Private participation: 1. By 2002: 2. New tertiary curriculum 2. MT human-resource policy * Education * 50% primary enroll. * Textbooks C. Safety Net and Poverty Reduction 1. National Disaster 1. Policy focus on: 1. Food Security Plan 1. Produce poverty profile Preparedness Strategy a. production subsidy 2. Food Security Plan 2. National Food and Nutrition b. consumption subsidy Strategy c. public works 3. Labor-intensive public works 2. Severance pay for retrenched PSEs IV. Statistics >>>>> Improve collection and reporting of national accounts, revenue, expenditure, BOP, monetary, and debt statistics. <<<<< Sources: Culled from Government of Ethiopia, Ethiopia: Policy Framework Paper, 1992, 1994, 1995, 1996, 1998a. FIRA Federal Inland Revenue Authority CBB Construction and Business Bank M & O Maintenance and Operations PER Public Expenditure Review PSE Public Sector Employees PIP Public Investment Program ETC Ethiopian Telecommunications Corporation NBE National Bank of Ethiopia SIP Sector Investment Program PIP Public Investment Program ER Exchange Rate AMC Agricultural Marketing Corp. EDDC Ethiopian Domestic Distribution Corporation AISCO Agricultural Inputs Supplies Corporation SOE State-owned Enterprise MOPED Ministry of Planning and Economic Development MOF Ministry of Finance MCH Maternal and Child Health 59