Ethiopia
Recent Economic Developments
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar
Close

This paper describes economic developments in Ethiopia during 1990–94. In 1992/93, there was a significant recovery in the economy, which was reflected in virtually all sectors, with real GDP rebounding by an exceptional 12.3 percent. Although the improved security situation in the country and, subsequently, the emergence of a favorable economic and political climate set the stage for the recovery, a number of additional factors contributed to this outcome. In the agricultural sector, good weather conditions, coupled with increases in land under cultivation and greater fertilizer usage, led to a bumper crop.

Abstract

This paper describes economic developments in Ethiopia during 1990–94. In 1992/93, there was a significant recovery in the economy, which was reflected in virtually all sectors, with real GDP rebounding by an exceptional 12.3 percent. Although the improved security situation in the country and, subsequently, the emergence of a favorable economic and political climate set the stage for the recovery, a number of additional factors contributed to this outcome. In the agricultural sector, good weather conditions, coupled with increases in land under cultivation and greater fertilizer usage, led to a bumper crop.

I. Real Sector

1. Aggregate output and expenditure

Following a period of modest and fluctuating growth in the 1980s (averaging 2.5 percent), events surrounding the collapse of the Derg regime in 1991 led to a substantial worsening of economic conditions in the Ethiopian economy, as reflected in a cumulative decline in economic activity of about 10 percent during 1990/91-1991/92 (Appendix Table I). The adverse impact of overall instability and security-related problems during this period was further magnified by variable weather conditions, depressed coffee prices, and foreign exchange scarcity, leading to a continuous deterioration in capacity utilization in the manufacturing sector.

Agricultural production, reflecting its significant sensitivity to favorable weather conditions, exhibited a relatively strong, albeit erratic, performance during 1989/90-1991/92--an average growth of 4 percent during 1989/90-1990/91, followed by a sharp drop in 1991/92 (2.6 percent). The latter also reflected a significant decline in the total area under cultivation as a result of the escalation of the civil war. In contrast, economic activity in the manufacturing and construction sectors declined by an annual average of 17.1 percent and 13.1 percent, respectively, during 1989/90-1991/92, mainly owing to a severe shortage of raw materials and foreign exchange. Output from distribution and other services also contracted by an average of 7 percent and 1 percent per annum, respectively, reflecting the overall decline in economic activity.

In 1992/93, there was a significant recovery in the economy, which was reflected in virtually all sectors, with real GDP rebounding by an exceptional 12.3 percent. While the improved security situation in the country and, subsequently, the emergence of a favorable economic and political climate set the stage for the recovery, a number of additional factors contributed to this outcome. In the agricultural sector, good weather conditions, coupled with increases in land under cultivation and greater fertilizer usage, led to a bumper crop, with the sector registering a strong 7 percent growth rate. In the manufacturing sector, a smooth flow of inputs through increased availability of foreign exchange led to a dramatic recovery in the capacity utilization rate, mainly in the public enterprise sector. Growth in manufacturing rebounded by an extraordinary 51.9 percent, albeit from an extremely low base. Small-scale private sector activity also began to develop, reflecting a modest revival from its historically repressed condition.

Developments during 1993/94 have been less favorable, even though the economy is estimated to have grown modestly (1.3 percent). An erratic distribution and pattern of rainfall throughout the country and inadequate usage of fertilizers and pesticides during the main rainy season are estimated to have contributed to a significant drop in agricultural output. This shortfall has been further aggravated by an unusually dry short rainy season, leading to severe drought in some northern regions, as well as in the southern lowlands. In addition, the pest problems that emerged during the 1994 short harvest in these regions, as well as in the western parts of the country, are estimated to have contributed to the sharp drop in harvested output. As a consequence, agricultural output is estimated to have declined by almost 6 percent. However, growth in the non-agricultural sector (9.2 percent), most notably in the manufacturing sector (13.2 percent), continued to be robust, thereby helping the economy maintain a positive growth for the year as a whole. This was mainly attributable to the continued recovery in the public enterprise sector, as well as in private sector activity, and to an improved investment rate.

The overall investment rate in the economy, after having stagnated at about 9 percent for four consecutive years during 1988/89-1991/92, improved considerably during 1992/93-1993/94, averaging 13.3 percent of GDP (Appendix Table III). Although still low even in comparison with the average for sub-Saharan Africa (17.3 percent in 1993), this represented almost a doubling of investment as a percent of GDP, compared with the 1990/91 level. In 1992/93, about half of the 12.5 percent gross capital formation rate derived from the private sector. In 1993/94, the investment rate further increased, to 14.1 percent, mainly reflecting a higher rate of capital budget implementation by the public sector. Domestic savings, while still weak, given Ethiopia’s extremely low per capita income level, also improved modestly during 1992/93-1993/94, after having dropped sharply to zero in 1990/91. This was attributable to a slight improvement in public sector savings, which increased from minus 1.7 percent of GDP in 1990/91 to 1 percent in 1993/94, as well as a revival in private sector savings (from 1.7 percent of GDP in 1990/91 to 4.6 percent in 1993/94). Although total consumption declined slightly during this period, the expansion in investment continued to be reflected in a large resource gap, which increased from an average of 6 percent of GDP percent during 1989/90-1991/92 to 8.8 percent during 1992/93-1993/94.

2. Sectoral developments

a. Agriculture

The agricultural sector--including principally agriculture and livestock, but, to a marginal degree, the forestry, fishing, and hunting sub-sectors--is by far the most important sector in the Ethiopian economy, constituting more than 50 percent of GDP (1993/94), and employing more than 80 percent of the labor force (Appendix Table II).

(1) Crops

Crop production is predominantly concentrated in staple crops, which are produced by smallholder, subsistence farmers. Irrigation is limited, such that production crucially depends on the amount and distribution of rainfall. Although the average level of rainfall varies considerably across the different regions, broadly, two harvests occur during any given year. The main harvest season (the meher crop) follows the long rains of July-September and the second harvest (the belg crop) follows the short rains of February-April. In some regions, mostly in the central highlands, the above-mentioned two rainy seasons tend to merge and the harvest generally occurs in October and January. More than 90 percent of total crop production takes place during the meher season, while the short season accounts for the remainder. However, some areas (e.g., Bale and Omo in the South, and Wello and Tigray in the North) depend crucially on the belg rains.

In a good year, cereal production, which mainly includes maize (28 percent of the total), teff (24 percent), barley (16 percent), wheat (13 percent), and sorghum (13 percent), is estimated at around 6 million tons. 1/ Of the total cultivated area, teff and maize cover more than 50 percent, followed by barley (17 percent) and wheat and sorghum (13 percent each). Maize, sorghum, and wheat have relatively high yields, and are normally planted during the long season; nevertheless, their yields, on average, at about 12 quintals per hectare, are considered extremely low even by African standards, mainly owing to rudimentary farming techniques as well as limited usage of fertilizers and improved seeds. It is estimated that about 15 percent of total crop production is marketable while the remainder is used for self-consumption or animal feed. In general, switching across crops (particularly toward cash crops) is uncommon in Ethiopia, because of the high priority that the farmers give to ensuring subsistence through the planting of staple crops and the lack of knowledge of the appropriate technology. 2/

Ethiopia’s principal surplus areas in the production of cereals are concentrated in the Oromiya (e.g., Arsi, and East and West Showa) and Amhara (e.g., East and West Gojam) regions, while many densely populated areas in the north (e.g., Tigray, Wello, and Gondar) as well as the pastoral south are known as structural deficit areas. The incomes of the smallholder farmers are particularly susceptible to drought, due to their limited ability to compensate for the impact of crop failure (e.g. lack of alternative employment opportunities, limited asset base, isolation from major markets, and low level of farm technology). The majority of the 4 million people in Ethiopia who are estimated to depend on food aid (even under normal circumstances) live in these regions. 3/

During 1992/93, cereal production is estimated to have grown by about 9 percent, with the highest growth achieved from the relatively high-yielding maize (20 percent) and sorghum (10 percent), followed by barley (4 percent), wheat (4 percent), and teff (3 percent). In 1993/94, reflecting vagaries of the meher season as well as drought during the short rainy season, crop production is estimated to have declined by 10 percent. 1/

Ethiopia’s main cash crops include coffee, oil seeds, and pulses. 2/ Coffee, the most important cash crop and Ethiopia’s principal export (see Section III.2a for further discussion), is mostly grown in Oromiya and the Southern Ethiopia Region, mainly by smallholder farmers (96 percent). Coffee is prepared and processed for the market in two different forms--sun-dried and washed. 3/ Exports are largely concentrated in sun-dried coffee (djimma, wollega, hararge, and sidamo), while exports of the higher-quality washed coffee are constrained by a lack of infrastructure (viz., washing stations, access roads, vehicles, training, etc.). Washed coffee is processed predominantly by service cooperatives (and a limited number of state farms), since the private sector has only recently been permitted to engage in coffee processing. There are about 130 washing stations in Ethiopia, 2 of which are owned by the private sector.

As part of ongoing reforms in the coffee sector, the Ethiopian Coffee Marketing Corporation (ECMC) was restructured into two separate enterprises--the Coffee Sales and Purchase Enterprise (CSPE) and the Ethiopian Coffee Export Enterprise (ECEE)--in late 1993. While the CSPE is engaged in coffee purchases and transfers from the interior (to the two export auction centers at Addis Ababa and Dire Dawa), the ECEE is in charge of buying coffee at these auction centers from both the CSPE and individual traders, and then exporting it. As a result of the liberalized licensing requirements for coffee exports that were introduced in August 1993 (reflecting significantly reduced issuance and renewal fees), the number of coffee exporters as well as private traders (both collectors (sebsaby) and transporters (akraby)). Have increased, thereby providing for greater competition in the various stages of the coffee marketing process. 1/ The monopoly position of the ECMC in the exporting of washed coffee also ended during 1991/92, as private exporters were allowed to participate in the auctions alongside the ECMC. 2/

Nevertheless, a number of regulations still pertain to the marketing of coffee; these appear to be aimed at controlling the quality and movement of coffee for export. These include: (i) the regulation that the amount of coffee that can be moved by an (unlicensed) individual must not exceed 3 kilograms; (ii) the requirement that all exportable coffee must be channeled through the auction and that all coffee sold in the domestic market must be coffee not acceptable for export (after inspection and grading at the auction centers); and (iii) the restriction that private exporters cannot make coffee purchases outside of the auctions.

In response to the sudden drop in coffee prices in mid-1989 (following the failure of ICO members to agree on quotas), and in order to restore incentives to coffee growers, producer prices were increased substantially, with the increase being financed through both a reduction in the coffee export surtax and a direct subsidy from the budget. More recently, beginning in the 1992 coffee season 3/, and in response to the volatility of international prices, floor prices were set by the CSPE for sun-dried coffee; these floor prices have remained unchanged since then. 4/ Owing to the recent surge in international prices (in mid-1994), these prices are no longer binding. Sun-dried coffee is regularly traded between the sabsabies and farmers on the basis of recent auction prices announced on the public radio and posted at local market places. As regards washed coffee, farmers receive a first payment from the service cooperatives before their washed coffee is sent to the auction market. They then get a second payment if the resulting auction price is above the first payment. Currently, about 60 percent of this difference is distributed to farmers 5/ as a second payment and the remaining 40 percent is reserved (by the cooperatives) for management and reinvestment for farmers. There are no data on the actual producer prices received by farmers. 1/

During 1989/90-1991/92, coffee production is estimated to have grown evenly, albeit modestly, by about 2.5 percent per annum, averaging slightly above 200,000 tons (Appendix Table VIII). After a slight decline in 1992/93, production is estimated to have increased by 5.7 percent in 1993/94, reaching 222,000 tons. While this may be attributed partly to the favorable impact of recent liberalization measures as well as to increases in international prices, the full impact of such measures on the growth in coffee production is presumably restricted in the short term to efficiency gains (during the picking of coffee beans), which is estimated to remain near 5 percent, given the usual cycle between the planting of new coffee and its harvesting. 2/ There is very little information on changes in the total area planted (presently estimated at some 400,000 hectares), although some additional planting is estimated to have taken place in traditional coffee growing areas, such as Eastern Hararge.

Arrivals at the auction centers, 3/ although lower than the record 1989 level, increased considerably to about 67,000 tons during January-May 1994, as compared with an average of 51,000 tons for the same five-month period extending from 1990 through 1993 (Appendix Table XI). While this is partly attributable to the diminished role of the parallel market in the channeling of coffee to the international market, 4/ the continued increases in international prices, which have substantially reduced the differential between the auction price and the domestic price, appear to have contributed to this outcome. However, existing regulations (as discussed above)--coupled with a traditionally high domestic demand for coffee, estimated to absorb about 35-40 percent of total coffee production--continue to induce a diversion of exportable coffee beans to the domestic market through irregular channels.

Pulses (horsebeans, chickpeas, haricots, and lentils) and oilseeds (neug, flax, rapeseed, sesame, and castor beans) constitute about 12-13 percent of crop production, and 17 percent of the total land under cultivation (excluding coffee and other cash crops). 5/ They are grown mostly as rotational crops by smallholder farmers in between the meher and belg seasons. In 1992/93, preliminary estimates suggest that production grew by about 3.4 percent.

(2) Livestock

Ethiopia’s livestock herd of 75 million is estimated to be the largest in Africa. It is difficult to calculate the sector’s value added, since a significant part of meat and dairy production is produced for subsistence purposes and in certain regions, particularly in the highlands, livestock is utilized only to support farming. Ethiopia’s livestock subsector continues to suffer from the vagaries of unpredictable weather conditions, disease, and the lack of a coherent plan for the development of the sector. Although hides and skins and leather are Ethiopia’s second most important export, the sector’s huge potential remains largely untapped. The impact of the recent drought is estimated to have been substantial for livestock herds.

(3) Fertilizer usage

Since the launching of the economic reform program, progress has been made in various areas, including in the usage of fertilizer, provision of credit and agricultural extension services to farmers, and the marketing and pricing of agricultural products. Despite the introduction of the Government’s “New (Fertilizer) Marketing System” in 1991/92, there have been difficulties in ensuring adequate levels of fertilizer usage by the smallholder sector. 1/ The monopoly of the Agricultural Input Supply Corporation (AISCO) in fertilizer procurement and distribution was removed, resulting in a multiple channel procurement and distribution system with significant private sector involvement in retailing, and to some extent, in wholesaling. At the end of 1993, more than 1,000 private retailers were registered by AISCO (about 900 of whom are operational) and an additional 600 are expected to be registered in 1994; these retailers operate alongside approximately 400 service cooperatives and 400 marketing centers, the latter managed by AISCO. It is estimated that the share of the private retailers in fertilizer distribution will increase to over 60 percent by the end of 1994. The total number of wholesalers in 1994 is likely to remain the same as in 1993 (at around 30), because of the large scale of the operations involved. Almost all fertilizer imports continue to be funded by donors through AISCO; in 1992/93 a private company started to import fertilizer as well.

Fertilizer usage in Ethiopia is considered extremely low; only about 15 percent of the farmers are estimated to use fertilizers. In 1993, fertilizer sales to the peasant sector (121,000 metric tons) were about 20 percent lower than in 1992 (156,000 metric tons). 1/ The reasons for the decline included a delay in the announcement of fertilizer prices, 2/ confusion over the associated subsidy scheme, lack of retail market outlets and credit facilities, and administrative constraints placed on the new private sector dealers in the context of the pan-territorial pricing policy pursued by the Government. The Government’s efforts to promote the benefits of fertilizer, which should have been launched at the time of price increases, were also limited. The 1994 sales target is about 160,000 metric tons, slightly above 1992, and is likely to be achieved, given favorable stock positions (partly owing to lower-than-expected sales during the preceding year) and the timely announcement of prices in mid-January. However, during January-April 1994, actual sales by AISCO to the peasant sector (27,510 metric tons) were lower than the level during the corresponding period in 1992 (33,476 metric tons), an outcome partly attributable to the delay in the commencement of the belg rains and the late planting season.

After four years without a direct financial fertilizer subsidy, the Government reintroduced a fixed subsidy of 15 percent on both types of fertilizers in February 1993, amounting to Br 26.43 per quintal for DAP and Br 23.41 per quintal for urea. In 1994, the subsidy was raised to 20-21 percent (to Br 36.96 and Br 31.98 per quintal, respectively). 3/ A National Committee, chaired by the Ministry of Agriculture, with the General Managers of the Agricultural and Industrial Development Bank (AIDB), Commercial Bank of Ethiopia (CBE), and AISCO as members, was recently established (along with regional committees), to facilitate the provision of fertilizer credit to needy farmers as well as to ensure the repayment of credit.

(4) Rural credit

As regards rural credit schemes, the Government recently enacted the Agricultural Cooperative Societies Proclamation (issued on February 1, 1994), which sought to promote the role of service cooperatives (named “cooperative societies” under the proclamation) in the allocation of credit in the rural areas. According to the Proclamation, the functions of the cooperatives will also include acceptance of deposits and the granting of loans to other institutions. It is expected that the AIDB and the CBE will both be involved in the scheme (contrary to the previous credit programs, which included only the AIDB).

(5) Agricultural Marketing and Pricing

In the areas of marketing and pricing of agricultural output, significant progress has been made in reorganizing the Agricultural Marketing Corporation (AMC), which has now been renamed the Ethiopian Grain Trade Enterprise (EGTE). The focus of the EGTE within the new market-oriented economy is significantly different from its past role. 1/ The main objective of the EGTE is to stabilize the grain market through intervention. In a normal year, based on the probability of grain supply fluctuations in the market-dependent population (i.e., urban, nomadic, and producers/consumers in deficit areas), about 300,000 tons of grain are assumed to be needed as a buffer stock and for the purpose of seasonal stabilization. Of this amount, 150,000 tons are planned to be purchased locally, while the rest would be supplied from abroad, principally from donors.

The planned operations of the EGTE were disrupted in 1993/94 because of the adverse impact of the drought. In order to prevent a further increase in producer prices, the EGTE purchased 18,000 tons in 1993/94 (compared with 150,000 tons the preceding year). Currently, the EGTE principally focuses on stabilizing the food situation in the country through the distribution of food aid. About 30,000 tons were provided to the drought-affected regions directly from EGTE’s stocks. As of mid-1994, the EGTE held stocks of 110,000 tons of grains, down from 160,000 tons in January 1994.

Although there is some competition at the wholesale level, which is due to its enormous infrastructure (trucks, warehouses, credit, etc.), the EGTE is by far the biggest participant in the grain market, with a substantial influence on producer prices. At present, producer prices are negotiated between the local committees (acting as agents for the EGTE in the corresponding regions) and the farmers, on the basis of indicative prices set by the EGTE. The EGTE, in turn, sets prices in line with regional price developments in the urban and rural markets.

The reorganization at the EGTE continued during 1993/94. As a result of its new focus on selected markets, its collection points have been reduced from 2,000 to 500. The number of trade centers (buying and selling) have been reduced from 104 to 66 and its coordinating branches now number 17, instead of 27. At the start of the reorganization process, the total staff numbered 4,300. Currently, its staff is 3,089 and it is intended to reduce the staff further to about 2,500. At present, the EGTE makes available its excess storage capacity to the private sector, nongovernment organizations, and AISCO. Though it is not involved in the provision of credit to farmers, the EGTE supplies some inputs to farmers (particularly fertilizer) in association with private traders.

(6) Agricultural land tenure

Since 1991, when the Government announced its policy of providing the farmers with security of land usership and eliminated restrictions on the rental of land and the hiring of labor, there have been no changes in policy on the issue of land ownership. It is expected that the issue will be considered following the establishment of a nationally elected government. 1/

b. Manufacturing

The manufacturing sector in Ethiopia is comprised of light manufacturing products such as construction materials, metal and chemical products, and basic consumer goods such as food, beverages, and textiles. The sector is mainly dominated by about 150 public enterprises (PE), which contribute more than 95 percent of the sector’s value added. The production of PEs is mostly concentrated in the food and beverages (50 percent) and textiles industries (15.3 percent); in addition, tobacco (12.4 percent), leather and leather products (6 percent), rubber, plastic, and cement (5.5 percent) are also important subsectors. 2/ Private sector manufacturing activity, although on a much smaller scale, follows a similar pattern. Of about 150 private sector firms, most are involved in bakery products, textiles, footwear, and furniture production.

During 1989/90-1991/92, manufacturing was adversely affected by the escalation of the civil war; shortages of raw materials, foreign exchange, and fuel; and the frequent breakdowns arising from obsolete machinery and power failures. As a result, the share of the manufacturing sector in GDP declined steadily, from around 5.7 percent in 1989/90 to 2.2 percent in 1991/92. In 1992/93, the sector’s share increased to 3.8 percent, owing to a sharp rebound following the elimination of some of these constraints (mainly the greater availability of foreign exchange). The average capacity utilization rate of the Pes climbed to almost 70 percent, from rates as low as 20 percent in many industries at the end of 1991/92. In 1993/94, the sector’s share increased further to 4.5 percent, though this was in part due to the contraction in agricultural production.

In the short term at least, the sustainability of the recovery in the manufacturing sector is likely to be influenced by the rapidity with which the private sector responds to market incentives as well as by the capacity of the PEs to adapt to the more competitive market environment that has been recently introduced in the economy. With respect to the latter, as a result of the Public Enterprise Proclamation enacted in August 1992, PEs elect their own management boards, assume autonomy in their pricing and production decisions, and are no longer supervised by the Ministry of Industry. 1/ The economic policy reforms that took place during 1992/93, most notably the liberalization of prices, the devaluation of the exchange rate, and the dismantling of the state distribution network, 2/ have created an unusual economic environment for PEs. Although most PEs succeeded in increasing output from unusually repressed levels and in improving their financial viability, the change in the cost structures following these reform measures made direct adherence by the PEs to the former cost-plus pricing principle difficult, partly because of low demand at the new, much higher prices.

In certain sectors, the need for restructuring is particularly pronounced. For instance, the textile sector currently suffers from serious overstocking, owing to a combination of supply and demand problems that have been developing over years, and which became acute during the transition to a market economy. On the demand side, changing tastes as well as competitive prices offered by private sector firms--most importantly by contraband trade from neighboring countries and increased sales of secondhand clothing--have resulted in a serious disruption in sales. Production, on the other hand, has been continuing, owing to the temporary availability of cotton supplies, and the concern to keep firms operating, given their sizable work forces. The tobacco industry is faced with similar problems arising from weak demand. Obsolete machinery is a major problem in many sectors, most notably in the food industry.

c. Energy

Ethiopia is excessively dependent on traditional fuels consisting mainly of wood, agricultural residues, and animal waste. Only about 6 percent of Ethiopia’s energy needs is met from modern energy sources such as petroleum and electricity. As a result, deforestation, with its adverse impact on soil productivity, has been an extremely serious problem.

Power generated by hydroelectric plants supplies most of Ethiopia’s electricity needs (over 95 percent), with total installed electricity capacity at about 370 megawatts. Ethiopia’s rich hydroelectric potential, with 40 river basins, remains almost completely untapped. Greater effort is being made to develop such energy sources. The construction of a new hydroelectric plant (rated at 180 megawatts) was supposed to start in 1992 as a result of an agreement with North Korea. The agreement, however, is currently being renegotiated; four other hydroelectric projects are under study. After almost complete stagnation between 1988/89 and 1991/92, electricity production increased by 11 percent in 1992/93 and 8.5 percent in 1993/94, reaching 1.3 billion kilowatt-hours (Appendix Table XIX). Although only 2 percent of the hydroelectric potential has been exploited, the country has excess capacity at the present rate of consumption. Thus the Ethiopian authorities have been looking into the possibility of marketing electricity to neighboring countries and progress has been made in the construction of transmission lines for the export of electricity.

Ethiopia imports crude oil (mainly from Saudi Arabia) for processing at the Assab petroleum refinery, which is now owned by Eritrea. Ethiopia has a contract with the Eritrean Government for processing the crude oil imported by Ethiopia; the contract is renewed every year. Petroleum products refined at Assab are considered “Ethiopian output” and are either shipped to Ethiopia or sold to Eritrea as an export; about 20 percent of the products refined at Assab are sold in the Eritrean market, the remainder is used in Ethiopia. Transactions are settled in birr, and the prices are revised every quarter based on international market prices. Additional petroleum products are imported from outside (and paid for in convertible currencies), and principally include kerosene, gasoline, and gas oil.

Recently, important steps have been taken to develop new energy sources for Ethiopia, while reducing dependence on forests and the import of petroleum products. Financed under a World Bank loan, a project has been initiated to develop the natural gas deposits situated in the Ogaden region. The Calub gas reserves, estimated at about 2.7 trillion cubic feet (in two reservoirs of 0.7 and 2.0 trillion cubic feet), are significant, with the upper layer (0.7 trillion cubic feet) also estimated to contain substantial associated liquids (LPG, gasoline, kerosene, and diesel/gas oil). The project’s contribution is expected to reach approximately 70,000 tons of oil equivalent per year (or about 8 percent of the current consumption of petroleum and electricity), providing energy sources for urban consumption amounting to about 400,000 tons/year of fuelwood (about 18 percent of current urban consumption).

d. Mining

Mining and quarrying are relatively unimportant in Ethiopia, contributing less than 1 percent of GDP. However, based on the findings of geological surveys, it is estimated that there is substantial potential for expansion. The Ethiopian Mineral Resource Corporation (EMRC) is responsible for the overall management of Ethiopia’s mineral resources (except for marble) from the early stages of exploration through production and exporting; private sector involvement is encouraged to tap the sector’s potential.

Gold, Ethiopia’s most important natural resource, is concentrated in the Legedenbi area (500 kilometers south of Addis Ababa in the Sidamo region). Average annual production is about 3 tons, which is fully sold to the National Bank of Ethiopia (NBE) for export. A project financed jointly by the AfDB is expected to increase annual production capacity to about 4.8-5 tons by the year 1996, assuming that private sector involvement increases. In addition to primary gold (produced by modern techniques), small-scale production takes place in the same region. Production from such sources is estimated to be around 500 kilograms. Tantalite production is about 30 tons per annum, and is principally produced at a pilot plant, also located around Legedenbi. The production potential at this particular reserve is up to 200 tons, but output has been limited by the competitiveness of the international markets.

Potash deposits exist around the very northern tip of Ethiopia by the Eritrean border (close to the Red Sea). Soda ash is produced for local consumption only, for use in the production of glass, bottles, detergents, etc.; current production is about 20,000 tons. If access to world markets could be ensured, production could be raised to 1 million tons. Salt is produced in very small quantities in a localized manner; large quantities are mostly imported from Eritrea. A reserve is being built in the Afar region in the north-east, which is expected to produce about 300-400 million tons of salt. There is an iron ore deposit in the western part of Ethiopia. Although it is not expected to be very rich, the Government is undertaking a feasibility study. Marble is handled by Natural Mining Company, a private enterprise. West Wolega, East Hararge, and Tigray have particularly rich marble deposits.

e. Building and construction

The construction sector was particularly repressed in the last years of the previous government, declining by a cumulative 56 percent during 1988/89-1991/92; this largely reflected shortages in construction materials as well as the overall decline in economic activity. As a result, the sector’s share in GDP declined from 3.9 percent in 1988/89 to 2.1 percent in 1991/92. In 1992/93, along with the overall recovery in economic activity, the sector rebounded by about 17 percent. This strong performance continued in 1993/94, with the sector’s share reaching 2.9 percent of GDP after another year of strong growth (10 percent). However, the share of GDP has remained lower than in the pre-1989/90 period.

f. Transportation

The most common form of transportation in Ethiopia is road transport, accounting for more than 90 percent of all inter-urban freight and passenger movement. The jointly owned single track Ethiopia-Djibouti railway line, although in poor condition, handles most of the remaining load. Civil aviation is not yet very important, although Ethiopian Airlines is considered to have established a highly profitable network of international and domestic routes. Given the rural nature of the economy, nonmotorized transport is also significant in areas where roads have not been constructed.

Ethiopia’s road density is among the lowest in Africa, with an estimated 15.2 kilometers of road per 1,000 square kilometers. The main road network extends radially from Addis Ababa with few interconnecting links; large areas of the country have no linkage to regional and economic centers. The public road network in Ethiopia consists of less than 20,000 kilometers of main and rural roads. Of the main roads, approximately 3,500 kilometers are paved and roughly 8,000 kilometers are graveled. It is estimated that about 65 percent of these roads are in poor condition, with the deterioration in road conditions having intensified over the last decade because of a lack of maintenance.

In December 1992, central control of road transport operations was abolished and the trucking industry was deregulated. Specifically, zonal transport offices (Ketenas), established according to the Proclamation of 1976 to control and regulate commercial road transport operations, were closed. State control of freight rates was removed, and the centralized system of route and cargo assignment to private operators was liberalized. Although the state-owned Ethiopian Freight Transport Corporation (EFTC) is still the largest single trucking operation in Ethiopia in terms of its massive infrastructure (cargo reservation and maintenance facilities as well as load capacity), private sector participation has been increasing. By the end of 1993, in addition to 4,000 trucks that operated in association with EFTC, six private companies, with a capacity of about 1,450 trucks, have been established.

After having risen by about 30-40 percent following deregulation, freight rates subsequently declined to levels lower than those prevailing in the pre-deregulation period, which is partly due to the increase in the supply of trucks, depressed demand, and to some extent, competition from unlicensed operators. Rates appear to vary depending on routes; this is attributable to a number of factors, some of which are associated with problems encountered by the trucking industry, including poor road conditions, the aging fleet and high vehicle operating costs, and an imbalance in the freight cargo load between inbound and outbound routes.

With the independence of Eritrea (with its ports of Assab and Massawa), Ethiopia is now a land-locked country. Assab is the principal port for imports to and exports from Ethiopia. Imports of dry, bulk, and oil cargoes enter through Assab, while other goods, particularly food aid, enter through Djibouti. In addition, smaller quantities enter through Massawa for the northern regions. The government-owned Ethiopian Shipping Lines Corporation is the major shipping company servicing Assab; it owns 11 vessels, with a combined denarius weight (DWT) of 80,000 tons.

3. Employment

Because of the rural nature of the economy, the modern wage sector employment in Ethiopia is small. Currently, the Government provides jobs for about 240,000 employees in the civil service, and another 245,000 are employed in the public enterprise sector, engaged in production and services. Current data on employment by private formal (wage or self-employed) and informal sectors are virtually nonexistent. A survey undertaken by the Ministry of Labor and Social Affairs (MLSA) in 1983, which included establishments employing 10 people and more, put private sector wage employment at slightly less than 200,000 (out of a total of about 700,000), or 37 percent of the public sector wage employment in the same year. Assuming that a similar relation presently prevails, the current size of modern sector wage employment (at those establishments employing 10 people or more) may be estimated at less than 1 million.

Over the last decade, high population growth coupled with less than satisfactory improvements in the productive capacity of the economy sharpened the existing imbalance in Ethiopia’s labor market. On the basis of the last census conducted in 1984, Ethiopia’s population is estimated to have reached 52 million (excluding Eritrea) in 1992, from 40.7 million in 1984, implying an average growth rate of 3 percent per annum. The growth rate of the urban population (6.4 percent) is estimated to have exceeded that of the rural population (2.7 percent) during this period, partly owing to migration. Given that employment opportunities have not expanded at the same rate, it is likely that the unemployment problem has become more acute in the urban centers.

Data on the registered unemployed, vacancies, and placements, as reported by the MLSA and compiled from the Labor Exchange Offices (LEOs), suffer from a number of statistical problems, including the voluntary nature of registrations as well as reporting issues (e.g., cases where vacancies were filled or applicants secured jobs by themselves are not regularly reported to the LEO). However, an increase in the registered unemployed during 1991/92-1992/93, coupled with a decline in reported vacancies, appears to indicate a worsening of the imbalance in the labor market (Appendix Table XXIII). While demobilization of the army, returnees from Eritrea, and an increasing number of graduates from the university are considered to have contributed to this outcome, some retrenchment also took place in the context of the recent restructuring of the ministries and the reform of the PEs. The Prime Minister’s Office (PMO) is in charge of the National Retrenchment Policy, with two committees--the Redeployment Committee and the Vocational Training Committee--dealing with the issue. Currently, policies are directed more toward the training of workers who are likely to be retrenched, as privatization of PEs commences, while keeping them on the payroll. Most retrenchment in the short term is expected to take the form of a redeployment of workers from the center to the regions; reportedly, regions are now operating with 70 percent vacancy rates, and thus there is room for most workers who are likely to be retrenched from the central government ministries to be redeployed at the regional level. There are also plans in the Government to subsidize some private sector firms to employ retrenched workers. The revival in private sector activity is envisaged to have dampened the unemployment problem to some extent.

Labor relations in Ethiopia are currently organized under the new Labor Code, which became effective in January 20, 1993, replacing the Labor Proclamation Order of 1975. The objectives of this proclamation are (i) to establish rules governing worker-employer relations according to basic principles of rights and obligations; (ii) to guarantee the right of workers and employers to form their respective associations and to engage in collective bargaining; and (iii) to strengthen and define the powers and duties of the government agency charged with ensuring that the law is adhered to with respect to collective bargaining, occupational safety, health, and the work environment. 1/ In line with the new Labor Code, LEOs are now operating as service agencies, rather than regulatory placement agencies. There are about 30 offices around the country, operating more or less autonomously, partly owing to regionalization. The unemployed still report to the LEOs, but on a completely voluntary basis. However, no private employment agencies are allowed in the system, consistent with the current policy stance of the International Labor Organization.

4. Wages

In October 1992, at the time of the exchange rate devaluation, the Government introduced significant wage increases, mainly for low-paid state employees. Increases were made on a progressively sliding scale, with the objective of raising the incomes of those groups whose wages had fallen significantly below the minimum consumption basket (calculated by the Wage Board). Following the increase in the minimum wage from Br 50 to Br 105 per month, many employees’ incomes had to increase several “increments” in order to be above the monthly minimum wage. Specifically, those with the lowest salaries of Br 50-60 per month received raises up to 110 percent, while the higher grades of the civil service had increases of just 5-6 percent. After the adjustments, the monthly average wage increased from Br 301 to Br 362. The old scale, based on 19 different grades across 6 categories of work (high-level professional, administrative, junior-professional, clerical, manual, and custodial), and a total of 10 wage increments, was maintained. Current regulations on civil service salaries prescribe that each employee be awarded one full increment every two years, according to a predetermined table of increments with respect to the grade and category of work. Owing to adjustments in October 1992, however, the current structure became obsolete. One implication of the adjustment was a further compression in civil service salaries, as indicated below.

Table 1.

Ethiopia: Wage Structure 1/

article image
Source: Office of the Wage Board.

Intervals are arbitrary; based on data provided by the Wage Board.

Figures do not add up to 100, owing to rounding.

At the time of the wage adjustments in the civil service, similar changes were adopted by PEs. Although wages broadly follow the same structure, in principle, the PEs are now autonomous in determining their wages. It is estimated that average wages at the PEs are broadly comparable to those prevailing in the civil service, although wages in the private sector are higher, reportedly, on the order of about 60 percent. 1/ There are no statutory minimum wages in the private sector.

A preliminary study on the reform of the civil service pay structure has now been completed within the broader context of the civil service reform. The study, which was prepared and carried out by the Wage Board, 1/ recommends modifications to the pay structure with a view to decompressing wages. Further analysis of the study is anticipated before the Government finalizes its decision on the magnitude and structure of pay adjustments.

5. Prices

Over the years, price developments in Ethiopia have largely been influenced by supply-side factors such as external shocks, volatility of agricultural production, and shortages of foreign exchange. 2/ During 1990/91-1991/92, the annual inflation rate averaged about 20 percent, as supply-side disruptions related to the end of the previous government were further aggravated by loose financial policies (Appendix Tables XXXXII). During 1992/93, despite the substantial adjustment of the exchange rate of the birr in October 1992, inflation declined to 10 percent, owing to a tightening of financial policies (including an improved fiscal position and positive real interest rates), as well as the favorable impact of a good meher harvest on food prices, coupled with an overall normalization in agricultural production. Annual inflation continued to decelerate during 1993/94, to 1.2 percent, mainly reflecting the continued positive effect of the good 1992/93 harvest as well as the tailing-off of the adverse effects of devaluation. Inflation as measured by the GDP deflator, which reflects developments in wholesale prices to some extent, also followed a similar pattern.

During January-May 1994, there was an increase in monthly inflation, reflecting pressures developing in the cereal market. The price index dropped in June by 2.0 percent, as a result of a decline in some other food items (e.g., meat and dairy products), while monthly inflation continued to be positive for cereals and nonfood items as a whole. 1/ On an annual basis, reflecting the impact of the good harvests in 1992/93, the increase in food prices fell below that of nonfood prices during 1992/93-1993/94.

CHART 1
CHART 1

ETHIOPIA DEVELOPMENTS IN THE ADDIS ABABA RETAIL PRICE INDEX, 1983–94

Citation: IMF Staff Country Reports 1994, 015; 10.5089/9781451812596.002.A001

Sources: National Bank of Ethiopia; and staff estimates.

Ethiopia has abolished all price controls except for three product categories, which remain subject to price controls at both the retail and the wholesale levels. These include petroleum and petroleum products, some pharmaceuticals, and sugar for household consumption. Controls over prices of a few essential items (cement, corrugated roofing, nails, reinforcing bars, khaki textiles, nylon textiles, traditional abujadid cloth, and cotton yarn) were eliminated in mid-1993. Price controls on spare parts, which also had been removed on a de facto basis in mid-1993, were removed statutorily in July 1994. In mid-May, the retail prices of petroleum products were increased with a view to taking account of the impact of adjustments in the official exchange rate on the import parity price. Utility prices remain under government control. Telephone tariffs were adjusted in July 1994. The present electricity tariffs, which have remained unchanged since March 1986, are currently being reviewed by the Ethiopian Electric Light and Power Authority. A study on water and sewerage tariffs has been completed. Passenger transport tariffs also remain under Government control.

6. Private sector investment

Since the launching of the economic reform program in late 1992, the private sector environment in Ethiopia has improved considerably. In line with the directives laid out in the new Investment Code (Proclamation No. 15/1992), Regional Investment Offices (RIO) 1/ have been established as one-stop offices to issue licenses for domestic investments exceeding Br 250,000. The Investment Office of Ethiopia (IOE), located in Region 14 (the Addis Ababa Region) issues licenses for foreign investments (exceeding US$500,000) and acts as the Regional Investment Office for Region 14. In addition, the IOE issues licenses and grants incentives on behalf of those Administrative Regions that have not yet set up an investment office.

Between July 1992 and June 7, 1994, the IOE and other functioning regional investment offices have issued investment certificates to 1,069 projects, of which eight were foreign investment projects (Appendix Table XV). The total capital implied by these project applications amounted to Br 7.3 billion (Br 6.9 billion issued by the IOE), of which foreign investment projects amounted to Br 677 million (9 percent). The total employment potential of the projects is estimated at some 60,000 permanent and 83,000 temporary jobs. The regional distribution of investment projects is overwhelmingly in favor of Region 14 (75.7 percent), followed by Oromiya (12.5 percent), Region 3 (4 percent), Tigray (3.3. percent), and the Southern Ethiopian Peoples Region (2.7 percent), while the remainder (1.8 percent) is shared by six other regions. Such a concentration is partly attributable to delays encountered at the regional level during the initial period following enactment of the Investment Code.

In terms of the sectoral distribution of investment licenses, manufacturing has the highest share (about 45 percent in value terms), followed by real estate (16.8 percent), agriculture (15.8 percent), hotels and tourism (6.1 percent), and construction (5.7 percent). Manufacturing investment is mostly concentrated in the food and beverages sector, followed by textiles and leather products, building materials, and plastic products. The value of four cotton projects accounts for about 55 percent of the investment licenses issued by the IOE in the area of agriculture. As of June 7, 1994, 114 projects (valued at Br 961,367) were under implementation and 66 projects (valued at Br 338,940) were operational. The implementation ratio (adjusted for those projects which are already operational) still remains low, at about 13.7 percent, owing to uncertainties over the Government’s urban land lease policy and to bottlenecks in the financing area. As regards the latter, given that the credit needs for the proposed investments (as of June 7, 1994) are estimated to exceed Br 3 billion, there is some uncertainty as to the capacity of Ethiopia’s banking system to provide the required credit for these investments.

In assessing the momentum of private investment, it would appear that there has been a slowdown in recent months. During 1992/93, the IOE issued licenses for 503 projects valued at Br 3.6 billion. A similar amount is likely to have been realized for 1993/94, as the value of total investment licenses issued by the IOE reached Br 3.2 billion as of June 7, 1994. However, in terms of the number of licenses issued, a considerable slowdown appears to have occurred; after a significant jump during the first quarter of 1993/94 (211 licenses), only 147 licenses were issued during January-May 1994. Nevertheless, there has been no increase in the number of returned licenses.

As regards licensing policies, the IOE continues to guide investor interests through its tax holidays (which extend 3 to 8 years) and tariff exemptions (mainly on machinery imports). In those areas where the IOE concludes an excessive buildup has been taking place, investors are explicitly advised to divert their interests to other areas. In addition to areas that are exclusively reserved for the Government, 1/ certain key areas are exclusively reserved for the domestic private sector, such as radio and television broadcasting; printing and publishing; retail and wholesale trade; import trade; banking and insurance business; small and medium-size air, rail and marine services (wet and dry cargo, and passenger services); traditional commodity exports; and food processing (for the local market).

7. Privatization

Ethiopia is in the very early stages of the privatization process. A Privatization Agency has been set up (Proclamation No. 87/1994, February 17, 1994) to study and execute the privatization/divestment process, and is in the process of organizing its work activities, including the hiring of additional personnel. Over the next few months, the agency will review the work that has been done by other institutions (e.g., Ministry of Industry), and then formulate the exact time frame and specific modalities for privatization. Some 29 hotels and restaurants and about 120 retail outlets, which were expected to be divested by the Government in the short term, are now expected to be fully disposed of in the coming months. About 100 PEs are planned to be privatized; of these, about 50 will be privatized in the short term (2-3 years). Privatization of state farms will be undertaken along with the remaining PEs over the longer term. However, a small amount of land under the state farms (about 6,500 hectares) has been disposed of or taken over by smallholders. In addition, a few enterprises were returned to their original owners and some are in the process of being returned.

8. Revised GDP figures

A project was initiated in 1992 in the Ministry of Planning and Economic Development, with technical assistance from the United Nations Development Program (UNDP), to revise Ethiopia’s national income accounts for the period 1980/81-1990/91. The project is expected to be finalized, and figures published before the end of 1994. The figures include estimates for 1991/92-1993/94. The main objectives of the project were: (i) to identify and correct methodological deficiencies in the old series after an in-depth study of the national accounts system; and (ii) to expand the system, to the extent possible, to cover hitherto uncompiled accounts. As the project continued, the objectives were enlarged to include estimation of series without Eritrea for the period 1980/81-1990/91 (for consistency purposes) as well as to devise methodologies for the estimation of regional domestic products. 2/

The coverage has been increased mainly in livestock, forestry, transportation, education, government administration, and defense. Revised estimation procedures have been applied in agriculture (crops, livestock, and forestry), mining and quarrying, small-scale industries and handicrafts, water, trade, hotels, and restaurants, domestic services, and other services. In addition, new data sources have been used, including: (i) the number of economically active population (based on the 1984 Population and Housing Census); (ii) surveys on small-scale industries; (iii) the energy survey for forestry; (iv) producer prices of various commodities; and (v) surveys on inputs to the agricultural sector. New aggregates were compiled on “Private Final Consumption Expenditure” and the “Change in Stocks.”

The revised GDP figures are broadly about 25 percent higher than the old figures in constant prices; in current prices, while a similar comparison applies for figures related to early 1980s, the differential grows during the latter half of the 1980s, yielding a revised series about 50 percent higher than the old series. The principal reason for the differential appears to be that in the revised series, the value of the agricultural sector is much higher in current prices, mainly owing to the adoption of market-determined producer prices in its valuation as opposed to controlled prices.

II. Public Finance

1. Overall fiscal performance

During the period 1988/89-1990/91, fiscal performance was strongly influenced by the civil war, the inadequate economic policies of the previous government with its orientation toward central planning, and the deepening economic and political crisis in the country. A relatively favorable budgetary outcome was achieved in 1988/89, but in the following years, fiscal performance worsened sharply. In 1989/90, the overall fiscal deficit (on a cash basis, including grants) reached a record level of 12.3 percent of GDP (14.5 percent excluding grants), reflecting both very low levels of revenues and external grants, and high government expenditures. During the last years of the previous regime, military expenditures were particularly large--averaging 9.7 percent of GDP during the period 1988/89-1990/91.

After the Transitional Government of Ethiopia (TGE) came to power in May 1991, the focus of fiscal policy began to shift. While revenue performance improved, thanks to the economic recovery and other factors, military outlays were radically reduced and spending on both the current and capital budgets was restructured toward infrastructure and social services, reflecting the priorities of the TGE’s economic reform program. Increasingly, as external grants and concessional loans were made available, the generation of counterpart funds from external funding became more important as a source of financing the budget. Also, Ethiopia began to benefit from the rescheduling of its external debt service obligations that resulted from a meeting of the Paris Club in December 1992.

The fiscal performance has improved substantially during the past three years, particularly in terms of reduced domestic bank financing of the budget. In 1991/92 and 1992/93, the overall fiscal deficit (on a cash basis, including grants) fell to about 7 percent of GDP; in 1993/94, the budget deficit (including grants) is estimated at 9.6 percent of GDP, largely reflecting an unexpected shortfall in counterpart funds generation, higher relief outlays associated with the drought, and a growth in donor-financed capital spending. Prior to 1991/92, domestic bank borrowing had played a major role in financing the fiscal deficit. In 1989/90, government borrowing from the domestic banking system peaked at 9.6 percent of GDP, but was sharply reduced to 3.2 percent of GDP in 1992/93 and to an estimated 1.9 percent in 1993/94. Foreign financing, which until 1991/92 had dropped sharply (to 1.2 percent of GDP) because of the reluctance of donors to support the policies of the previous regime, rose to 3.9 percent of GDP in 1992/93 and 7.1 percent of GDP in 1993/94. During 1992/93 and especially in 1993/94, foreign financing in terms of structural adjustment credits and loans that generate counterpart funds gained significantly in importance.

2. Revenues and grants

a. Overview

In 1988/89, owing to favorable economic growth and the strong controls associated with a centrally planned economy, revenue performance was good, with a revenue to GDP ratio of 23.1 percent, almost two thirds of which was from tax revenue. In the following two years, however, the revenue performance worsened progressively, largely because of the intensifying political crisis. By 1991/92, the revenue-to-GDP ratio had fallen to the very low level of 10.7 percent of GDP. The drop in nontax revenue was particularly strong (from 9.1 percent of GDP in 1988/89 to 2.9 percent in 1991/92), mainly reflecting a sharp deterioration in the profitability of public enterprises (with a dramatic decline in capacity utilization rates), and the inability of the Government to command profit remittances as in the past. During the past two years, the revenue-to-GDP ratio rose considerably, to 11.9 percent in 1992/93, and an estimated 13.6 percent in 1993/94. At the same time, a significant change occurred in the tax and nontax revenue structure with the change from a centrally planned to a market economy. Receipts from external grants also rose sharply during the past two years. Largely, this reflected receipts from donors associated with the World Bank-organized Emergency Recovery and Rehabilitation Program (ERRP), as well as grants provided by the European Union (EU), including the STABEX scheme and structural adjustment support.

b. Tax revenue

Tax revenue rose from a very low level of 7.8 percent of GDP in 1991/92 to 8.2 percent of GDP in 1992/93, and an estimated 10.7 percent of GDP in 1993/94, reflecting not only the economic recovery, but also the implementation of important tax and tariff policy reforms and successful efforts to strengthen the institutional capacity of the tax and customs administration. The composition of tax revenue during the whole period changed substantially; most notably, the share of direct tax revenue in total revenue declined from more than 40 percent in 1988/89 to about 31 percent in 1993/94; and the share of import duties and taxes rose substantially from 15 percent to 41 percent during the same period. In addition, export tax revenue declined from about 6.9 percent to 1.7 percent in the period 1988/89-1993/94.

The business profit tax and profit transfers from the public enterprise sector were the most important revenue instruments during the time of the previous government. These tax bases effectively collapsed at the end of the 1980s with the deteriorating performance of public enterprises and parastatals. Revenue from these sources reached its lowest level in absolute terms during the years 1991/92 and 1992/93, reflecting the aftereffects of both the economic and political crisis and the transition to a new Government. For the TGE, restoring tax revenue levels and revising the tax system in a way commensurate with the systemic change from a centrally planned economy to a market-oriented system were important tasks and formed an integral part of the economic reform program.

Specifically, prior to 1991/92, Ethiopia’s tax system was integral to the needs of a command economy; the tax system was very complex, it had a narrow base, and tax rates were set at punitively high levels. Tax assessment and collection methods were administratively cumbersome, although carried out under the discipline connected with a centrally planned economy. Tax reforms undertaken during 1992/93 and 1993/94 have focused on changing this system toward one that is consistent with a market economy, i.e., a liberalized economy with a rapidly growing private sector. These reforms have included a reduction of the maximum rate of the personal income tax and the rationalization of the large number of sales and excise taxes into a simpler system with fewer tax rates, and more services made subject to the sales tax. In a bid to broaden the tax base and make the tax system more equitable, the taxation of rental incomes was introduced. All export taxes other than on coffee were abolished. With the devaluation of the birr, the Government also sought to significantly reduce tariff rates. In addition, it eliminated the higher sales tax rate on imports, incorporating the differential rate into the new tariff structure (see below).

In the context of its regionalization policy, the TGE in 1992/93 delegated some tax collection tasks to the Regions, including the agricultural income tax and land-use fee, the profit and sales taxes on small-scale traders, and the personal income tax from regional government employees and enterprises owned by the Regions (see section II.4 below). Also, the TGE began to strengthen the tax and customs administration. In May 1994, steps were initiated to establish an autonomous Revenue Authority, comprising both the Inland Revenue Administration (IRA) and the Customs Administration.

Concerning specific taxes, after falling to a low level in 1991/92, revenue from the personal income tax increased slightly during 1992/93 and 1993/94 (see Appendix Table XXIV). The personal income tax is levied on the wages and salaries of civil servants and public enterprise employees. This tax was reformed in October 1992; the maximum tax rate was lowered from 85 percent to 50 percent, and the number of brackets was reduced from 17 to 10. However, this reform left the incidence for most low-income wage earners largely unchanged. Also, at the same time, substantial salary and wage increases took place in the public sector, which more than offset the negative impact of the reduction of the maximum tax rate on revenue.

With regard to the business profit tax, in 1993/94, a strong increase in revenue was achieved, reflecting the beginning of the recovery process in the economy during 1992/93; revenue from this tax rose by 43.7 percent compared with the previous year. 1/ However, this was still lower than the level of receipts collected in 1988/89 and 1989/90. Taking into account the substantial improvements in capacity utilization and production output in both the public and private nonagricultural sectors, the buoyancy of the business profit tax during 1993/94 was lower than expected. Further improvements in revenue performance from this tax were hampered by tax evasion in the face of high tax rates, cumbersome collection methods, and the weak institutional capacity of the IRA and the Regional Finance Bureaus.

Other forms of income taxes yielded only limited revenues during the period under consideration. The rental income tax, which was introduced in July 1993, did not produce any noteworthy revenue in 1993/94, owing to delays in the preparation of directives by the Ministry of Finance for the collection of this tax by the Regional Finance Bureaus. General administrative weaknesses in the Regional Administrations also hampered revenue collections. Agricultural income tax and rural land-use fee receipts recovered considerably during 1992/93 and 1993/94, but were still limited by various factors. First, and most importantly, the land-use fee is uniform across the country and does not take into account the considerable variation in land quality and differences in cash-crop growing potentials; substantial revenue from the commercial farming sector is thereby forgone. 2/ Second, the agricultural land tax, which has a maximum tax rate of 89 percent (but a flat rate of 50 percent for all incorporated commercial farms), has proven to be impractical from the perspective of the tax administration; collections appear mostly confined to the minimum tax of Br 10 per farmer. Third, although revenue from the agricultural income tax is largely collected in agricultural surplus areas (e.g., Oromia, Amhara), drought and production shortfalls in large parts of the country during 1993/94 made it impossible to increase revenue collection from this tax further.

Revenue from domestic sales and excise taxes also recovered during the past two fiscal years. After falling by almost 30 percent in 1991/92, receipts rose by 32 percent in 1992/93 and by another 13 percent in 1993/94. However, the increase in revenue in 1993/94 was much lower than expected, considering that the tax reforms introduced in August 1993 resulted in a broader tax base for the sales tax on services, and that higher exfactory prices prevailed for a large number of products. In part, the shortfall may be explained by an increase in arrears of about Br 50 million in sales and excise taxes from public enterprises in the textile, beverages, and tobacco sectors. These enterprises produced large stocks that they were unable to sell because of strong competition from smuggled imports (such as secondhand clothing, alcohol, and cigarettes), and noncompetitive output prices that reflected high production costs and the structural problems of the enterprises concerned.

Revenue from customs duties and taxes rose substantially in 1992/93 and 1993/94, reflecting both higher imports (especially from franco valuta imports subject to high tariff rates) and the impact of the October 1992 devaluation, which had a major effect on the value of imports in local currency terms, particularly during 1993/94. 1/ Revenue from import duties and taxes rose by about 70 percent in 1992/93 and again by about 72 percent in 1993/94. The tariff reforms that the Government introduced in August 1993 also contributed to this outcome--despite a substantial reduction of the maximum tariff rate (from 230 percent to 80 percent) and a reduction in the average nominal tariff rate from 34 to 29 percent. 2/ In addition, the Government has begun to restructure and strengthen the Customs Administration, involving both external technical assistance and the hiring of an external company charged with preshipment inspection. However, revenue was forgone as a result of the smuggling of goods, especially second-hand clothing, cigarettes, and alcohol via Djibouti and other neighboring countries. Recent protocols signed between the Government and neighboring countries (including Djibouti and Eritrea) should improve the exchange of information and result in a reduction of smuggling.

Not surprisingly, revenue from export taxes fell considerably from 1988/89 until 1991/92. In 1992/93, the Ethiopian Government abolished all export taxes, except for coffee, in a bid to promote the expansion and diversification of exports. 1/ Owing to the recovery in the Ethiopian coffee sector, the devaluation of the birr, and rising world market prices for coffee since mid-1993, revenue from export taxes increased ‘Substantially during 1993/94.

During the past fiscal year, the Government made progress in its efforts to collect tax and customs arrears, although sales tax arrears from public enterprises have built up as mentioned above. At the end of fiscal year 1993/94, total tax and customs arrears amounted to about Br 640 million and Br 150 million, respectively, most of which had been incurred by public enterprises and parastatals prior to 1991/92. Involving the Supervising Authority for Public Enterprises, the IRA and the Customs Administration have started to negotiate repayment schedules, and agreements have been reached with a number of public enterprises and parastatals, including the Telecommunications Authority, the Ethiopian Shipping Lines Corporation, and some 15 other public enterprises from various sectors (e.g., grinding mills, printing press companies, and woodworking factories).

c. Nontax revenue

Nontax revenue is principally derived from public enterprise and parastatal receipts, consisting of capital charges, residual surplus, interest payments, and from 1993/94 onward, state dividends. 2/ After having reached a historically high level of 9.1 percent of GDP in 1988/89, nontax revenue plummeted by two thirds until 1991/92, with the ratio to GDP declining to about 3.0 percent in 1990/91 and 1991/92 (see Table 2). The decline reflected the sharp drop in payments received from the public enterprise sector, resulting from the deterioration in financial performance and the inability of the Government to command profit transfers from these enterprises.

Table 2.

Ethiopia: Summary of Government Finance, 1988/89–1993/94

article image
Source: Ministry of Finance; and staff estimates.

Adjusted for extrabudgetary outlays financed from commercial external borrowing before 1991/92.

During the past two fiscal years, the ratio of nontax revenue to GDP remained relatively low. However, sizable profit transfers were derived from a number of public enterprises and parastatals, including the Commercial Bank of Ethiopia, the Telecommunications Authority, the Ethiopian Shipping Lines, the National Bonded Warehouse, the Petroleum Corporation, and the Agency for the Administration of Rented Houses. Also, profits remitted from the NBE increased substantially, to Br 90 million in 1992/93 and Br 151 million in 1993/94. A one-time boost in nontax revenue was recorded in 1992/93 through a grant payment from Israel associated with the earlier emigration of the Felashas (see Appendix Table XXIV).

In 1992/93, the Government initiated a process of restructuring in the public enterprise and parastatal sector. Proclamation No. 25/1992 was issued, through which about 160 of the total of more than 200 public enterprises and parastatals were given an autonomous status. According to this law, these state enterprises and parastatals are now formally under a Supervising Authority within the Prime Minister’s Office, but have autonomy in terms of wages and staffing policy, price policy, and investment and financing decisions (see also Section 1.7). While all other public enterprises and parastatals continue to pay capital charges, residual surplus and interest payments, those enterprises covered by Proclamation No. 25/1992 began to pay state dividends in 1993/94. The Government has devised a schedule, according to which those enterprises to be divested in the short term pay a higher state dividend (up to 90 percent of net profits) than those that will be retained under government ownership over the medium to long term (which will pay up to 50 percent of net profits). At the end of 1993/94, outstanding arrears from public enterprises and parastatals amounted to about Br 200 million, mostly from the period prior to 1991/92.

d. External grants

External grants plummeted from about Br 800 million (4.7 percent of GDP) in 1988/89 to about Br 460 million (2.2 percent of GDP) in 1991/92 as donors significantly curtailed their support of the previous government. During 1992/93 and 1993/94, the flow of external grants increased again, reflecting (i) the realization of counterpart funds from the sale of imported goods and foreign exchange associated with the ERRP; (ii) counterpart funds realized through the EU’s STABEX scheme; and (iii) donors’ recognition of the economic reform efforts of the TGE. In 1992/93 and 1993/94, Ethiopia received grants in kind and untied cash and counterpart funds amounting to 4.1 percent of GDP and an estimated 6.6 percent of GDP, respectively.

After having played only a very minor role before, counterpart funds generated from external grants and loans have become increasingly important for financing expenditures in the Government’s budget over the past three years. In 1993/94, total counterpart funds amounted to 3.6 percent of GDP and financed 14.4 percent of expenditures (see Appendix Table XXIX). However, in 1993/94, the generation of counterpart funds was below expectations, reflecting both lower aid flows than anticipated and the accumulation of arrears, mostly from the Agricultural Inputs Supply Corporation (AISCO) for the payment of fertilizer received under loans and commodity grants.

In April 1993, supported by the Fund and the World Bank, the Government concluded an agreement with a number of donors with regard to the budgeting and management of counterpart funds. Six donors have signed this Memorandum of Understanding on Counterpart Funds to date. 1/ The agreement involves arrangements for the collection of counterpart funds in pooled accounts and the general principles to be followed in the utilization of counterpart funds, the most important of which is that counterpart funds should not be used to finance extra budgetary expenditures. In financing outlays from the pooled counterpart funds accounts, the Government agreed to adhere to a number of important principles, including transparency, verifiability, and cost-effectiveness. A Counterpart Funds Administration Unit (CPFAU) was subsequently established within the Ministry of Finance to monitor the application of the memorandum and any other matter relating to the collection and utilization of counterpart funds. With technical assistance provided by the European Union, the Ministry of Finance has begun to strengthen the capacity and authority of this unit in the management of counterpart funds.

3. Public expenditures

During the last years of the previous regime, total public expenditures rose to very high levels, reflecting both a growing bureaucracy and increasing outlays to finance the civil war; in 1988/89 and 1989/90, total expenditures were 35.8 percent and 32.4 percent of GDP, respectively (see Table 2). In the following years, public expenditures declined, with cutbacks being most drastic on capital outlays; the ratio of expenditures to GDP declined to 20.3 percent in 1991/92. In 1992/93 and 1993/94, as the economy recovered and the new Government began to implement its program of economic reform and development, public expenditures also recovered and rose to 23.2 percent and 29.9 percent of GDP, respectively. Intrinsic to this recovery were government efforts to restructure public spending away from military outlays and toward expenditures on basic economic infrastructure and social services.

a. Current expenditures

In line with the priorities of its economic reform policy agenda, the TGE in 1992/93 began to restructure current outlays in terms of the functional structure of the budget. Current spending on general services (including organs of state, judiciary, defense, public order, and security) and productive activities in economic sectors (e.g., industry, mining, and commerce) was contained; conversely, current outlays for basic economic infrastructure (e.g., roads) and social sectors (health, education, community services) were increased substantially. For example, current outlays for education and health, which had fallen to the low level of 2.8 percent in 1990/91, rose to 3.4 percent of GDP by 1993/94. In contrast, military spending, which had averaged 9.7 percent of GDP in the period 1988/89-1990/91, was cut back to 2.7 percent of GDP in 1992/93, and to 2.3 percent of GDP in 1993/94 (see Appendix Table XXV).

There was also a shift in the structure of the current budget when classified on an economic basis. Prior to 1992/93, the public sector wage bill was severely repressed, with a cap imposed on salary increases for all civil servants earning above Br 600 per month. Almost 80 percent of all employees in the civil service received salaries below Br 400 per month, while an estimated 40 percent of civil servants had salaries below the poverty line; this number appears to have risen to between 50 percent and 70 percent through 1992. 1/ In October 1992, when the Ethiopian birr was devalued, the Government adjusted public sector wages to address this problem and to compensate for the expected cost of living increases associated with the devaluation. The minimum wage was raised by 110 percent, from Br 50 to Br 105 per month; percentage increases were substantially lower for higher salary grades. 2/ Also, the general salary cap was abandoned. Despite the salary adjustment, in 1993/94, about two thirds of all civil servants still received salaries below Br 400 per month. Moreover, the October 1992 adjustment resulted in a further compression of civil service salaries; remuneration levels for high-ranking Government officials remain significantly below those of counterparts in the private sector.

The total wage bill increased from 5.8 percent of GDP in 1992/93 to 7.1 percent of GDP in 1993/94 (see Appendix Table XXVI), mainly reflecting the full impact of the October 1992 salary adjustment. Also, in 1993/94, costs associated with the regionalization policy contributed to the increase; in the context of rationalizing central government ministries and strengthening the regional administration employees, some 4,000 staff were transferred from the Central Government to the Regions (for details see Section II.4 on regionalization). However, this development masks the fact that the wage bill for military personnel was significantly reduced during the past two years.

Since 1991/92, the Government has not increased the total number of civil servants, and the automatic hiring of university and college graduates has been abolished. At the end of fiscal year 1993/94, the Government employed about 242,000 civil servants, including Central Government and Regional Administrations. In September 1993, the Government published a document on retrenchment policy, which provides guidelines and principles on the implementation of the planned retrenchment from both the civil service and the public enterprise and parastatal sector, including compensation programs and packages that will be made available.

Prior to 1991/92, outlays on materials and supply for operations and maintenance (excluding the military) had been allowed to deteriorate dramatically. Over the past two fiscal years, while outlays on materials for military purposes declined strongly, progress was made in terms of increasing materials and supply expenditures in other sectors. In 1993/94, total expenditures for materials rose slightly to 3.3 percent of GDP, although there was still an imbalance between spending on wages and expenditures for materials.

Outlays for grants and contributions comprise payments to international organizations, grants to individuals (e.g., scholarships), and grant contributions to various government institutions. Prior to 1991/92, general price subsidies for agricultural products (especially food grain), which were untargeted and heavily biased toward the urban population, featured prominently in the budget. 1/ These subsidies were abolished in fiscal year 1992/93 except for a modest coffee price subsidy. In the following year, the coffee price subsidy was fully terminated, while a subsidy on fertilizer was introduced. This subsidy, which was targeted at about Br 50 million during 1993/94, was designed as a temporary measure to promote fertilizer usage among small-scale farmers and to increase agricultural production. 2/ In 1993/94, higher-than-expected expenditures for drought relief were incurred through the Relief and Rehabilitation Commission (RRC). Pension payments rose modestly in 1992/93 and somewhat more rapidly in 1993/94, owing to the larger number of civil servants who took early retirement.

During the period 1989/90-1991/92, the Ethiopian Government began to accumulate external interest arrears, amounting to about Br 210 million. Through an agreement reached with Paris Club creditors in December 1992, the TGE was able to reschedule a substantial part of its foreign debt, including arrears (see Chapter VI.5). Consequently, the authorities were able to keep external interest payments at a relatively low level in 1992/93 and 1993/94 (0.5 percent and 0.6 percent of GDP, respectively). In contrast, domestic interest payments increased substantially during the past two fiscal years--from an average of 1.1 percent of GDP for the period 1988/89-1991/92 to 1.6 percent of GDP in 1992/93 and 2.8 percent of GDP in 1993/94. This reflects that a substantial part of the Government’s internal debt, totaling about Br 10 billion, and taken up mostly by the previous government at low interest rates (2-5 percent), has had to be serviced at the market interest rate of 12 percent since October 1992.

Expenditures directly linked to external assistance, and mainly in the form of grants in kind (including technical assistance), were relatively low until 1991/92, but rose strongly during 1992/93 and 1993/94. In part, this reflected high levels of food aid as well as a substantial USAID commodity grant to assist in the rehabilitation of the textile industry.

In 1992/93, the Government began to design social safety net measures for retrenched civil servants and employees from the public enterprise and parastatal sector as well as other groups adversely affected by the country’s political transition and economic reform program, including ex-servicemen of the previous government’s army, refugees and returnees, and other impoverished groups in the rural and urban areas. A steering committee was given responsibility for developing the Government’s policy in this area and supervising the implementation of safety net measures. The following elements are part of the Government’s safety net program: (i) public works programs for the unemployed, especially in the urban areas; (ii) agricultural production subsidies for ex-servicemen, refugees, returnees, and poor farmers; (iii) special projects for elderly people, the disabled, orphans and widows; and (iv) severance payments and other programs for retrenched civil servants and workers from public enterprises and parastatals.

During 1992/93 and 1993/94, the Government spent Br 37 million and about Br 50 million, respectively, on social safety net measures. However, the implementation of these measures occurred at a considerably slower pace than anticipated because of delays in the finalization of the safety net policy as well as in the preparation of guidelines and mechanisms for program implementation at the regional level. The generally weak administrative set-up in a number of Regions, and their general inexperience with implementing safety net programs, were also contributing factors.

Compensation payments to retrenched workers from the public enterprise and parastatal sector have been the most important safety net measures over the past two years. Severance payments and other compensation packages were made available for up to 20,000 workers who were laid off from the Ethiopian Building Construction Authority, the Maritime Transport Authority, and other public enterprises and parastatals, including state farms. Apart from providing severance payments, the Government has also begun to design a number of programs and projects specifically geared to this target group within the context of its retrenchment policy, including a revolving fund to provide credits for retrenched workers who want to start their own businesses.

b. Capital expenditures

Capital expenditures, which had been relatively large through the late 1980s, were sharply eroded in the last few years of the previous government. Capital outlays fell to 6.1 percent of GDP in 1990/91, and further, to 4.6 percent of GDP in 1991/92 (see Table 2), as the new Government sorted out its investment priorities and as the disruptions to government administration from the years of civil war were addressed. During the following two fiscal years, capital expenditures recovered to 8.3 percent and 11.1 percent of GDP, respectively, reflecting the new Government’s commitment to rehabilitate existing infrastructure and initiate new development projects as well as its reassertion of control over the central government ministries. During the past two years, this policy was supported by increasing external financing of capital projects; total external financing of capital projects increased from about Br 300 million in 1991/92 to about Br 900 million in each of the following two years (see Appendix Table XXVIII).

The implementation rate of the capital budget gradually improved from well below 70 percent in previous years to 71 percent in 1992/93, and an estimated 87 percent in 1993/94. Much of the emphasis during the early years of the new Government was on the completion of long-outstanding projects, rather than on the introduction of new projects. Even through 1993/94, much of the outstanding project portfolio is still in need of completion. During 1993/94, the Government embarked on a detailed assessment of its strategies in the different sectors. Together with the World Bank, the Government also engaged in a detailed Public Expenditure Review (PER), which played an important role in the formulation of the 1994/95 budget.

Within the capital budget, substantial expenditure restructuring took place during the past two years, with the focus shifting away from investments in general and economic services sectors, and toward projects in the infrastructure and social services areas. The share of investments in mining, industry, and commerce in the total capital budget fell from more than 25 percent in the years 1989/90 and 1990/91 to 14.8 percent in 1992/93 and 13.2 percent in 1993/94 (see Appendix Table XXVII). In contrast, the share of social services rose from 6 percent in 1991/92 to 12.5 percent in 1992/93 and 18.9 percent in 1993/94. Also, government capital spending on infrastructure (especially roads) increased considerably over the past two years. Within sectors, the TGE also began to shift its capital spending priorities; in 1992/93 and 1993/94, more emphasis was put on primary education and primary health care in the social sectors. In the agricultural sector, substantially less was spent on investments in state farms, while projects focusing on small-scale farmers received more attention.

4. The regionalization policy and fiscal implications

a. Overview

In 1992, the TGE initiated a far-reaching regionalization policy through Proclamation No. 7/1992, which established 14 administrative regions and granted the right to establish self-governments by a “nation, nationality, or people” within prescribed areas. 1/ This policy represented an important departure from the strong tendency of centralizing government administration under previous regimes, and an important step toward the establishment of fiscal federalism and the development of a democratic political system. In June 1992, elections for Regional Councils were held in most of the country, forming the basis for establishing Regional Governments. At present, owing to the merger of four Regions into one, there are 10 official Regions. 2/

Proclamation No. 41/1993, which became effective in January 1993, established the respective powers and duties of Central and Regional Governments. It provided the legal basis for the establishment of 20 ministries and four commissions at the central government level, and outlined the powers and duties of these ministries and commissions, including the responsibility of giving assistance and advice to the Regional Administrations. Also, according to this Proclamation, each Regional Government was allowed to establish, as necessary, up to 20 regional bureaus, following very similar sectoral divisions as applied for the establishment of central government ministries. The Proclamation further outlined the powers and functions of these bureaus.

An administrative base is already in existence in a number of Regions, since the previous government had maintained 30 administrative regional centers. In a few Regions--notably Tigray, Amhara, Oromia, and Addis Ababa--the administrative capacity has traditionally been strong. Shortly after having initiated its regionalization policy, the TGE began to transfer staff from the Central Government to the Regions; by January 1994, about 4,000 professional staff had been redeployed to the Regions. There have also been movements of personnel between Regions. Substantially larger transfers are envisaged during 1994/95 within the context of the restructuring and civil service reform of the Central Government. This will address the recognized shortage of skilled personnel in many of the Regions. In 1993/94, the Government also began a program to strengthen Regional Administrations with support from UNDP.

b. Tax policy and revenues

Overall tax policies are set by the Central Government, but the Regions have substantial scope for designing and implementing their own tax instruments. Proclamation No. 33/1992, which became effective in October 1992, defined the division of revenue-sharing responsibilities between the Central Government and the Regional Administrations. Revenue instruments are categorized as regional, joint, and central instruments.

Regional revenue instruments include the rural land use fee and agricultural income tax; the profit tax and sales tax on individual traders; the profit tax, sales tax, and personal income tax from enterprises owned by Regions; the personal income tax from regional government employees; the rental income tax on property belonging to the Regions; the tax on income from water transportation; royalty and rent of land from other than large-scale mining activities; and charges and fees for licenses and services rendered by the Regional Governments. Revenue sharing applies to the profit tax, personal income tax, and sales tax from jointly owned enterprises; profit tax, sales tax, and personal income tax for incorporated businesses; all taxes on large-scale mining, petroleum, and gas operations; and forest royalties. The Central Government collects revenue from levies on imports and exports, the personal income tax on central government employees and international organizations, the profit tax on enterprises owned by the Central Government, the rental income tax for houses in the possession of the Central Government, taxes on lotteries, and fees and service charges on services provided by the Central Government. According to current practice, the Central Government collects revenue from the profit tax, sales tax and personal income tax on businesses operating across regional boundaries (e.g., transport companies) and other large-scale incorporated private enterprises, as well as the income tax, royalty and rents from large-scale mining activities.

c. Budget preparation and implementation in 1993/94

According to Proclamation No. 41/1993, each Region should prepare and implement its own current and capital budget proposals; regional budgets are then included in the overall consolidated budget of the Central Government. 1/ This process was initiated for the first time for 1993/94, and it appeared to operate relatively smoothly, although only about half of the Regions were able to prepare their budgets independently. In the other Regions, the Ministries of Finance and Economic Development and Planning had to provide assistance to the Regional Finance Bureaus in forecasting revenues and budgeting current and capital expenditures. However, much progress has been made in the meantime; reportedly, most Regions were able to prepare the 1994/95 budgets without central government assistance.

In the budget preparation process for 1993/94, revenue forecasts were subject to a high degree of uncertainty with regard to the capacity of individual regions to collect revenue from the tax and nontax instruments legally assigned to them. Given the distribution of responsibilities for revenue collection and the existing ownership structure of public enterprises and parastatals, only about 20 percent of all revenue (Br 806 million) was expected to be collected by the Regions, while the Central Government would collect about 80 percent in the 1993/94 budget. There were also large differences in revenue shares of individual Regions; Afar, Oromia, and Addis Ababa accounted for about three fourths of the total revenue to be collected by Regions, while three Regions (Afar, Benshangul, Gambella) accounted for less than 1 percent each in the Regions’ revenue share (see Table 3).

Table 3.

Ethiopia: Distribution of Revenues, Expenditures and Central Government Grants, 1993/94 1/

article image
Sources: Ministry of Finance; and staff estimates.

Budget data.

Block grants received from the Central Government.

World Bank data.

Southern Ethiopian Peoples’ Administrative Region (former Regions 7, 8, 9, 10 and 11).

The extremely high expenditure per capita figure shown for Gambella Region may be the result of an underestimated population size.

Very preliminary data on the outcome of the 1993/94 budget suggest that on average, the collection of revenue by the Regions has been somewhat lower than anticipated. In general, both the budget data and the preliminary outcome of 1993/94 mask the fact that relatively large tax bases in a number of Regions, especially Oromia and Addis Ababa, have so far remained unexploited to a significant extent, notably the commercial sector in agriculture, the large numbers of unincorporated businesses in the industrial sector, and self-employed professionals in the services sectors. However, since Proclamation No. 33/1992 has become effective, Regional Administrations now have a strong incentive to improve their own tax and nontax instruments and intensify assessment and collection efforts on the group of taxes for which the revenue wholly accrues to the Regions.

The Regions are expected to finance about 25 percent of their total expenditure with own revenues, the other 75 percent being financed through block grants from the Central Government. These grants appear to have been allocated in a way that reflected the Regions’ own revenue potential as well as economic and development policy priorities and such criteria as population density. For example, taking into consideration its large revenue base, Addis Ababa (Region 14) was only allocated a relatively small grant from the Central Government, whereas those Regions with a very narrow own revenue base were expected to receive very substantial grants from the Central Government (Table 3).

On the expenditure side of the 1993/94 budget, the Regions were expected to be responsible for about 38 percent of total outlays, with roughly the same shares in the current and capital budgets, 38 percent and 37 percent, respectively. However, the Regions’ shares in individual sectors varied greatly, with the highest shares in health (85 percent) education (81 percent), and agriculture and natural resources (60 percent).

The analysis above (Sections II. 3 and II.4) has shown, that the implementation of budgeted expenditures during 1993/94 was hampered by the weak administrative capacity in the Regions. This was particularly so with regard to the execution of some programs under the current budget, notably social safety net measures. More importantly, capital budget implementationat the regional level was generally more difficult than originally envisaged. Until late into 1993/94, capital budget spending at the regional level was very slow and capital outlays were ultimately concentrated at the end of the fiscal year. It also appeared that in a number of Regions, capital budget implementation ratios were considerably below the level achieved by the Central Government.

d. Domestic and foreign financing

In 1993/94, the Regional Administrations depended on their own revenue and transfers from the Central Government to finance their operations; preliminary data suggest that Regions did not borrow from the domestic banking system. Although domestic bank borrowing by Regions is allowed de jure, de facto this has not yet happened. The Central Government has indicated its strong determination to contain domestic bank borrowing by the Regions. With regard to foreign financing (and external grants). The Central Government has full control over the negotiation and spending of foreign funds; moreover, the management of foreign currency is also reserved for the Central Government. Nevertheless, in 1993/94, the Central and Regional Governments have begun to develop a policy that will allow Regions a greater degree of autonomy to negotiate external grant assistance programs with donors.

III. Balance of Payments

1. Overall developments. 1989/90-1993/94

Ethiopia’s balance of payments has been significantly affected by weather conditions, the terms of trade, increasing lack of competitiveness of the birr as inflation accelerated followed by improved competitiveness with the October 1992 devaluation, the civil war and change of government, and the emergency relief and rehabilitation efforts provided by the international community (Appendix Table XXXI). During 1989/90, exports fell by 16.7 percent, reflecting mainly a decline in coffee prices; the current account deficit (including transfers) remained broadly unchanged from the previous year, at 1.5 percent of GDP, owing largely to a sharp cutback in regular imports. 1/ Reflecting the successive overall balance of payments deficits of the previous five years, gross official reserves in terms of weeks of imports declined to only slightly more than one week at the end of 1989/90. Moreover, as the overall balance of payments deficit was too large to be financed by international reserves, Ethiopia accumulated external payments arrears for the first time.

Exports in 1990/91 fell further to their lowest level in 16 years, mostly because of a drop in recorded coffee shipments, but most other exports fell in value as well, in response to falling prices and increased smuggling through neighboring countries as the exchange rate became increasingly uncompetitive. Imports rose mainly because of an increase in commodity assistance and aircraft purchases. Growing interest obligations and a reduction in the normally strong external earnings by Ethiopian Airlines turned the services balance negative for the first time. While grant receipts rose sharply (reflecting mainly famine relief and emigration-related transfers late in the year), this was insufficient to offset the deterioration in the trade and services accounts, and thus the current account (including transfers) widened to 2.7 percent of GDP.

Developments in 1991/92 mainly reflected the initial recovery and reconstruction following the change of government in May 1991, although conditions remained difficult after mid-1991 despite the end of the large-scale military conflict. Severe congestion was reported at the port of Assab, Eritrea, capacity utilization in manufacturing was only 20-40 percent (owing to a severe shortage of imported inputs and parts), and a lack of fertilizer and other inputs limited recovery in agricultural production despite favorable weather conditions. Official reserves in weeks of imports strengthened, but this was largely due to the further compression of regular imports and a continuing accumulation of external payments arrears. Exports of coffee continued to drop sharply (with the continuing weakening of world coffee prices) and other exports also fell, primarily because of the further increasing overvaluation of the official exchange rate.

In 1992/93, the balance of payments began to be influenced by the stabilization and structural policies that were implemented, starting in October 1992 with the major devaluation, 1/ the increasing availability and access of the private sector to foreign exchange from donor sources, and the initial steps taken to liberalize the foreign trade and domestic marketing systems. Despite weakening world prices, export earnings increased by 41 percent, as the volume of coffee exports nearly doubled (reflecting an increasing share of exports through official channels). Imports (net of aircraft purchases) rose by 5.8 percent, in line with real economic growth. In view of the severe shortage of imported inputs and spare parts in prior years, this growth was less than anticipated, owing mainly to investment bottlenecks, tight private sector credit policy, delays in tariff reform, and slow disbursements under the Emergency Recovery and Reconstruction Project (ERRP). Overall, the trade balance worsened modestly during 1992/93. Service earnings weakened during 1992/93, as fare increases authorized for Ethiopian Airlines did not completely cover the full impact of the exchange rate depreciation. Combined with declines in private and official transfers to levels more in line with the years prior to 1991/92, the current account balance (including official transfers) worsened during 1992/93. The capital account improved, however, as loan disbursements increased by 21.3 percent, although even this proved less than expected, as project implementation lagged and ERRP inflows picked up only toward year-end.

The current account (including official transfers) remained broadly unchanged during 1993/94 (although increasing slightly as a share of GDP). Exports increased by 23 percent, with a significant increase in noncoffee exports, reflecting improved competitiveness following the October 1992 devaluation. Coffee exports benefitted from the higher international prices that began to emerge in May 1994; however, initial reports suggest that the volume of exports barely changed, reflecting disruptions arising from the restructuring of the ECMC, strong domestic coffee demand, and increasing competition in the world market for high-quality varieties. Imports (net of aircraft purchases) grew by about 12 percent, reflecting large disbursements under the ERRP and increased availability of foreign exchange from the fortnightly foreign exchange auctions held starting in May 1993. In the services account, an increase in net interest payments was partly offset by an improvement in other services such as travel (including tourism) and transportation (including Ethiopian Airlines). The capital account swung sharply into surplus, largely because of disbursements under the World Bank structural adjustment credit (SAC) and the Emergency Recovery and Reconstruction Project (ERRP), leading to a substantial overall balance of payments surplus. As a result, international reserves strengthened significantly at the NBE and the CBE, with official reserves increasing from 14.7 weeks of imports (net of aircraft purchases) at end-June 1993 to 24.3 weeks at end-June 1994; however, some of these reserves were not freely available because they were being blocked in double signature accounts with donors or were already committed for imports.

2. Merchandise trade

During the 1989/90-1991/92 period, the trade and exchange system was very restrictive, with a significantly overvalued real exchange rate, 1/ a foreign exchange surrender requirement, administrative allocation of foreign exchange primarily to the government sector, high and widely dispersed tariffs, and a burdensome trade licensing system. The overvalued exchange rate led to rationing and thereby: (i) provided incentives for traders to operate in the parallel (illegal) foreign exchange market at a more depreciated rate; (ii) induced exporters to ship through unofficial channels (smuggling) to avoid having to surrender their foreign exchange; and (iii) protected import-competing firms because of the high cost of acquiring foreign exchange through the parallel market (thereby compressing imports). Additionally, trade volumes fell to a low level during 1990/91, owing to the war and subsequent change of government in May 1991.

In 1992/93 and 1993/94, the authorities devalued Ethiopia’s currency, the birr, developed a foreign exchange auction system, liberalized the tariff structure and other trade-related taxes, and streamlined the trade licensing system (see Section IV). These reforms have stimulated and helped return exports to official channels, reduced protection of import-competing firms, and increased the supply of foreign exchange to private traders, thereby stimulating import purchases.

a. Exports

Ethiopia’s main exports are coffee, gold, leather and leather products, chat, petroleum products, sugar and molasses, textiles, pulses, and fruits and vegetables (Appendix Table XXXII). Among these categories, coffee is by far the most important export good, constituting on average (1989/90-1993/94) about half of Ethiopia’s export value. Coffee exports were affected by low world coffee prices during the 1989/90-1991/92 period, as well as by restrictions in the domestic market and by an increasingly overvalued exchange rate. Increases in world coffee prices, devaluation, and domestic market reform stimulated coffee exports, particularly through official channels, during 1992/93 and 1993/94.

Since Ethiopia’s share in the world market is currently under 2 percent, receipts from coffee exports have been heavily influenced by exogenous developments in the world coffee market. Specifically, coffee prices declined sharply from 1988 to 1992, when they stood at their lowest levels in 20 years. Coffee prices only began to strengthen in 1993, especially during the second half of the year; price increases continued through the remainder of 1993/94. The main reasons for the rise in prices during 1993 were reductions in world stocks, as world production fell short of consumption, and the implementation of an export retention scheme by the major coffee-producing countries.

Ethiopian coffee also came under increased competition in the world market, especially for the high-end varieties for which it had been highly competitive. Reduced demand in some of Ethiopia’s principal markets for high-quality varieties of coffee (viz., Japan, Western Europe, and Saudi Arabia), together with the emergence of new suppliers of competing high-end varieties (such as Yemen), led to downward pressure on prices for some of Ethiopia’s high-quality coffee varieties such as Harar coffee.

Domestic market factors have gradually facilitated a strengthening in volume. The market was heavily regulated during 1989/90-1991/92, with the sector subject to marketing controls, restrictions on entry into the unwashed coffee export sector, the monopoly position of the government-owned ECMC regarding exports of washed coffee, and the overvaluation of the exchange rate. These factors reduced producer incentives and encouraged smuggling. 1/ As a result, the volume of coffee exports dropped during 1990/91 and 1991/92.

However, the coffee market was substantially reformed during 1992/93 and 1993/94. The October 1992 devaluation substantially increased incentives for export. The export licensing system was streamlined and minimum capital requirements were eliminated for exporters in December 1992, lowering the cost of exporting coffee and opening up access to the coffee export market to a wider range of private traders. The reduction of coffee export taxes in August 1993 also improved incentives for coffee exporters. Importantly, the ECMC was split into separate domestic and export enterprises in November 1993, placing private exporters on an equal footing with the new ECEE. These reforms have improved incentives for exporting through official channels and for improving the quality of coffee by, for instance, improving harvesting and storage methods and by bringing more coffee beans to washing stations. 2/ Many washing machines were destroyed during the change of government in May 1991 and these are now being rebuilt. As a result of these domestic market reforms, export volume nearly doubled during 1992/93.

With the recent strengthening of international coffee prices since summer 1993, Ethiopia’s coffee exports increased during 1993/94. However, the increase in export volume was less than expected, owing to the time needed for trees planted (following the change of government) to bear fruit, the diversion of coffee into the domestic market (as strong local demand boosted domestic prices above export prices), reduced government shipments in the months following reorganization of the former ECMC, and the roughly half-year lag between world coffee price increases and the pass-through into shipments for export.

Leather and leather products (hides and skins) are Ethiopia’s principal manufactured exports. In value terms, the share of hides and skins exports have fluctuated between 14 percent and 18 percent of total exports during the period, with 1992/93 exports at the lower end of this range, owing to a sharp drop in export prices (corrected for devaluation) during 1992/93. Starting from a high base in 1989/90, exports of hides and skins declined in terms of both value and volume during 1990/91 and 1991/92 because the sector suffered from an increasingly overvalued exchange rate, excessive regulation, and the effects of the war. During 1992/93, supply increased owing to devaluation, liberalization of domestic assembly activities, and integration of previous war zones into the national economy; this led to an increase in export volume and, correcting for devaluation, a decrease in export prices. Overall, export value increased substantially during 1992/93. The export volume continued to increase significantly during 1993/94 and this was accompanied by higher prices.

Exports of chat increased dramatically during the period, more than tripling in value terms from 1990/91 to 1992/93. This trend reflects the liberalization of the domestic market for chat, which was government controlled.

Petroleum products exports (surplus residues from the refining process of crude petroleum from the oil refinery at Assab, Eritrea) have followed the general trend in exports, dropping from 1990/91 to 1991/92 and then rising in 1992/93. The growth in export volume during 1992/93 (38 percent) was more than double that in export unit value, reflecting the increasing capacity utilization of the refinery at Assab.

In general, the volume of exports of pulses, oilseeds, oilseed cake, sugar and molasses, and fruits and vegetables has been affected by weather conditions and by changing foreign demand. Additionally, the export volume of each of these commodities fell very substantially during 1991/92, given the displacement of farmers as a result of the civil war. The loss in volume was especially significant for pulses, sugar, and molasses, given their historical importance in total export value. While there is substantial scope for increased exports of these commodities, developments during 1993/94 suggest that the potential for growth in these exports lies in the future. Gold exports in Ethiopia’s balance of payments consist of purchases of gold by the NBE from the Ethiopian Mineral Resources Development Corporation; an offsetting debit entry is contained in the capital account under short-term capital (net).

b. Imports

By end-use category, Ethiopia’s main imports consist of semifinished goods, crude petroleum and petroleum products, transport and industrial capital goods, and durable and nondurable consumer goods (Appendix XXXV). Other imports include raw materials and agricultural capital goods. Total imports (valued in birr) fell during 1991/92 and rose during 1992/93. Most of the increase in the birr value of imports during 1992/93 may be attributed to the 142 percent increase in the birr/U.S. dollar official rate (from Br 2.07 = US$1 to Br 5.0 = US$1) in October 1992; the average exchange rate for imports in 1992/93 depreciated by about 66 percent, while the birr value of total imports (excluding aircraft) increased by 97 percent. Most of the remaining increase in total imports can be accounted for by increased imports of crude petroleum, which expanded once an agreement had been reached with Eritrea on the terms for Ethiopia’s use of the Assab refinery. Imports of raw materials and capital goods were weaker than expected during 1992/93, principally because of delays in the implementation of projects under the ERRP.

The growth of imports during 1993/94 reflected the pickup of disbursements under the ERRP. Imports of cereals and other items, which had been expected to surge, reflecting the needs of those adversely affected by the erratic weather conditions of 1993, were held to normal levels, principally because of delays in the provision of drought relief supplies by donors, together with constraints at the port of Assab and in the making of deliveries to the interior of Ethiopia.

c. Terms of trade

Ethiopia’s terms of trade deteriorated modestly during 1990/91 and then recovered during 1991/92 (Appendix Table XXXVI). In 1992/93, Ethiopia experienced a terms of trade deterioration of over 20 percent, which was the largely to decreases in the export unit values of coffee, leather, and leather products associated with a weakening in world market prices. The terms of trade appear to have improved markedly in 1993/94, owing to stronger world prices for Ethiopia’s key exports; in the case of coffee, the rebound in world prices began in mid-summer 1993.

d. Direction of trade

During the period 1988-92, Ethiopia’s major trading partners were the European Union (EU), the United States, Japan, Saudi Arabia, Sweden, and the former Soviet Union (FSU) (Appendix Tables XXXIX and XL). During the period 1989-91, the share of imports from the EU averaged 41 percent of total imports (including commodity aid); the share fell to 35 percent during 1992. The share of Eastern European countries had been about 20 percent of total imports during 1989-90 (with the FSU by far the most important supplier), but fell to 3 percent in 1991 and 1 percent in 1992, as trade with the FSU dropped precipitously. Imports from the United States were steady at 5 percent of total imports except for 1991, when they rose to 13 percent. Japanese imports were fairly steady during the period at about 7 percent of total imports. Imports from Saudi Arabia, which had averaged 2 percent of total imports during 1989-1990, increased substantially to 10 percent during 1991 and 17 percent during 1992. Within Africa, imports from Djibouti more than doubled in birr terms during 1992, principally because of blockages at the port of Assab.

During the same period, Ethiopia’s exports to the EU averaged about 45 percent of its total exports. After having decreased to 39 percent in 1990, the EU’s share increased to 52 percent in 1991 and then dropped to 40 percent in 1992. There was a notable drop in exports to Germany, especially during 1992, owing to decreased demand for high-quality Harar coffee. Exports to the United Kingdom increased strongly in 1992. Exports to Eastern European countries amounted to 11 percent of total exports in 1989, falling to 5 percent in 1990, 1 percent in 1991, and 0.2 percent in 1992; this reflected a steady decline in purchases by the FSU during the period, which had been the largest export buyer in the group of Eastern European countries. The share of exports to the United States declined steadily over the period (from 15 percent in 1989 to 4 percent in 1992), while Japan’s share in total exports increased from 11 percent in 1989 to 23 percent in 1991, falling slightly to 22 percent in 1992. Saudi Arabia increased its share of total Ethiopian exports substantially, from an average of 8 percent during 1989-91 to 20 percent in 1992, owing to exports of coffee, live animals for religious ceremonies, and other agricultural products such as flour.

3. Services and transfers

Ethiopia’s services account has traditionally recorded a surplus, resulting mainly from net receipts from transportation and government services (Appendix Table XLI). Net receipts from transportation have been attributable principally to passenger and port services of Ethiopian Airlines and Ethiopian Shipping Lines. Government services include mainly embassies (both foreign embassies in Addis Ababa and Ethiopian embassies abroad), international organizations located in Addis Ababa, and foreign technical assistance. During 1989/90-1992/93, the combined surplus from these services more than offset net payments on investment income, travel, and (during 1990/91-1992/93) in other services. Other services include claims received from and premiums paid on nonmerchandise insurance provided by nonresidents. The travel account includes foreign exchange and travelers’ checks from foreign travelers in Ethiopia. While net receipts in other areas of the services account were stable or declining, transportation receipts grew by 48 percent in 1991/92, and by 109 percent in 1992/93 (reflecting the depreciation); receipts fell during 1990/91, owing to the war and change of government.

The strong growth trend in transportation receipts is mainly due to the opening of new and profitable international routes by Ethiopian Airlines and efficiency gains in existing operations. As a result, net receipts from nonfactor services increased from Br 144.5 million in 1990/91 to Br 312.7 million in 1992/93. With net payments on investment income rising from Br 138.8 million in 1989/90 to Br 348.7 million during 1992/93 owing to the devaluation, the services account declined from Br 106.6 million in 1989/90 to minus Br 7.6 million in 1990/91, rising to Br 27.4 million in 1991/92, and then falling to minus Br 36.0 million in 1992/93.

Ethiopia’s transfer receipts increased steadily during 1989/90-1992/93, reflecting increased flows of foreign assistance from both private agencies and official organizations. Private transfers are accounted for mainly by donations from nongovernmental organizations (NGOs), while transfers from private individuals have assumed lesser but growing importance. Official transfers now come from the EU under the Stabilization of Export Earnings (STABEX) scheme and other programs, and also from bilateral donors such as the United States and Japan.

4. Capital account

Disbursements of public long-term loans declined during 1989/90-1991/92, but grew substantially during 1992/93; repayments increased steadily during the period prior to 1992/93, with an increase in birr terms during 1992/93 reflecting the devaluation (although amortization payments fell slightly in terms of SDRs) (Appendix Table XLII). In the period prior to 1992/93, almost all foreign loans received by Ethiopia were in the form of project assistance. While the transport and communications sector had long been the recipient of a large portion of official long-term loan disbursements, this emphasis especially increased during 1992/93 owing to aircraft purchases of Br 510 million by Ethiopian Airlines, along with social services; conversely, the share of the industrial and energy sectors in total disbursements declined. Disbursements (excluding loans to Ethiopian Airlines) declined from Br 969.4 million in 1988/89 to Br 566.4 million in 1991/92; during 1992/93, disbursements under the World Bank’s SAC were accompanied by those of bilateral donors such as the United States, for a total of Br 1,282.8 million.

About two thirds of public sector long-term borrowing in 1989/90 was disbursed by bilateral donors; during 1992/93 nearly two thirds was accounted for by multilateral borrowing. Of the multilateral lenders, the African Development Bank disbursed the largest amount during 1992/93 (Br 394.6 million), with the World Bank disbursing Br 299.7 million under its SAC and other facilities. The Fund disbursed Br 97.7 million under the first-year structural adjustment facility (SAF) arrangement. Among the bilateral lenders, the FSU and the former Yugoslavia were traditionally important sources of official loans for Ethiopia, but these are currently insignificant. Official loans have increasingly been supplied by the United States, with its share of bilateral official loan disbursements reaching 98 percent during 1990/91-1992/93.

Reflecting large borrowing in previous years, amortization payments (on an accrual basis) on public long-term loans increased in birr terms during the period, although as noted above, in 1992/93, this principally reflected the depreciation of the birr. Net disbursements have mostly been absorbed by the Central Government. In fact, the remainder of the public sector experienced net outflows of long-term capital during 1992/93. Also, amortization payments have been financed partly by running arrears, which amounted to Br 1.3 billion at end-1991/92 (excluding arrears on ruble-denominated debt to the FSU of Rub 625.9 million at end-December 1992). In view of its foreign exchange shortage, Ethiopia unilaterally decided to postpone payments on non-OECD bilateral loans as of November 1989 and on OECD bilateral loans starting from October 1990, with the exception of collateralized loans (Ethiopian Airlines) or those owed to contractors with work in progress. Excluding ruble-denominated debt to Russia, the United States, Italy, Libya, and Germany (including debt to the former German Democratic Republic) were among Ethiopia’s more important creditor countries.

In addition to net official long-term capital flows, the capital account also includes net short-term capital flows. This consists of net trade credits and other net flows. The latter item, other net flows, consists of working balances of Ethiopian Airlines and Ethiopian Shipping Lines, as well as entries offsetting commodity gold exports 1/ and, to maintain consistency with Ethiopia’s monetary survey, certain inflows of official transfers. 2/

5. External debt

Total new external long-term debt commitments contracted by the Central Government fluctuated considerably during the 1989/90-1992/93 period, ranging from Br 166.8 million in 1990/91 to Br 3.3 billion in 1992/93 (Appendix Table XLV). At the same time, there were significant variations in terms (maturity, grace period, and interest rates), with the average maturity tending to increase during the period. In general, these observations also apply to debt contracted with suppliers and commercial bank creditors during the 1989/90-1990/91 period, but no new suppliers’ or commercial bank credits were contracted during 1991/92 and 1992/93 (Appendix Table XLVI). This latter development reflects the Government’s efforts to borrow only from official creditors on concessional terms.

Ethiopia’s disbursed outstanding external debt (including interest arrears, but excluding ruble-denominated debt to Russia) increased from minus Br 6,409.8 million (49.6 percent of GDP) at end-June 1991 to Br 7,985.0 at end-June 1992 (50.1 percent of GDP) and Br 19,024.1 million (92.0 percent of GDP) at end-June 1993, the latter increase reflecting the devaluation (Appendix Table XLVII). The share of debt owed to multilateral lenders (including the World Bank Group) remained at 27 percent during this period, while the share of other multilateral organizations (including the African Development Bank/Fund) increased from 14 percent at end-June 1991 to 16 percent at end-June 1993. For bilateral creditors, the share of OECD countries in outstanding debt increased from 30 percent at end-June 1991 to 33 percent at end-June 1992, declining to 31 percent at end-June 1993. The United States, Italy, and Germany (including the former German Democratic Republic) were Ethiopia’s largest bilateral OECD creditors. The share of non-OECD bilateral creditors declined from 29 percent at end-June 1991 to 26 percent at end-June 1993. Among these, Russia, the former Czechoslovakia, and Bulgaria were the largest former-CMEA creditors of Ethiopia, while Libya, the former Yugoslavia, and the Democratic People’s Republic of Korea, were Ethiopia’s largest non-OECD, non-CMEA creditors. 1/

Ethiopia’s debt service ratio increased during the 1989/90-1992/93 period as a result of the increase in Ethiopia’s external debt during the 1980s. As a percentage of exports of goods and nonfactor services, the debt service ratio increased from 45.3 percent in 1989/90 (accrual basis) to 82.1 percent in 1991/92; the debt service ratio (prior to debt relief) fell to 71.8 percent during 1992/93.

In December 1992, representatives of Ethiopia and 12 countries, with observers from five other countries and four international organizations, met for the first time at a Paris Club meeting to discuss Ethiopia’s request to alleviate its external debt service obligations to official creditors. 2/ In the Agreed Minute of December 16, 1992, the representatives of the participating countries agreed to reschedule, under enhanced concessional terms, outstanding arrears as well as debt service falling due during a three-year consolidation period ending in late 1995. The rescheduling applied to all publicly guaranteed debt contracted prior to end-December 1989. Ethiopia agreed to seek debt rescheduling on comparable terms from official bilateral creditors other than the participating countries. Multilateral and private commercial debts were excluded, as well as debts incurred by Ethiopian Airlines. This rescheduling reduced the total debt service ratio (including nonparticipating countries as well as multilateral organizations) to 65.4 percent during 1992/93.

IV. Exchange and Trade System 3/

1. Exchange arrangement

The currency of Ethiopia is the Ethiopian birr, which was pegged to the U.S. dollar from February 1973 until May 1993. At that time, a foreign exchange auction was introduced, and two rates were in force: an official exchange rate, applicable to a limited range of transactions, and a secondary exchange rate, which was equal to the marginal rate from the most recent auction. The official rate was set at Br 2.07 - US$ 1 from February 1973 to September 1992. With effect from October 1, 1992, the official rate was raised to Br 5.00 - US$ 1. Subsequently, it was raised to Br 5.13 on April 1, 1994, to Br 5.58 on May 16, 1994, to Br 5.66 on June 1, 1994, to Br 5.59 on July 1, 1994, and most recently stands at Br 5.57 effective August 1, 1994. The official exchange rate is applicable to the following foreign exchange transactions: imports of petroleum products, pharmaceuticals, fertilizer, as well as official debt service and government contributions to international organizations and its foreign offices.

Since May 1, 1993, the Government has made foreign exchange available through a biweekly Dutch auction system, with the successful bidders receiving an allocation of foreign exchange based on the exchange rate contained in their bid. 1/ The amount to be auctioned by the NBE has been publicly announced in advance since Auction 4 (June 12, 1993), although the amount actually disbursed may differ from the announced amount. Bids for foreign exchange in the auction may be used to import items not included on a negative list. Applicants must deposit 100 percent of the birr equivalent in advance. Successful bidders must withdraw the foreign exchange within 30 days of the auction. There are penalties (e.g., they may not be allowed to participate in the subsequent auction) if successful bidders decline to use the foreign exchange. The marginal exchange rate from the auction (“secondary market exchange rate”) serves as the exchange rate until the next auction for all transactions outside the auction, except for transactions where the official rate is applicable (listed above).

The marginal rate from the auction depreciated through the first 34 auctions (though not always steadily). After a tentative start, during which the marginal rate was equal to the official rate (Br 5.00 - US$1) during the first four auctions (May 1, 15, 29, and June 12, 1993), the marginal rate depreciated rapidly to Br 5.90 through the ninth auction (August 21). Increased spending by government ministries during the end of the Ethiopian fiscal year, which ends July 7, may have been partly responsible for increased demand during Auctions 5 (June 26) and earlier. Thereafter, during Auction 11 (September 18), the rate fell to Br 5.10, and then fell further in Auction 12 (October 2) to Br 5.01, posing problems for coffee exporters, whose birr receipts on existing dollar-denominated contracts were consequently reduced. The authorities responded by decreasing the amount of foreign exchange actually sold, from US$10.3 million in Auction 10 (September 4) to US$5.5 million in Auctions 13 and 14 (October 16 and 30). In subsequent auctions, the rate depreciated steadily, reaching Br 6.29 during Auctions 27-29 (April 30, and May 14 and 28, 1994), dipping to Br 6.15 in Auction 33 (July 23), and settling at Br 6.16 in Auction 34 (August 6, 1994). This was accompanied by an increase in the amount sold, which peaked at US$19.4 million during Auction 28 and subsequently declined to US$9.5 million in Auction 34.

CHART 2
CHART 2

ETHIOPIA DEVELOPMENTS IN THE FOREIGN EXCHANGE AUCTION, MAY 1993 – AUGUST 1994 1/

Citation: IMF Staff Country Reports 1994, 015; 10.5089/9781451812596.002.A001

Sources: Notional Bank of Ethiopia; and staff estimates.1/ By “Dutch” foreign exchange auction held every two weeks; Auctions No.1 - No. 34
CHART 3
CHART 3

ETHIOPIA DEVELOPMENTS IN INDICES OF NOMINAL AND REAL TRADE–WEIGHTED EXCHANGE RATES, 1980–94 1/

(1980=100)

Citation: IMF Staff Country Reports 1994, 015; 10.5089/9781451812596.002.A001

Sources: International Financial Statistics: and National Bank of Ethiopia.1/ Based on marginal rate of foreign exchange auction since May 1993.

An examination of detailed bidding information shows the leading commodity groups and the average bid size, as well as the importance of the private sector, the latter having received only limited allocations of foreign exchange prior to the introduction of the auction system. Through the first 34 auctions, 43.9 percent of foreign exchange went to private bidders. Also, successful bids were US$66,800 on average during Auctions 1-34; the average bid size increased during Auctions 20-31, with 9 of these 12 auctions showing average bids greater than US$66,800. The leading commodity categories in auction bids during the first 25 auctions were: machinery and equipment (an average of 21.8 percent of bids were in this category), spare parts (20.7 percent), chemicals and raw materials (19.8 percent), and consumer goods (14.3 percent). These shares were highly variable from auction to auction but there appears to have been an upward trend in the share of machinery and equipment, especially starting with Auction 19 (December 25, 1993). The authorities’ actions to narrow the negative list effective September 1, 1993 (announced September 5) may have been responsible for the increase in bids for the newly eligible items, since the shares of chemicals and raw materials and building materials increased substantially during Auction 11 (September 18), as did the share of consumer goods during Auction 12 (October 2). A second narrowing of the negative list in March 1994 was not associated with comparable changes in the sectoral composition of bids.

Buying and selling rates for certain other currencies are set on the basis of the relevant rate for the U.S. dollar and the previous day’s closing rate of the currency concerned against the U.S. dollar in London. The NBE does not deal with the public; its transactions in U.S. dollars with authorized dealers take place at the secondary rate. Authorized dealers must observe the secondary rate for the U.S. dollar and the prescribed commission charges, which accrue to the NBE, of 0.5 percent buying and 1.5 percent selling; they are also authorized, but not obliged, to levy service charges for their own account of up to 0.25 percent buying and 0.75 percent selling and, for currencies other than the U.S. dollar, to include the margin charges applied by the correspondents abroad.

Reflecting movements in the secondary rate from the auction since May 1993, as well as movements in the U.S. dollar and in relative prices between Ethiopia and its trading partners, the trade-weighted real effective exchange rate of the birr appreciated by 30.2 percent from 1989/90 to 1991/92, depreciated by 41.2 percent during 1992/93, reflecting the first devaluation in the official rate in October 1992, and then movements of the secondary rate from the auction. The real effective rate continued to depreciate during 1993/94 as demand through the auction pushed up the marginal rate (in birr per U.S. dollar).

Regarding the (illegal) parallel market for foreign exchange, there is some indication that the volume of transactions has shrunk significantly, although information on the parallel market is limited. Consistent with the decline in volume, the parallel market premium over the auction rate has declined from about 50 percent in May 1993, when the auction was introduced, to about 10 percent currently.

2. Administration of control

All transactions in foreign exchange must be carried out through authorized dealers under the control of the NBE. All payments abroad require licenses issued by the Exchange Controller. All exports are licensed by the Exchange Controller to ensure the surrender of foreign exchange proceeds and shipments require permits issued by that office.

3. Prescription of currency

Outgoing payments are normally made in convertible foreign exchange appropriate to the country of the recipient or in U.S. dollars. The net proceeds of exports must be surrendered in a foreign currency that is freely convertible, or in any other acceptable foreign currency.

4. Nonresident accounts

Nonresidents may, subject to exchange control approval, open nonresident accounts either in birr or in foreign currencies at authorized banks. Deposits to these accounts can be made only in foreign exchange. Balances on nonresident foreign currency accounts may be freely transferred abroad.

A joint venture may be permitted to open foreign currency, transferable or nontransferable birr accounts to purchase raw materials, equipment, and spare parts not available in the local market.

5. Imports and import payments

Payments abroad for imports require exchange licenses. Certain imports, mostly consumer goods, may not be financed on an acceptance basis. Imports financed by suppliers’ credits require prior approval and are limited to raw and intermediate materials, pharmaceuticals, and machinery and transport equipment.

Under the franco valuta (own foreign exchange) arrangement, importers can bring in imports financed with their own sources of foreign exchange. This system was reintroduced in October 1989, and further liberalized in May 1990 when the import restriction on private vehicles was abolished. In January 1992, the Council of Ministers liberalized the regulations regarding franco valuta trade still further. A special franco valuta permit is no longer required for most imports, such as personal effects (including a maximum of one automobile every two years) and goods used to carry out an importer’s licensed activities. The new regulations specify that in cases where a permit is still required, the Ministry of Trade will issue the permit within 48 hours of application.

In August 1993, a new harmonized tariff schedule was implemented. Previously, the maximum tariff had been 230 percent and there had been 25 different tariff bands. The new tariff schedule has a maximum of 80 percent and the number of bands has been narrowed to 10. A minimum tariff of 5 percent was introduced on over 300 previously zero-rated items; the number of zero-rated items declined to 78. All but 2 specific duties were converted into ad valorem duties and the differential sales tax on imports (24 percent vs. 12 percent on domestic goods) has been eliminated (i.e., harmonized at 12 percent) with the extra tax effectively incorporated into the tariff structure. Most exemptions have been eliminated. The overall nominal weighted average statutory tariff was thus reduced from 41 percent in 1991/92 to 29 percent in 1992/93.

6. Payments for invisible transactions

There are quantitative restrictions on foreign exchange payments for invisibles (such as travel, education abroad, remittances by foreign employees in Ethiopia, and medical payments abroad). Payments for invisible transactions require exchange licenses. Invisibles connected with trade transactions are treated on the same basis as the goods to which they relate.

Regarding travel, there are different restrictions on foreign exchange availability for business travel, tourism, and government travel. For business travel by exporters and importers only, the travel allowance is currently US$120 per day for a maximum of 20 days (except under exceptional circumstances). For tourism, there is a maximum of US$50 total per year; business travel for purposes other than exporting and importing is subject to the limit applicable to tourists. For government travel, there is a schedule of rates that vary by country and city, based on cost of living. Effective July 1994, the requirement to go to the commercial bank and to the Inland Revenue Administration prior to exit (supporting evidence requirement) has been totally eliminated. However, an exit visa is still needed.

After providing for payment of local taxes, foreign companies may, in principle, remit dividends on their invested and reinvested capital in any currency. The exchange controller checks the company’s profit/loss statement on a monthly basis. Airlines are allowed to repatriate all “excess funds,” defined as operating revenue minus expenses. Share companies (i.e., companies with foreign equity participation limited to 48 percent) such as oil companies and truck dealers, are allowed to transfer only the service charges (charges for technical services provided by the home office as specified in a service agreement). Profits (dividends) cannot be remitted.

7. Exports and export proceeds

Exports of most cereals to any destination other than Djibouti are prohibited. All commodity exports require permits from the Exchange Controller and some require, in addition, the approval of specified public bodies. For exports on a c.i.f. basis, exporters must obtain full insurance from the Ethiopian Insurance Corporation. Foreign exchange receipts are to be surrendered to the NBE, generally within three months, and the export proceeds are received in an appropriate currency. Exports of hides and skins, which are regulated, are prohibited until the needs of local factories have been met.

The export licensing system has been streamlined and fees have been reduced. A general export licensing system (for commodities other than coffee and tea) was implemented in December 1992. At that time, the half-million birr minimum capital requirement for exporters was also abolished. Separate export licenses are required for coffee and tea. The number of private coffee exporters has been significantly expanded and coffee export license fees were reduced in August 1993. Additionally, a duty drawback scheme was implemented for exports in August 1993; duties paid on raw materials used in production for export are to be refunded, although the administrative modalities of this scheme are still being worked out.

8. Proceeds from invisible transactions

Foreign exchange receipts from invisible transactions must be surrendered. Travelers may bring in Br 10 in Ethiopian currency. Foreign exchange must be declared by travelers on entry, and its re-export is subject to authorization, except for temporary visitors. Reconversion of birr must be supported by documentary evidence of prior exchange of foreign currency.

9. Capital

All receipts of capital in the form of foreign exchange must be surrendered. Exchange control authorization is required, and a registration of capital inflows with the exchange control authorities establishes evidence of receipt, which is required for repatriation. All recognized and registered foreign investments may be terminated on presentation of documents regarding liquidation and payment of all taxes and other liabilities.

Under a decree promulgated in July 1989, foreign investors are permitted to hold a majority share in a joint venture, except in precious metals, public utilities, telecommunications, banking and insurance, transport, and retail trade sectors. Exemptions from income taxes are granted for up to five years for new projects, and for up to three years for major extensions to existing projects. Imports of investment goods and spare parts for joint ventures are also eligible for exemptions from customs duties and other specified import levies in the case of new projects or major expansions of existing projects.

In May 1990, a Special Decree on Investment was promulgated, which removed the restrictions on the size of activity and substantially liberalized restrictions on the types and sectors of activity for private participation. It also guaranteed the right of foreign investors to remit profits and dividends and the proceeds from the sale of assets. Under the Decree, investors in agriculture, industry, construction, and hotel services were accorded such incentives as exemption from customs duties and income tax; incentives were made more attractive for larger investments and investments in priority areas.

In May 1992, the TGE issued a detailed proclamation to encourage investment in the domestic economy designed to promote diversification of the production base, increased employment opportunities, and improved export incentives. 1/ Building on the May 1990 Special Decree on Investment, the proclamation guaranteed the right of foreign investors to remit their profits and dividends and to pay related fees, royalties, and other foreign exchange costs. Further, proceeds from asset (or share) sales relating to these foreign investments were allowed to be fully repatriated and were exempt from the payment of any taxes. However, the code reserved certain sectors (including most of the financial, energy, and rail/air transportation sectors) for investment by the Government. The code also treated domestic investors somewhat more favorably than foreign investors in these and some other areas.

Borrowing abroad requires exchange control approval and is restricted. Authorized banks may freely place their funds abroad, except on fixed-term deposits, but they may not acquire securities denominated in foreign currency without the permission of the NBE. In addition, they need prior approval of the NBE to overdraw their accounts with foreign correspondents, to borrow funds abroad, or to accept deposits in foreign currency.

10. Gold

The ownership of personal jewelry, of which gold or platinum forms a part, is permitted. Newly mined gold is sold by the Ethiopian Mineral Resources Development Corporation to the NBE. Imports and exports of gold in any form other than jewelry requires exchange licenses, except for imports and exports by or on behalf of the monetary authorities.

V. Money and Credit

1. Monetary and credit developments 1/

a. Overview

During 1988/89-1990/91, the annual growth rate of broad money increased from 8.9 percent in 1988/89 to 18.6 percent in 1990/91 (Appendix Table XLIX). The rapid growth in net domestic assets (NDA) was the major factor affecting the money supply; NDA movements mainly reflected variations in domestic credit to the Government, which depended on central bank financing for its budget deficits. In contrast, net foreign assets, which had once reached Br 576.4 million (US$ 278.5 million) at end-June 1986, fell sharply and became negative in 1989/90 before rising to Br 270.7 million in 1990/91. Interest rates were unchanged during the period.

From its inception, the TGE sought to adopt disciplined fiscal and monetary policies; as a result, broad money growth declined to 13.3 percent in 1991/92, and 14.0 percent in 1993/94; broad money growth increased to 16.2 percent in 1992/93, but this was a year when the currency was devaluated 58.6 percent in U.S. dollar terms. Compared with pre-TGE periods, the principal factor affecting the money supply shifted to movements in net foreign assets, reflecting the 1992 devaluation and the mounting inflows of foreign aid from the international community (Appendix Table L). In the meantime, the growth of net credit to the Government has slowed from its recent peak of 29.3 percent in 1989/90, declining to 10.6 percent in 1992/93, and further to 6.3 percent in 1993/94.

The velocity of circulation of broad money declined steadily during the period 1988/89-1991/92, from 2.97 in 1988/89 to 2.27 in 1991/92. It rose slightly to 2.49 in 1992/93, before declining again to 2.32 in 1993/94. 2/ There was still an increase in velocity in the period 1991/92-1992/93, even when the change in birr currency notes in Eritrea during this period is excluded. The share of quasi-money (time and savings deposits of the CBE) in broad money decreased from 26.8 percent in 1988/89 to 22.8 percent in 1990/91, but subsequently increased to 26.5 percent in 1992/93 and 28.8 percent in 1993/94. It is likely that some of this increase reflected a movement of funds from non- interest-bearing narrow money to interest-earning deposits, given the large revision in the interest rate structure in October 1992.

Through the periods of the previous Government and the first year of the TGE (1991/92), overall domestic credit grew basically in line with developments in net credit to the Government. While domestic financing of the government deficit (including grants) rose from 4.3 percent of GDP in 1988/89 to 9.6 percent in 1989/90 and 6.2 percent in 1990/91, overall credit growth rose from 7.7 percent in 1988/89 to 17.1 percent in 1989/90, and then dropped to 11.9 percent in 1990/91. Domestic credit growth to the nongovernment sector, in sharp contrast, decelerated from 20.0 percent in 1987/88 to 1.2 percent in 1989/90 and turned negative, to minus 1.2 percent in 1990/91.

With the commencement of the structural adjustment efforts, the TGE began to reduce its deficit financing through the domestic banking system, with a small reduction in 1991/92 (to 5.7 percent of GDP), and more significant reductions in 1992/93 and 1993/94 (3.2 percent of GDP and 1.9 percent, respectively). Since 1992/93, the Government has attempted to ensure that overall domestic credit growth is consistent with the overall macroeconomic policy framework, rather than simply being driven by the pace of government credit expansion. Thus, the reduction in government credit expansion has been offset by an increase in credit to the nongovernment sectors (public enterprises, cooperatives, and private and individuals).

As briefly mentioned above, NFA became the major determinant of movements in broad money, as the TGE’s economic reform efforts gathered pace (including the impact of the devaluation) and as the inflow of foreign aid increased. During 1988/89-1990/91, the last three years of the pre-TGE period, NFA had contributed an annual average of 1.5 percent to broad money growth. Thereafter, NFA’s contribution increased from 12.9 percent in 1991/92 to 86.5 percent in 1992/93, and further to 132.8 percent in 1993/94.

It should be noted, however, that a literal interpretation of these figures would lead to an over estimation of the role of NFA, since some of the factors that explain this large contribution in the latter years had, in effective terms, no monetary consequences. Specifically, the effects of the devaluation, from 1992/93 onward, show up both as an increase in NFA as well as an offsetting increase in the valuation account component of “other items net” (OIN). Also, certain foreign-aid-related funds, which are of a “blocked nature,” were included in both NFA and OIN at the time that their ownership was transferred from the donor institution to the TGE (see below). 1/

In effect, these NFA-related developments in OIN (in terms of the accounting of the monetary survey) simply result in offsetting developments in NDA, thus masking to some extent the TGE’s efforts to enhance credit to the nongovernment sectors. The NDA’s contribution in broad money developments declined as that of NFA increased and turned negative in 1993/94.

During the period 1988/89-1990/91, interest rates were effectively negative in real terms. In October 1992, the Government raised the overall level of interest rates, with a view to achieving positive real interest rates. The spread between deposit and lending rates was compressed somewhat, and the structure of rates was changed in order to eliminate all differentials by ownership and most of the differentials across sectors. Among sectors, agriculture, construction, and housing were treated preferentially by setting the lending rates slightly lower than for other sectors, in line with the Government’s development priorities.

b. Credit operations of the monetary system

A major shift in the composition of the loan portfolio of the banking system (including the NBE) took place over the period 1991/92-1993/94 (Appendix Table LI). As of end-June 1991, about two thirds of the loans and advances 2/ of the NBE and the CBE were to the Government. A fifth of the portfolio represented advances to the AIDB, 3/ while public enterprises and private sector enterprises (including individuals) accounted for only 7 percent and 4 percent, respectively. Taking account of the fact that loans to the AIDB were virtually nonperforming, only about 11 percent of total loans were effectively allocated to the nongovernment sector. During 1991/92-1993/94, credits were increasingly allocated toward the private sector. Of the total net increase in loans and advances, the Government’s share declined from 86.2 percent in 1991/92 to 45.5 percent in 1992/93 and further to 21.3 percent through end-March in 1993/94. 4/ Conversely, the private sector’s share increased from 1.5 percent to 25.3 percent and to 53.7 percent in the same period. The share of public enterprises rose from 10.1 percent in 1991/92 to 22-24 percent in 1992/93-1993/94. As a result, at end-March 1994, the distribution of the loan portfolio was as follows: 63.6 percent to the Government, 14.5 percent as advances to the AIDB, 10.9 percent to public enterprises, and 10.0 percent to the private sector (of the remainder, loans to cooperatives and to the HSB were 0.3 percent and 0.7 percent, respectively).

In terms of the sectoral allocation of credit, there has also been a significant change (Appendix Table LII). 1/ As of end-June 1991, the shares of loans and advances among the major sectors were: exports (19.0 percent), construction (18.2 percent), industry (17.8 percent), imports (16.5 percent), and domestic trade (12.2 percent). Agriculture accounted for only 4.4 percent of total nongovernment credit. Most credit in recent years has flowed toward domestic trade, imports, exports, and the industrial sector. This reflects a number of factors, including the increased credit demand for raw materials and spare parts during the economic recovery, the cost increase induced by devaluation, the introduction of the foreign exchange auction, the improvement in the capacity utilization in the manufacturing sector, the relatively higher price increases for manufactured products, which encouraged potential borrowers in the industrial sector, and the gradual but steady shift from government to private borrowers.

As a result, at end-March 1994, the share of domestic trade significantly increased (to 25.5 percent) at the expense principally of loans to the construction and the agricultural sectors (which declined to about 8 and 2 percent, respectively). The shares of the other sectors remain broadly unchanged.

The efforts made by the Government in recent years to encourage private sector participation in the economy have also resulted in a shift of credit from public enterprises to the private sector. Focusing on credit allocation within the nongovernment sector (excluding the NBE loan to the AIDB), the share of public enterprises in total nongovernment credit declined from 63.6 percent to 49.9 percent, while the share of the private sector increased from 34.2 percent to 45.7 percent between end-June 1991 and March 1994. Credit to public enterprises continued to predominate in those sectors where the private sector was slow to develop.

Despite its dominant role in the economy, loans and advances to agriculture declined between 1988/89-1992/93, picking up only in 1993/94. As the Government drastically curtailed its involvement in the sector (with the reduced role of the state farms), credit to the state farms dropped sharply. While credit to the cooperatives and the private sector experienced some increase, their absorptive capacity for borrowing (including the lack of an adequate legal framework for the establishment of peasant cooperatives) remained limited. In addition, through 1993, both the private sector and cooperatives largely relied on their own funds for investment financing. Credit demand was also weak, given the relatively small number of investors entering into the commercial agricultural sector and the weak realization rate of private agricultural projects. The small increase in agricultural product prices, particularly food prices, relative to output prices for other sectors, reduced the rate of return to investment in agricultural projects. The collapse of the AIDB also meant the absence of an institutional vehicle for agricultural lending. Furthermore, the absence of collateral and clear legal rights to land represented obstacles to the CBE playing a greater financial role, and many peasants and their peasant association defaulted on previous fertilizer loans.

Construction and housing also experienced a decline in their share of total nongovernment credits. Outstanding loans and advances to this sector constituted 15-18 percent of the total during 1988/89-1991/92, but declined thereafter, to less than 8 percent as of end-March 1994. Loans to the sector were restrained owing to: (i) a contraction in demand resulting directly from higher interest rates since late 1992, and indirectly, as a consequence of the significant increase in the cost of acquiring land; (ii) the closure of the Building Construction Authority enterprise; (iii) the reduction in the number of new government projects; and (iv) the slow project implementation rate in the budget.

c. Domestic liabilities of the monetary system

During 1989/90-1991/92, broad money grew at an annual average rate of 16.5 percent, which was substantially higher than the annual nominal GDP growth rate of 8.5 percent (Appendix Table XLIX). Despite the negative rate of interest on deposits (which averaged minus 9.7 percent based on the Addis Ababa Retail Price Index (AARPI) 1/), quasi-money (interest-earning deposits) increased by an average of 12.3 percent, reflecting forced saving by the private sector owing to policies that restricted private expenditure. In 1992/93, the real interest rate on deposits increased to zero percent and, in 1993/94, to more than 5 percent; quasi-money increased by 28.5 percent and 24.2 percent, respectively, during these two years. The high growth in quasi-money was caused by a shift within households’ portfolios from real to financial assets, from non-interest -earning currency and demand deposits to interest-earning time and saving deposits. Between 1991/92 and 1993/94, the share of quasi-money in broad money increased from 23.9 percent to 28.8 percent.

Conversely, currency in circulation (cash), which increased 48.1 percent in 1990/91 as a share of broad money, declined steadily in importance, to 43.3 percent by 1993/94. The cash-dominant nature of the economy as well as the moderate response of cash demand to interest rate changes can be explained by several factors: (i) the primitive stage of development of payment systems such as time-consuming procedures for transferring and withdrawing money from the banks, the lack of networks, which makes it difficult for people to reach banks, and the high associated transactions costs of depositing funds at the banks; (ii) uncertainties regarding the political situation; (iii) a lack of information regarding banking services; and (iv) a lingering suspicion of banks in terms of their maintaining the confidentiality of an individual’s financial information vis-à-vis the Government.

Despite the ending of the discrimination in deposit rates, interest-earning deposits of public enterprises and cooperatives have shown little movement (Appendix Table LIII). During 1989/90-1991/92, savings and time deposits by public enterprises and cooperatives decreased from Br 23.9 million (0.7 percent of total deposits) at end-June 1989 to Br 19.5 million (0.4 percent) at end-June 1992, and then increased marginally to Br 28.1 million (0.4 percent) at end-March 1994. The latter increase can be explained partly by the growth of demand for liquid assets to cover the cost-push effect of devaluation, the elimination of budgetary support for certain enterprises, and the increased wage bill, which was financed from own resources.

In effect, as of March 1994, the private sector was virtually the sole depositor in interest-earning deposits, its share accounting for more than 99 percent of savings deposit and more than 96 percent of time deposits. The interest-rate responsiveness of short-term saving deposits was larger than that of time deposits in absolute terms, while the rate of change was smaller, owing to the difference in the size of both types of deposits (as of March 1994, private saving deposits stood at Br 2,753.0 million while time deposits were only Br 349.6 million). 1/ The large magnitude of saving deposits is attributable to the low speculative demand for money of households rooted in low income levels.

d. Specialized banks

Loans constituted about 90 percent of the total assets of specialized banks and almost all of these loans were to the nongovernment sector. During 1989/90-1991/92, specialized banks’ lending to the nongovernment sector grew at an annual average rate of 5.4 percent; it grew 6.0 percent in 1992/93, and 7.6 percent as of end-December 1993. The AIDB has been the main vehicle for medium- and long-term lending and also deals with short-term lending to the agricultural sector. Agricultural credit, particularly to state farms, accounted for 76.2 percent of the AIDB’s total lending at end-December 1993, although most of those loans were nonperforming. The Housing and Savings Bank (HSB) extends loans for residential and commercial buildings to state enterprises, housing cooperatives, and private individuals. While the majority of loans in terms of the number of commitments are for residential buildings, the majority of loans in terms of loan amounts are for commercial buildings.

On the liability side of specialized banks’ balance sheets, as of end-December 1993, the major items were: borrowing from the NBE (57.0 percent of total liabilities), time and savings deposits (14.3 percent), and foreign loans--mainly from the IDA, African Development bank, and bilateral donors (7.9 percent). For the HSB, most time deposits were from the CBE, EIC and the PSSA, which carried a 1 percent interest rate.

2. Financial system

Through the 1970s and 1980s, the financial sector of Ethiopia mirrored the centrally planned nature of the economy. Presently the financial sector comprises: (i) the central bank (NBE); (ii) a commercial bank (CBE) with 154 branches; (iii) two specialized banks--the AIDB, specializing in medium- and long-term credits (mainly to agriculture), and the HSB, which services the housing and construction industry; (iv) the Ethiopian Insurance Corporation (EIC), providing insurance services; (v) the Pension and Social Security Authority (PSSA); (vi) some 400 savings and credit cooperatives (SACC) 1/; and (vii) about 4000 service cooperatives (SC) 2/.

With the exception of the SACCs and the SCs, all of these institutions are owned by the Government. The banking sector is dominated by the NBE and CBE. Since the number of branches differs substantially between the CBE and other banks, a scheme was set up such that the CBE takes care of the disbursement of loans on behalf of the HSB, with a commission paid to the CBE in areas where HSB branches do not exist. SACCs grant loans to members who might not qualify for credit elsewhere and who cannot afford the high loan rates offered in either the informal sector or by the other financial institutions.

In a country where more than 80 percent of the population live in rural areas, operational financial institutions are, in general, limited to urban areas. Nationally, per capita deposits, measured by deposits (time and savings deposits) of the CBE, are estimated at below 65 birr in 1993/94.

3. Monetary policy instruments

a. Credit allocation

The NBE regulates and monitors loans provided by the banking system, and determines which activities are to be financed and under what conditions, particularly with regard to credit terms, maturities, and collateral. Each bank provides credit to the specific economic sectors in which the bank specializes, and all transactions that exceed certain amounts have to be reported to the NBE for approval 1/. The CBE provides mostly short-term credits for working capital. The AIDB is the main source of credit for cooperatives and state farms in the agricultural sector and public and private enterprises in the other sectors. The HSB concentrates on activities related to building, including private dwellings.

Within the banking system, loans are, in principle, disbursed against physical collateral, such as vehicles or project plans (in the case of new projects). The formal collateral requirements of the banking system, viz., the ratio at the value of the collateral (estimated by the registered price of the collateralized good) to the amount to be borrowed, was revised in December 1992, when the ratio was reduced from 250 percent to 100 percent in the case of the CBE (the AIDB maintained its collateral requirement ratio of 125 percent). A joint working group (comprising of the NBE, CBE, and the Ministry of Industry (MOI)) has also been reviewing the formalities of the registration scheme and a possible broadening in the range of eligible forms of collateral (currently limited to only registered vehicles). Pending this review, agricultural machinery such as tractors, combines, and harvesters, which used to be rejected as collateral, have become acceptable as collateral as they have been registered by local regions. In addition, efforts have also been made to make the collateral requirements operationally more flexible for loans related to fertilizer imports, coffee exports, and transport vehicles (where no collateral is being required), and to differentiate the collateral policy according to customers’ credibility ratings.

In providing credit to the Government (through direct advances or the acquisition of Treasury bills and bonds), the recently issued Monetary and Banking Proclamation (Proclamation No. 84/1994, January 30, 1994) stipulates limits on the Government’s borrowing ability from the NBE and commercial banks based on the average annual ordinary revenue (AAOR) for the three preceding fiscal years for which accounts are available. Specifically, the outstanding stock of (i) direct advances (overdraft facility) from the NBE are not to exceed 15 percent of AAOR; (ii) treasury bills purchased by the NBE and other banks are not to exceed 25 percent of AAOR; and (iii) government bonds purchased by the NBE and other banks are not to exceed 50 percent of AAOR. 1/ Regional governments are required to submit any request for bank financing to either the Ministry of Finance or the Ministry of Planning and Economic Developments, along with a feasibility study for any project requiring financing as well as the full regional budget and revenue forecast. Approval of the relevant ministries would be required, taking account of the overall financing constraints on the overall borrowing by the Government. Commercial banks are not allowed to extend direct advances or other credit facilities, whether directly or indirectly, to the Central Government and the Regions.

In the agricultural sector, loans were provided only to service and producers’ cooperatives and state farms in the pre-TGE period. Since 1991/92, the latter two institutions have either been dismantled or become insolvent. Newly organized cooperative societies (operating under a recently passed Proclamation) are able to receive deposits from and extend loans to their members. Nevertheless, a large majority of the rural population have little or no access to official credits, and are forced to rely on the curb market for loans. Despite wide recognition of the existence and importance of the informal market in the economy, relatively little is known about its working and structure.

Based on available information, two of the most important informal institutions are the iqqub and the iddir. The iqqub is a saving association based on pre-established homogeneous groups, where each member agrees to pay a small sum periodically into a common pool so that each member can, at a given point in time, borrow the full amount in the pool; when a loan is repaid, another member of the iqqub is then able to borrow (there is usually a specified sequencing of borrowers). Iqqub loans are lent at zero nominal interest rate. The iqqub is generally used for the purchase of consumer durables, payment of school fees, clothing, and the like. Depending on the size and location, some iqqqub have linkages with a bank. The iddir is a non-profit making institution based on solidarity, friendship, and mutual assistance among members. It runs an insurance program to meet emergency situations. The risks covered by iddirs include funeral expenses, financial assistance to families of the deceased, medical expenses, and losses due to fire or theft. Members have to pay certain amount to a common fund. Most of the common funds are deposited in a bank.

b. Reserve and liquidity requirements

The CBE is required to maintain, at the end of each week, 1/ at least 5 percent of its total net deposit liabilities (defined as the sum of demand deposits, saving deposits, and time deposits, less uncleared checks paid and uncleared foreign effects) in a nonremunerated reserve account with the NBE and to hold liquidity (defined as the sum of cash, reserve deposits at the NBE, demand balances with other banks, and government securities maturing within 370 days) of at least 20 percent of their total net current deposits (defined as net deposits less deposits payable after 30 days of notice). 2/ An NBE committee is currently working to establish a new ratio on the basis of Proclamation No. 84/1994. The CBE’s reserve and liquidity holdings at the NBE have consistently exceeded the legally required levels, owing both to the low levels of economic activity in the past and the binding effect of the credit ceiling recently imposed by the NBE in connection with the macroeconomic adjustment program. 3/

c. Refinancing facilities

The NBE provides short- and long-term refinancing facilities to banks and other financial institutions. Owing to the considerable excess liquidity in these institutions, this refinancing facility has not been used since the mid-1970s for the CBE, while the bulk of the loans extended to the AIDB were converted to government bonds in March 1994 (see below).

d. Interest rates

In July 1986, the lending rates of the CBE and specialized banks were reduced in order to encourage investment. Rates were differentiated in order to favor certain sectors, including the state-owned enterprise sector. Time and savings rates were raised to encourage long-term deposits from small savers. These rates remained unchanged until October, 1992, when they were adjusted upward in order to realize positive real interest rates and eliminate differentials according to the type of ownership (Appendix Table LVIII). As of end-June 1994, sectoral differences in lending rates were narrowed but not fully liberalized, in order to make the changes more acceptable to the public and to provide some preferential borrowing rates for priority sectors.

Since October 1992, the lending rate to the Government from the financial sector has been 12 percent (revised upward from 3 percent) for short term loans and 13 percent (revised upward from 5 percent) for loans with maturity of more than over five years. Differentiation of rates by ownership 1/ have been eliminated. In lending to other sectors of the economy, the preferential rates for housing construction, agriculture, exports, and imports of agricultural inputs (ranging from 4.5 to 7 percent) have been significantly reduced. As of end-June 1994, lending rates ranged between 10 percent and 14 percent according to sector. The term structure of lending rates to the nongovernment sector is slightly upward sloping, instead of being flat, as in the previous structure.

With regard to time deposits for individuals, savings and credit cooperatives, and self-help organizations, the term structure of interest rates is upward sloping, allowing greater returns to longer term deposits; interest rates on time deposits with a maturity of one year or longer range from 11.5 percent (revised from 6 percent) for one year to 12.0 percent (revised from 7.5 percent) for five years. Time deposits with a maturity of less than one year earn interest either 10.5 percent or 11.0 percent. Effective July 1989, savings deposits cannot be held by financial institutions and government-owned enterprises. Interest rates on savings deposits for individuals, savings and credit cooperatives, and private and self-help organizations have been revised upward to 10 (from 6) percent.

e. Payment system

Since there is only one commercial bank and neither a clearing house nor an interbank market exists in Ethiopia, the present payments system in Ethiopia is quite rudimentary. Intrabranch settlements are, in principle, conducted through branch accounts held at the head office of the CBE, except for some regional branches, which are given some autonomy to settle among branches in the same region. Money transfers are conducted by telegram, mail, or the issuance of draft (checks), followed by the sending of a credit summary as an official payment instruction in the case of the former two methods. A payment instruction is issued for all transactions (gross basis) and sent to the head office via mail, and the head office nets out the transactions and feeds them back to each branch with the results. The normal time lag between an actual transaction and its final settlement is about ten days. Interbank settlements (settlements among the CBE, HSB, and AIDB) are all conducted using the CBE account. The branches of the HSB and the AIDB open current accounts at one of the branches of the CBE and settle their transactions through certified payment order in the case of the Government or by cash payment orders (or drafts) in the case of the private sector. There are no cable transactions between the HSB and the AIDB. With respect to settlements between the CBE and the NBE, a reserve account held at the NBE is used, and both banks exchange daily a form of document that confirms the transactions that occurred during the previous day (i.e., a voucher system).

4. Financial sector reform

Since mid-1991, the TGE has made efforts to move progressively to a stable market-based economy, with the banking system expected to play a key role in channeling deposits efficiently to productive investments on the basis of economic considerations. Within this context, the NBE has sought to create an environment conducive to the establishment of an efficient and competitive financial sector, supported by an effective regulatory and prudential supervision system.

a. Restructuring and recapitalization of the AIDB

The AIDB was established to extend loans to the agricultural and industrial sectors. The AIDB drew its funds mainly by borrowing from the NBE, and loans were highly concentrated in the state farm sector with no collateral required. Owing to the serious operating inefficiencies of the state farms (and their consequent inability to repay loans), and the subsequent dissolution of many state farms since 1988/89, more than 90 percent of the AIDB’s loans to some 90 state farms (over Br 2 billion including principal and interest) were nonperforming. These losses destroyed the profitability of the AIDB and fully eroded its capital base.

To restructure the AIDB’s financial position, a proclamation was issued in December 1993, which allowed for the transfer of AIDB’s debt claims from the NBE’s books to those of the Government. A special interest-free Government bond with a maturity of ten years, with equal annual repayments 1/, was issued in March 1994 by the Government in the amount of Br 1.3 billion to the NBE in exchange for the Government’s assuming the NBE’s claims on the AIDB. A similar, second special bond of Br 85 million was issued and exchanged for the AIDB’s remaining debt claims on state farm loans. In effect, combined with the drawdown of the NBE’s provisions for bad debt, the nonperforming loans for state farms financed by the NBE and AIDB’s own resources have thus been written off, and an independent trust fund established for the purpose of managing the sale or liquidation of state farms in default and for the application of proceeds to compensate the Government in its repayment of the special bonds.

With respect to the recapitalization of the AIDB, a proclamation is now awaiting gazetting, which will allow the AIDB to increase its capital to the level of Br 250 million (from a current loss position of negative Br 131.5 million). 2/

b. New legislation

The Monetary and Banking Proclamation (No. 83/1994) was issued on January 30, 1994 and redefines the status, functions, and authority of the NBE under a free market economic environment. The legislation was formulated so as to provide for a stronger and more autonomous role for the NBE. The legislation states the NBE’s relations with the Government and other financial institutions through its supervision of credit, the setting of foreign exchange rates, and the framework of interest rate policy. The legislation stipulates that 20 percent of the net profit of the NBE is to be allocated to the General Reserve Fund, while 80 percent is to be credited to the account of the Ministry of Finance.

As noted above, the Proclamation “Licensing and Supervision of Banking Business” (No. 84/1994) was issued on January 31, 1994, and allows for the establishment and supervision of private domestic commercial banks (with equity participation limited to investors of Ethiopian nationality). It states the requirements for opening a commercial bank, including a minimum capital requirement 2/ as well as the stipulation that banks can only be established on a limited share basis, 3/ in order to protect a bank from the dangers that could arise from minority rule and decisions. The legislation also specifies that a minimum liquid assets position and reserve balance must be maintained and that the NBE is the sole organ allowed to issue a license to a financial institution. Upon receipt of an application for license, the NBE is required to decide on its acceptability within 90 days. No bank can pay dividends unless it has fully covered all of its capital expenses; at the end of each financial year, a bank is required to transfer at least 25 percent of its net profits to a Legal Reserve Account until the latter account reaches the required minimum capital amount. Subsequently, the NBE issued a directive in May 1994 that established criteria for the selection of the members of the Board of Directors of a private bank, fee rates, the content of application forms, and other relevant information that has to be submitted along with the application.

Finally, the “Licensing and Supervision of Insurance Business Proclamation” (No. 86/1994), which was issued on February 1, 1994, established the main requirements for establishing new insurance companies, including the minimum paid-up cash capital requirement (Br 3 million for the general insurance business, Br 4 million for the life (long-term) insurance business, and Br 7 million if both services are given). If an insurance company is established by a share company, each shareholder’s investment is limited to no more than 20 percent of the total equity. The law also specifies the requirements for the issuance of a license, the application for license, share registration, required capital, reserves, and solvency margin. This proclamation was followed by a Directive (No. SIB/1/94) in June 1994, which set the criteria for the selection of members of the Board of Directors of a proposed insurance company, as well as the fees, contents of the application form, and other relevant information that have to be submitted along with the application.

VI. Assessing Eritrea’s Impact on Ethiopia’s Program

1. Overview

One of the difficulties that emerges in the analysis and programming of monetary developments in Ethiopia is the statistical problem of separating out movements in the money stock between Eritrea and Ethiopia, given their de facto currency union. The issue arises principally because Ethiopia’s monetary survey, as currently compiled, does not explicitly distinguish between currency notes that are in Ethiopia proper rather than in Eritrea. This implies that movements of the overall monetary aggregate for Ethiopia may not yield an accurate measure of the actual development of liquidity within Ethiopia, given the possibility of bilateral payment surpluses or deficits with Eritrea. This chapter suggests several conceptual adjustments that would allow for a more meaningful interpretation of developments in Ethiopia’s monetary aggregates (although these tasks are complicated by limitations of the available data base).

2. Possible adjustments

The level of broad money, as presently reported in Ethiopia’s monetary survey, is larger than the actual level of broad money in Ethiopia. Specifically, the currency in circulation component of the survey not only includes currency in circulation in Ethiopia, but also total currency holdings in Eritrea. An attempt to reduce broad money by the amount of currency in Eritrea would require a corresponding adjustment in Ethiopia’s net domestic assets (NDA) classification, which measures the net claims of Ethiopia’s banking system on Ethiopian residents. Some of Ethiopia’s NDA reflect claims on Eritrean residents that gave rise to their birr deposits as well as currency-in-circulation in Eritrea. Although a precise adjustment of individual elements of the Ethiopian monetary survey is not possible, some broad accounting adjustments may be undertaken at a conceptual level.

On the liability side, Ethiopia’s currency in circulation would need to be adjusted for the amount of total currency notes in Eritrea, i.e., currency in circulation in Eritrea as well as the vault cash of the Eritrean banking system. On the asset side, counterpart adjustments would be made by subtracting these items from Ethiopia’s net domestic assets. 1/ In practice, the amount of total currency in circulation in Eritrea is not known and needs to be estimated. 2/ However, Eritrea’s provisional monetary survey treats (net) claims on the Ethiopian banking system as part of Eritrea’s net foreign assets, and thus this component, i.e., Eritrea’s birr-denominated net foreign assets (NFABR), is directly observable from the current presentation of the Eritrean accounts. 3/ Ex post adjustments to the Ethiopian accounts can therefore be undertaken on the basis of figures observed in the Eritrean accounts, even though the Ethiopian banking system does not keep a separate record of its liabilities to and claims on Eritrea (i.e., Eritreans are effectively treated as residents rather than nonresidents). As noted above, however, the distribution of currency in circulation between the two countries would still need to be estimated (as of the beginning of each period).

In essence, these two components capture intra-union monetary flows, i.e., the bilateral balance of payments between Ethiopia and Eritrea, and are largely driven by the relative stance of macroeconomic policies pursued in each country. While a change in cash in circulation in Eritrea reflects, inter alia, a net migration of currency that remains outside banks, changes in the (net) claims of the Eritrean banking system on Ethiopia provide an indicator of how much the former accumulates (or loses) reserves and liquidity vis-à-vis the latter (and vice versa).

An important conceptual distinction should be made. As in any standard monetary survey, the factors affecting the area-wide money supply include rates of (net) domestic credit expansion in each country as well as the balance of payments position vis-à-vis the rest of the world. Therefore, as long as accumulated claims between the two countries remain as reserves in the (corresponding) banking system, area-wide broad money should not be affected. 1/ In that case, intra-union monetary flows would imply a lower level of broad money for the country running the deficit and an adjustment to that effect would need to be made. This broadly reflected the situation for Ethiopia as of June 1993, when Eritrea held a sizable surplus in its birr-denominated NFA position. 2/

3. Numerical illustrations

On the basis of the preceding discussion, Eritrea’s monetary accounts can be integrated within an overall analysis of monetary developments in Ethiopia during 1992/93 (June-June), in order to illustrate the possible impact of Eritrea on monetary developments in Ethiopia. Two scenarios are indicated in Table 4. The first two columns (A and B) indicate the unadjusted monetary survey for Ethiopia and Eritrea; Column C indicates the adjustments required to produce a more economically meaningful monetary survey for the birr area as of June 1992. In making this adjustment, it is necessary to make an assumption on the amount of currency in circulation in Eritrea; column C assumes that 7 percent of total currency in circulation is held outside the Eritrean banking system on June 1992, in line with the relative population shares of the two countries. Ethiopia’s adjusted (effective) NDA is then derived by subtracting NFABR and cash in circulation in Eritrea from the unadjusted (reported) NDA.

Table 4.

Ethiopia: Adjusted Monetary Survey, 1991/92–1992/93

(In millions of birr)

article image

As reported in the Eritrean monetary survey.

Estimate.

Column D illustrates the impact of two additional assumptions on developments during 1992/93, viz., that (i) 10 percent of the total increase in NFABR during this year (of about Br 524 million) was due to the return to the banking system of currency already in Eritrea (hence the Br 50 million decline in Eritrea’s cash in circulation; and that (ii) the remaining increase in the Eritrean NFABR was derived from new flows of cash from Ethiopia, implying a reduction in the latter’s cash in circulation. 1/ Such a flow of notes from Ethiopia, equivalent to its running a bilateral balance of payments deficit with Eritrea, could reflect a bilateral current account deficit and/or net positive capital flows to Eritrea.

An immediate implication of the adjustments is that adjusted NDA in Ethiopia is reduced to Br 7,542 million and Br 7,267 million in June 1992 and June 1993, respectively, from the unadjusted amounts of Br 8,585 million and Br 8,782 million. As a consequence, Ethiopia’s adjusted broad money records a much lower growth rate of 12.4 percent, compared with an unadjusted growth of 16.2 percent. In Eritrea, broad money declines from an unadjusted 99 percent to an adjusted 66 percent, reflecting that the adjusted definition includes an additional component (i.e., Eritrea’s currency in circulation). On the other hand, area-wide broad money grows by 18.8 percent, as compared to Ethiopia’s unadjusted 16.2 percent.

In column E, a similar exercise is carried out by assuming that none of the increase in NFABR in 1992/93 is derived from a return to the banking system of currency in circulation from within Eritrea. As would be expected, Ethiopia’s money supply growth is correspondingly lower (11.7 percent) while Eritrea’s is higher (71.1 percent). This simply reflects the assumption that all money is now assumed to come from Ethiopia while currency in circulation in Eritrea remains unchanged over the period. Area-wide growth in broad money is unaffected by this compositional movement.

These variants demonstrate that without knowledge of financial developments within Eritrea and its balance of payments with Ethiopia, it would be difficult to assess how tight monetary policy was in Ethiopia during 1992/93. The variants also indicate that overall area-wide broad money growth is not highly sensitive to changes in cash in circulation in Eritrea and NFABR, since neither NDA nor NFA (denominated in convertible currencies) is assumed to change in any of the alternative variants.

VII. Estimation of Demand Curve for Foreign Exchange Auctions

The introduction of a foreign exchange auction system in Ethiopia since May 1, 1993 has generated a rich data base that can be used to estimate a demand function for foreign exchange. As of June 25, 1994, 31 auctions had taken place, generating some 1,000 observations pooled in a cross section-time series data base consisting of the bid price and the corresponding cumulative demand at each auction. Data were also available on the number of eligible bidders and the excess demand generated in each auction. Various specifications of the demand function were considered in estimating a demand equation. On the basis of the simulation of Auction 31 (discussed here) and earlier auctions, the model would appear to have relatively high predictive power.

1. Specification of equations

The model specification not only reflects the usual price quantity relationships, but also an attempt to capture the lagged response of bidders to any excess demand pressures revealed in the previous auctions. In addition, an aspect of the game theoretic behavior of bidders may be captured by testing their response to the marginal bid which cleared the market at the previous auction; in effect, this rate gives bidders a sense of which bid rates are likely to be successful in the current auction. The model specification is as follows:

( 1 ) B R i t = α β ( C D i t ) + γ ( M R t 1 ) + σ ( E D t 1 ) + ϵ t
( 2 ) BR it MR t-1 = λ ω ( CD it ) + δ ( ED t 1 ) + z t

Where:

BRit - bid rate i for auction t;

CDit - cumulative demand i for auction t;

MRt-i - marginal rate for the previous auction; and

EDt-1 - excess demand for the previous auction;

εt, zt - error terms

Equation (2) implicitly constrains the coefficient of MRt-1. to be unity and therefore considers bidders’ behavior in terms of the difference between their bid rates and the previous marginal rate. In general, this equation did not provide a good fit, with R2 less than 0.5.

The number of eligible bidders for the previous auction, EBt-1, was included as an additional independent variable, but was found not to improve the fit of the equation significantly. A possible reason could be that the two variables EDt-1 and EBt-1 are strongly correlated with each other. As a consequence, the variable EBt-1 was dropped from the equation.

A semi-logarithmic form of the model 1/ was run to see if this specification would be able to capture any nonlinear relationship between the explanatory variable (the bid rate) and the set of independent variables. The fit improved slightly with the semilogarithmic specification.

2. Structural issues

Two main developments occurred over the auction periods, which might have led to a structural shift in the demand curve. First, as of the fifth auction, the quantity to be auctioned was preannounced. To capture the possible effect of this development on the demand for foreign exchange, a dummy variable, ID, was included. 1/ The second development was the narrowing of the import negative list as of September 4, 1993 (Auction 10) and a further narrowing of the list as of March 5, 1994 (Auction 23). Two dummy variables, NL1 and NL2, were used to capture any possible effects on the demand function 2/.

Various combinations of the three dummy variables were specified in the model to find the best fit. The specification that best captures the two developments appears to be that with the two slope dummy variables ID and NL2. Including NL1 as well did not improve the fit any further because both NL1 and NL2 were essentially capturing similar effects. Indeed, NL2 would be a better variable to include as the earlier narrowing of the negative list was not adequately publicized.

3. Estimation results and simulations

Table 1 presents the results of the most significant equations estimated. Equations Ml and M2 were estimated on the basis of the first 30 auctions, and were used for the purpose of simulating the results of Auction 31, of June 25, 1994. Equations M3 and M4 include the results of Auction 31. All equations have good R2 with the semilogarithmic specification providing a slightly better fit. The t-statistic and F-statistics are all significant. The demand curve appears to be relatively stable, as the coefficients change little when the Auction 31 data are added. The simulation of Auction 31 is shown in Table 6, together with a comparison of the actual bid rates. The predictive value of the equations is relatively strong, except for the extremely low bid rates. In particular, the prediction of the marginal rate by both equations differs from the actual marginal bid rate (6.22) by less than 1 percent.

In simulations of earlier auctions, the predicted values have in general fallen within 3 percent of the actual marginal rates. The percentage errors have tended to be smaller over periods where the demand for foreign exchange has remained relatively stable.

Table 5.

Ethiopia: Demand Estimates for Foreign Exchange

article image
Notes: t values in parenthesis CDt = cumulative demand EDt-1 = excess demand lagged one auction MRt-1 = marginal bid rate lagged one auction ID.CDt = pre-announcement of auction quantity dummy NL2.CDt = narrowing of import negative list dummy
Table 6.

Ethiopia: Prediction of Auction 31 Bid Rates (Birr)

article image

APPENDIX

Table I.

Ethiopia: Growth Rate of Sectoral GDP at Constant 1980/81 Factor Cost, 1988/89–1993/94

(Annual growth rate in percent)

article image
Source: Staff estimates based on discussions with the Ethiopian authorities on the revised GDP estimates soon to be finalized. Data exclude Eritrea after 1990/91.
Table II.

Ethiopia: Gross Domestic Product by Sector at Current Factor Cost, 1988/89–1993/94

article image
Source: Staff estimates based on discussions with the Ethiopian authorities on the revised GDP estimates soon to be finalized. Data exclude Eritrea after 1990/91.
Table III.

Ethiopia: Expenditure on Gross Domestic Product at Current Market Prices, 1988/89–1993/94

article image
Source: Staff estimates based on discussions with the Ethiopian authorities on the revised GDP estimates soon to be finalized. Data exclude Eritrea after 1990/91.

Includes changes in stocks.

Table IV.

Ethiopia: Gross Domestic Product by Sector at Constant 1980/81 Factor Cost, 1988/89–1993/94

(In millions of birr)

article image
Source: Staff estimates based on discussions with the Ethiopian authorities on the revised GDP estimates soon to be finalized. Data exclude Eritrea after 1990/91.
Table V.

Ethiopia: Indices of lmplicit GDP and Sectoral Deflators, 1988/89–1993/94

(1980/81 = 100)

article image
Source: Staff estimates based on discussions with the Ethiopian authorities on the revised GDP estimates soon to be finalized. Data exclude Eritrea after 1990/91.

Two columns are presented for 1990/91; value of the index varies including and excluding Eritrea.

Table VI.

Ethiopia: Estimates of Agricultural Production of Major Crops, 1988/89–1991/92 1/

article image
Source: Central Statistical Authority.

Excluding Eritrea, Tigrai, Ogaden, and Assab Autonomous Regions.

Private holdings.

Production data from cooperatives and state farms ceased to be compiled since 1990/91.

Table VII.

Ethiopia: Estimates of Cultivated Areas Under Major Crops, 1988/89–1991/92 1/

article image
Source: Central Statistical Authority.

Excluding Eritrea, Tigrai, Ogaden, and Assab Autonomous Regions.

Private holdings.

Production data from cooperatives and state farms ceased to be compiled since 1990/91.

Table VIII.

Ethiopia: Estimates of Coffee Production, Marketing, and Stocks, 1988/89–1993/94 1/

(In thousands of metric tons)

article image
Source: Ministry of State Farms, Coffee and Tea Development.

Estimates (except for exports) based on the findings of the Coffee Sector Survey.

As recorded on loading at port rather than at customs stations.

Represents unofficial exports, handling losses, etc.

Table IX.

Ethiopia: Coffee Exports to ICO Member and Nonmember Countries, 1988/89–1993/94 1/

(Quantity in metric tons; value in millions of birr)

article image
Source: Ministry of Coffee and Tea Development.

ICO, International Coffee Organization.

Initial quota.

Quotas suspended.

Estimates.

Table X.

Ethiopia: Coffee Prices and Shipments, 1988/89–1993/4

article image
Source: Data provided by Ethiopian authorities.

Monthly prices weighted by monthly export volumes of unwashed coffee; “Djimma, UGQ” is a standard blend.

Beginning– and end–of– month surtax rates weighted by monthly local export volumes; midpoint shown.

Monthly average auction price for total arrivals (washed and unwashed) sold at auction in Addis Ababa, weighted by arrivals.

Arrivals in Addis Ababa and Dire Dawa for inspection by the Ministry of State Farms, Coffee and Tea Development.

Ministry of Coffee and Tea Development data, which differ in timing and place of recording from customs returns. Unit values are derived from unrounded data.

Table XI.

Ethiopia: Monthly Export and Arrival Volumes, Coffee Prices, and Surtax, January 1991–May 1994

(1980/81 = 100)

article image
Source: Data provided by the Ethiopian authorities.

Auction price less estimated transport and handing costs.

For unwashed and washed coffee combined (mostly unwashed) at Addis Ababa, weighted by same month’s arrivals.

Domestic retail price at Addis Ababa.

At beginning and end of month.

Percentage excess of the domestic price over the auction price.

ICO, International Coffee Organization; U.S. cents per pound.

Table XII.

Ethiopia: Value of Manufacturing and Processing at Constant Prices, 1988/89–1991/92

(In millions of birr)

article image
Source: Ministry of Industry.
Table XIII.

Ethiopia: Output of Selected Public Sector Industrial Corporations at Constant 1978/79 Prices, 1988/89–1991/92

(In millions of birr)

article image
Source: Ministry of Industry.

The share companies include Addis Tyre, Ethiopian Pulp and Paper Factory, Ethiopian Crown and Ethio–Japanese Synthetic Textiles.

Table XIV.

Ethiopia: Income Statement for Selected Public Sector Corporations, 1989/90-1991/92

(In millions of birr)

article image
Source: Ministry of Industry.

Budget figures are included in 1989/90.

Estimates are included m 1989/90 when the accounts are not closed.

Table XV.

Ethiopia: Summary of Investment Projects, July 1992–June 7,1994

(In millions of birr, unless otherwise specified)

article image
Source: Investment Office of Ethiopia.

Thousands of birr per employee.

Table XVI.

Ethiopia: Comsumption of Petroleum Products, 1989/90–1993/94

article image
Source: Ethiopian Petroleum Corporation.

Asphalt consumption for fiscal year 1990/91 indicates nine months’ sales data only.

Table XVII.

Ethiopia: Ex Refinery and Retail Prices of Petroleum Products, 1989–94 1/

(In birr per liter)

article image
Source: Ethiopian Petroleum Corporation.

Retail prices vary from one location to another, depending on transportation costs.

Table XVIII.

Ethiopia: Domestic Production of Refined Petroleum Products, 1988/89–1993/94 1/

article image
Source: Ethiopian Petroleum Corporation.

Mainly from imports of crude oil.

Nine months’ production.

Table XIX.

Ethiopia: Electricity Production, 1988/89–1993/94

(In thousands of kilowatt–hours)

article image
Source: Ethiopian Electric and Power Authority.

Based on sales.

Table XX.

Ethiopia: Retail Price Index For Addis Ababa (Excluding Rent) January 1991 – June 1994

article image
Sources: Central Statistical Authority; and National Bank of Ethiopia.
Table XXI.

Ethiopia: Retail Price Index For Addis Ababa (Excluding Rent), 1990/91–1993/94

(1963=100)

article image
Sources: Central Statistical Authority; and National Bank of Ethiopia, Quarterly Bulletin.

The weights presented reflect the consumption pattern of households with monthly income of Br 400 or less in 1963. They do not add up to 100 because rent is excluded from the index. The statistical weights in the above item can be found by expressing the group weights as a percentage of 85.4.

Table XXII.

Ethiopia: Retail Price Index For Food Items in Addis Ababa, 1990/91–1993/94

(1963=100)

article image
Sources: Central Statistical Authority; and National Bank of Ethiopia, Quarterly Bulletin.
Table XXIII.

Ethiopia: Registered Unemployed and Reported Vacancies by Occupational Classification, 1988/89–1992/93

(In number of persons)

article image
Source: Ministry of Labor and Social Affairs.
Table XXIV.

Ethiopia: Central Government Revenues and Grants, 1988/89–1993/94

(1980/81 = 100)

article image
Source: Ministry of Finance; and staff estimates.

Excise, turnover, and transaction taxes through 1988/89; sales tax until 1992/93; sales and excise taxes thereafter.

Includes transaction taxes on petroleum products and on alcohol and tobacco through 1989/90.

Excise and transaction taxes through 1989/90; sales taxes until 1992/93; sales and excise taxes thereafter.

War levy through 1990/91 (closing transfer in 1991/92), 1992/93 includes transfer for the evacuation of Felashas.

Table XXV:

Ethiopia: Central Government Current Expenditures by Functional Classification, 1988/89–1993/94

article image
Source: Ministry of Finance.

Data not adjusted for extra budgetary (commercial) defense outlays through 1990/91.

Includes housing and community services.

Includes expenditures on social safety net measures of Br 36.9 million in 1992/93 and Br 50.0 million in 1993/94.

Table XXVI.

Ethiopia: Central Government Current Expenditures by Economic Classification, 1988/89–1993/94

article image
Sources: Ministry of Finance: and staff estimates.

Includes expenditures on social safety net measures of Br 36.9 million in 1992/93 and Br 50.0 million in 1993/94.

Table XXVII.

Ethiopia: Central Government Capital Expenditures, 1988/89–1993/94 1/

article image
Sources: Ministry of Finance; and staff estimates.

Includes outlays other than investment proper, such as operating expenses of certain development agencies.

Mainly funds at the Agricultural and Industrial Development Bank (AIDB).

Includes urban development and housing, social security, manpower and training, and sport.

Breakdown is available only in closed account actuals.

Table XXVIII.

Ethiopia: Central Government External Borrowing by Sector, 1988/89–1993/94 1/

article image
Source: Ministry of Finance.

External financing for projects on the basis of budgetary accounts.

Mainly funds of the Agricultural and Industrial Development Bank (AIDB).

Technical assistance for the Ministry of Finance.

Table XXIX.

Ethiopia: Counterpart Funds Generation from External Grants and Loans, 1991/92–1993/94

article image
Sources: Ministry of Finance, and staff estimates.

Contributions to the Economic Recovery and Rehabilitation Programme (ERRP).

Excluding structural adjustment loans of the World Bank and the AfDB.

Commodity credit.

Table XXX.

Ethiopia: Balance of Payments, 1988/89–1993/94 1/

article image
Sources: Staff estimates.

Estimates for 1992/93 and 1993/94 are adjusted for valuation effects by multiplying SDR values by the period average of secondary (auction market) foreign exchange rates.

Data cover year ended July 7, where available, otherwise June 30.

Excludes ruble denominated debt service to the former Soviet Union, except for 1988/89. Interest and amortization payments have been adjusted for debt cancellation by France in 1988/89 and by the Federal Republic of Germany and France in 1990/91 except for the years in which cancellation occurred.

Increase in net foreign assets denoted by minus sign. Excludes the accounts of the Djibouti branch of the Commercial Bank of Ethiopia.

Table XXXI.

Ethiopia: Balance of Payments, 1988/89–1993/94

article image
Sources: National Bank of Ethiopia; Ministry of Finance; Ministry of Planning and Economic Development; and staff estimates.

Data cover year ended July 7, where available, otherwise June 30.

Excludes ruble denominated debt service to the former Soviet Union, except for 1988/89. Interest and amortization payments have been adjusted for debt cancellation by France in 1988/89 and by the Federal Republic of Germany and France in 1990/91 except for the years in which cancellation occurred.

Excluding technical assistance and similar transfers recorded in the budget.

Increase in net foreign assets denoted by minus sign. Excludes the accounts of the Djibouti branch of the Commercial Bank of Ethiopia.

Table XXXII.

Ethiopia: Exports, 1988/89–1992/93 1/

(Value in millions of birr; volume in thousands of metric tons)

article image
Source: National Bank of Ethiopia.

Data based on custom records, except for gold, which is reported by the National Bank of Ethiopia.

Data cover year ended July 7, where available (otherwise June 30).

Including beverages and other industrial products.

Table XXXIII.

Ethiopia: International Prices for Selected Commodities 1/

article image
Source: National Bank of Ethiopia.

Prices are based on export permits authorized; f.o.b prices in March and April for all years.

Cost and freight prices.

Table XXXIV.

Ethiopia: Exports by Private Sector (Selected Commodities), 1988/89–1992/93

article image
Source: Ministry of Trade.
Table XXXV.

Ethiopia: Value of Imports, c.i.f. by End Use, 1988/89–1992/93 2/

(In millions of birr)

article image
Source: National Bank of Ethiopia.

Based on customs records.

Includes in 1988/89 Br 110.7 million, in 1990/91 Br 283.2 million, in 1991/92 Br 2333 million, and in 1992/93 Br 510.0 million for import of airplanes by Ethiopian Airlines.

Table XXXVI.

Ethiopia: Merchandise Trade Unit Value and Volumes, 1988/89–1992/93

article image
Sources: Staff calculations and estimates, based on information provided by the National Bank of Ethiopia.

Estimates are based on partner countries’ exports unit values.

Includes “other exports and re–exports.” The calculation of unit values and volumes of exports is based on the moving weighted average of the export values, lagged by one year, of the corresponding unit values and volumes.

Table XXXVII.

Selected Effective Exchange Rates, 1988/89–1992/93

article image
Sources: IMF, International Financial Statistics; and staff estimates.

Ended June 30.

Based on marginal rate of foreign exchange auction since May 1993.

Table XXXVIII.

Ethiopia: Percentage Shares of Private and Public Enterprises at Foreign Exchange Auctions 1/

article image
Sources: National Bank of Ethiopia, and staff estimates.

Foreign exchange auctions have been conducted biweekly since May 1,1993.

Table XXXIX.

Ethiopia: Recorded Imports by Country of Origin, 1989–92

(In millions of birr)

article image
Source: National Bank of Ethiopia; based on customs records.
Table XL.

Ethiopia: Exports by Country of Destination, 1989–92 1/

(In millions of birr)

article image
Source: National Bank of Ethiopia.

Includes re–exports.

Table XLI.

Ethiopia: Balance in Services and Transfers Accounts, 1988/89–1992/93

(In millions of birr)

article image
Sources: National Bank of Ethiopia, and Ministry of Finance.

Includes freight and insurance.

Accrual basis. Includes IMF charges and interest payments. Excludes interest payments on ruble–denominated debt to the former Soviet Union from 1989/90 onward.

Table XLII.

Ethiopia: Capital Account, 1988/89–1992/93 1/

(In millions of birr)

article image
Sources: National Bank of Ethiopia, and Ministry of Finance.

On accrual basis.

Data cover fiscal year ended July 7, where available (otherwise June 30).

Excludes amortization payments on ruble–denominated debt to the former Soviet Union from 1989/90 onward.

Estimated from exchange control record of export permits and import licenses compared with inward bills for collection; results are regarded as indicative for lags and leads in payments.

Working balances of Ethiopian Airlines and Ethiopian Shipping lines.

Table XLIII.

Ethiopia: External Public Debt Service, 1988/89–1992/93 1/

(In million of birr)

article image
Sources: Staff estimates based on data from the Ministry of Finance; National Bank of Ethiopia; and Ministry of Planning and Economic Development.

Excludes ruble–denominated debt and debt service to the former Soviet Union from 1989/90 onward.

Includes repayments of Trust Fund loans, and repurchases from IMF.

Includes IMF charges and interest payments.

Includes debt cancellation of Br 6.4 million on principal and Br 1.1 million on interest.

Table XLIV.

Ethiopia: External Public Debt Disbursements, 1988/89–1992/93

(In millions of birr)

article image
Source: National Bank of Ethiopia.

Data cover year ended July 7 where available, otherwise June 30.

Includes suppliers’ credits and commercial bank disbursements.

Former Council for Mutual Economic Assistance.

Table XLV.

Ethiopia: New External Public Debt Commitments, 1988/89–1992/93

article image
Source: Ministry of Finance.

Includes loans and credits approved by the Ethiopian Government, but for which suitable projects have not yet been identified.

Terms are calculated as weighted averages of total annual new commitments.

Table XLVI.

Ethiopia: Suppliers’ and Commercial Bank Credits Contracted, 1987/88–1991–92 1/

article image
Source: National Bank of Ethiopia.

Maturity, grace period, and interest rates are weighted averages.

No new suppliers’ or commercial bank credits were contracted during 1991/92.

Table XLVII.

Ethiopia: External Public Debt Outstanding by Lender June 1989–June 1993 1/

(In millions of birr)

article image
Source: Ministry of Finance.

From 1989 onward the ruble–denominated debt to the Soviet Union is excluded.

Includes suppliers’ credits and commercial bank disbursements.

Former Council for Mutual Economic Assistance.

Table XLVIII.

Ethiopia: External Arrears, June 1991–June 1993 1/

(In millions of birr)

article image
Source: Ministry of Finance.

Outstanding at end of period.

Excludes ruble–denominated debt to the former Soviet Union.

Table XLIX.

Ethiopia: Monetary Survey, June 1989–June 1994 1/

article image
Source: National Bank of Ethiopia.

Excludes the accounts of the Djibouti branch of the Commercial Bank of Ethiopia. Transactions of the branches in Eritrea are included only through April (for the National Bank) and May (for the Commercial Bank), 1991. Thereafter, their balances are held constant in the monetary survey of Ethiopia.

The AIDB debt transfer operations (March 1994) are not reflected in the figures for June 1994 in order to maintain comparable data series.

Based on GDP excluding Eritrea after 1990/91.

Table L.

Ethiopia: Factors Affecting Changes in Money Supply, June 1989–June 1994 1/

article image
Source: National Bank of Ethiopia.

Excludes the accounts of the Djibouti branch of the Commercial Bank of Ethiopia. Transactions of the branches in Eritrea are included only through April (for the National Bank) and May (for the Commercial Bank), 1991. Thereafter, their balances are held constant in the monetary survey of Ethiopia.

The AIDB debt transfer operations (March 1994) are not reflected in the figures for June 1994 in order to maintain comparable data series.

Table LI.

Ethiopia: Loan Portfolio of the Banking System, June 1989–March 1994 1/

article image
Sources: National Bank of Ethiopia; and Commercial Bank of Ethiopia

Includes provision for doutful debts; excludes the account of the Djibouti branch.

The AIDB debt transfer operations (March 1994) are not reflected in the figures for March 1994 in order to maintain comparable data series.

Excludes equity investment.

Excludes deposit claims on specialized banks.

Table LII.

Ethiopia: Sectoral Breakdown of Commercial Bank Claims, June 1989–March 1994 1/

article image
Sources: National Bank of Ethiopia, and staff estimates.

Excludes the accounts of the Djibouti branch.

Adjusted to exclude long–term credit to, shares in, and deposits with the Housing and Savings Bank.

Co-financing of Special Housing Projects, and loans to the HSB.

Table LIII.

Ethiopia: Commercial Bank Lending and Deposits, June 1989–March 1994

(In millions of birr)

article image
Sources: National Bank of Ethiopia.

Includes provision for doutful debts; excludes the account of the Djibouti branch.

Excludes equity investment

Excludes deposit claims on specialized banks.

Excludes the accounts of Djibouti branch.

Includes domestic banks and other financial agencies and nontransferable accounts of nonresidents.

Table LIV.

Ethiopia: The Commercial Bank of Ethiopia, Reserve and Liquidity Position June 1989–June 1994 1/

article image
Source: National Bank of Ethiopia

Position at the last business day of the last week of the month.

Demand, savings, time deposits less uncleared checks paid, uncleared effects–foreign.

The requirement is 5 percent of net deposits.

Net deposits less deposits at notice (able to withdraw after 30 days’ notice).

The requirement is 20 percent of net current deposits.

Table LV.

Ethiopia: Central Government’s Domestic Financing, 1988/89-1992/93 1/

(In millions of birr)

article image
Sources: National Bank of Ethiopia; and Ministry of Finance.

As shown in monetary data on June 30 and July 7, and on a closed budgetary accounts basis, which reflects backdating of National Bank advances.

Long–term government securities; includes interest bearing notes issued in lieu of bonds, and interest–free promissory notes arising from past currency issue.

Government bonds and treasury bills.

Calculated as a residual. Includes use of currency balances, unrecorded financing, and other discrepancies.

During the fiscal year.

Table LVI.

Ethiopia: Consolidated Balance Sheet of the Specialized Banks, June 1989–December 1993 1/

article image
Sources: National Bank of Ethiopia, and staff estimates.

The Agricultural and Industrial Development Bank (AIDB) and the Housing and Savings Bank (HSB).

Claims on state enterprises, sooperatives, private businesses, and individuals.

Mainly fixed deposits at the HSB by the Ethiopian Insurance Corporation, the Pension Commision, and the Commercial Bank if Ethiopia (CBE).

Mainly a long–term credit from the CBE to the HSB. Excludes deposits and shareholdings but includes accrued interest.

Includes AIDB capital (Br 100 million), HSB capital (Br 15.7 million, of which Br 3 million is CBE shareholdings), reserves, provisions, profits and loss accounts, etc.

Table LVII.

Ethiopia: Specialized Bank Lending, June 1989–December 1993

article image
Source: National Bank of Ethiopia.

Includes provisions for bad and doubtful debts. Balance sheet excludes such provisions but includes past investment (mostly nationalized) and miscellaneous claims.

Loans granted during the fiscal year, and credit outstanding at the end of the fiscal year.

Table LVIII.

Ethiopia: Structure of Interest Rates, as of June 30,1994

(In percent per annum)

article image
Source: National Bank of Ethiopia.

ANNEX Ethiopia: Summary of the tax system as of June 1994

(All amounts in birr)
article image
article image
article image
article image

Bibliography Fund documents

  • Annual Report on Exchange Arrangements and Exchange Restrictions, 1993 International Financial Statistics

  • SM/91/122, June 14, 1991 (Ethiopia-Recent Economic Developments)

  • SM/93/102, May 13, 1993 (Ethiopia-Recent Economic Developments)

Ethiopian documents

  • National Bank of Ethiopia, Quarterly Bulletin

  • Ministry of Planning & Economic Development, Economic Reform Programme: Impact Analysis

  • Technical Committee for Macroeconomic Development, Report on Macroeconomic Development in Ethiopia

1/

Figures are averages for 1988/89-1991/92, and are stated as percentages of total staple crop production (Tables VI, VII). Data are not available for 1992/93-1993/94, owing to the suspension of the annual crop survey conducted by the Central Statistical Authority (CSA); presently, the CSA is mainly involved in the preparation of the forthcoming census.

2/

There has been some switching to root crops, but principally in the southern areas in order to cope with drought.

3/

This implies, on average, an annual minimum food aid requirement of roughly 500,000 tons, assuming an annual cereal requirement of about 120-130 kilograms per person.

1/

These figures are very preliminary.

2/

In addition, small amounts of sugar cane and cotton are grown, mainly in state farm plantations along the Awash river in the Afar region. The latter has been adversely affected during the transition to a market economy, as most cotton growing state farms were dissolved, and smallholders switched to staple crop and livestock production, leading to a decline in national lint cotton production, to about 5,000 tons at present from approximately 30,000 tons during the Derg regime. There are efforts to study the problems of the sector with a view to reviving it. In addition, fruits and vegetables appear to have gained in importance in recent years, although no statistics have been compiled on production.

3/

In dry processing, fully ripe red cherries are directly picked from the tree and dried in the sun on raised beds, providing air circulation and avoiding soil contamination. The more capital-intensive wet processing method involves pulping, fermenting, as well as drying on elevated wire trays.

1/

There are six types of coffee licenses: coffee export, coffee transport, coffee cleaning, coffee collecting, coffee trade auxiliary, and coffee washing.

2/

Private exporters have been allowed in the exporting of sun-dried coffee since late 1989.

3/

The coffee season extends from October through September.

4/

These prices were set at 4.2 Br/kg for djinirna, 6.15 Br/kg for sidamo, 7.0 Br/kg for nekempte, and 9.1 Br/kg for hararge. The setting of floor prices was complicated for some time because of low international prices as well as the continuing demand in the domestic market, with domestic prices remaining, on average, higher than the export prices.

5/

The exact share varies across service cooperatives.

1/

However, prices can be deduced from the auction price data by adjusting for taxes and transportation and handling costs.

2/

Normally, coffee trees come into production 3-4 years after planting and are in full production at 6-8 years.

3/

Arrivals at the auctions increase during October through April, following the coffee harvest in October.

4/

However, leakages to Djibouti, Somalia, and Sudan are still estimated to be high.

5/

Averages for 1989/90-1991/92.

1/

Prior to 1991/92, AISCO used about 2,900 Service Cooperatives (SC) as the only retail outlets for fertilizer retailing. During 1990/91, a considerable number of SCs were destroyed and damaged as a result of the civil unrest and ceased to function. Following the substantial reduction in the number of retail outlets (down to about 550), AISCO, with the assistance of the United Nations Food and Agricultural Organization (FAO), introduced a New Marketing System.

1/

There is no domestic production of fertilizers in Ethiopia. Two types of fertilizers are imported--Diammonium Phosphate, or DAP (about 85 percent) and urea (15 percent).

2/

Given the regular crop cycle in Ethiopia, in order to increase fertilizer usage during a particular crop season, price announcements need to be made around December of the preceding year or in January. The 1993 crop season price was only announced in May 1993, which was already past the belg season and behind the planting phase of the meher season.

3/

The subsidy equals the difference between the cost of fertilizer to the Government and the price charged to the farmer. When farmers purchase directly from the marketing centers, they are expected to pay an additional Br 7 per quintal, which is the margin for private traders (included as part of the subsidy).

1/

Previously, farmers were forced to surrender a minimum of 50 percent of agricultural output to the AMC at fixed prices, allowing a small margin over cost. Checkpoints were established along major grain routes, and illegally transported farm produce was confiscated. One consequence was the reduction of the smooth flow of grains from surplus to deficit areas. This led to very low prices in surplus areas and extremely high prices in deficit areas. On the other hand, AMC purchases were mostly allocated to urban areas, institutions (such as flour mills) and, more importantly, the military. Given the heavy involvement of AMC in the grain trade, grains were transacted at artificially low prices, which bore little relation to market prices. In March 1990, the quota system governing the surrender of grains was abolished, and merchants were allowed to operate freely.

1/

For historical background of land issues, see Ethiopia: Recent Economic Developments, SM/93/102, May 13, 1993.

2/

Recent data on manufacturing pertaining to 1992/93-1993/94 are not available because of disruptions related to the restructuring at the Ministry of Industry. Percentages reported are based on the gross value of production in 1989/90, presumably a normal year, and before the onset of acute shortages. Also see Tables XII-XIV.

1/

The Supervisory Authority is in charge of evaluating the performance of PEs as well as in organizing their external auditing. The Authority is also represented in a PE’s Board of Directors for the purpose of assessing the PE’s investment budget.

2/

At the level of retail trade, the private sector has already started to play a more active role. Regarding wholesale trade, the previous institutions inherited from the earlier regime still occupy the dominant share, although participation by the private sector is now allowed. The former institutions are now autonomous entities with their own Board of Directors and do not receive subsidies from the Government. The Ethiopian Domestic Distribution Corporation and the Ethiopian Import-Export Corporation were merged. The new organization will continue to be involved in wholesale trade as well as in the importation of consumer goods.

1/

No data are available on the structure of wages in the private sector.

1/

The old structure within which the Wage Board used to operate has been dismantled; now the Board directly reports to the Prime Minister’s Office.

2/

Inflation is measured by the Addis Ababa Retail Price Index (AARPI), which suffers from a number of statistical problems in terms of its coverage and expenditure weights. See Ethiopia: Recent Economic Developments, SM/91/122, 6/14/91.

1/

About 57 percent of the AARPI is composed of food items; consequently, the seasonal pattern of inflation, as measured by the AARPI, is largely determined by the seasonal pattern of the food index. A simple regression on food price inflation (covering the last ten years), including monthly seasonal dummies and controlling for the escalation of war in 1990/91 and the drought in 1984/85, indicates that the pattern of monthly inflation in food prices follows broadly the developments in the crop pattern. Specifically, monthly inflation is significantly lower in December and starts increasing gradually thereafter during January through April; no significant trend is observed during May through July. In August, inflation appears significantly higher, and then remains stable till November.

2/

There were 4 RIOs as of June 7, 1994.

2/

Under the revised figures, Eritrea constitutes broadly about 7 percent of Ethiopia.

1/

For most enterprises, this tax is assessed on the basis of profits from the past fiscal year. For incorporated businesses, a flat rate of 50 percent is applied; for unincorporated business, the tax has 18 brackets and a maximum tax rate of 59 percent (for taxable income above Br 24,000/year).

2/

Members of agricultural producers’ cooperatives pay Br 5 per hectare, while other individual farmers pay Br 10 per hectare.

1/

During 1992/93, most imports that came into the country before April 1993 were still recorded for customs valuation purposes at the old exchange rate of Br 2.07 = US$1, even though the devaluation took place in October 1992. This reflected the fact that most of these imports were shipped or contracted for prior to the devaluation announcement.

2/

Also, the minimum tariff rate was raised to 5 percent, the number of zero-rated items was reduced (which also led to a reduction in the number of bands from 25 to 10), all but three specific duties were converted into ad valorem rates, the sales tax on imports was reduced to the level of the sales tax on domestic goods (12 percent), and the import duty drawback mechanism was simplified and expanded.

1/

Previously, the export of a number of other products--including cotton seed, sugar, and hides and skins--had been taxed at various rates.

2/

In addition, nontax revenue includes charges and fees, receipts from sales of goods and services, pension contributions, reimbursements and property sales, and other extraordinary receipts. Under the previous government, a war levy had also been introduced; this was abolished in 1990/91.

1/

The European Union, the World Bank, the African Development Bank (AfDB), the Canadian International Development Agency (CIDA), the Swedish International Development Agency (SIDA), and the Netherlands. Also, the Ministry of Finance and representatives of donors began to hold regular consultation meetings on counterpart funds issues.

1/

According to the World Bank, the poverty line for urban households in 1990 was Br 244 per month.

2/

The salary adjustment was accompanied by a reform of the personal income tax, which favored the recipients of higher salaries relatively more through the reduction of tax rates. For details on this, see Chapter IV.2; for more details on the public sector wage policy, refer to Chapter II.4.

1/

According to a World Bank estimate, for example, urban consumers (about 13 percent of the population) received about 60 percent of the food grain subsidy. Moreover, some of these subsidies were hidden in extrabudgetary programs.

2/

In fact, it can be argued that this explicit subsidy replaced the implicit subsidy on fertilizer resulting from the overvalued exchange rate prevailing prior to October 1992.

1/

A total of 63 population groups were identified, of which 46 were considered large enough to form their own governments.

2/

The names of the 10 official Regions are: Tigray (Region 1), Afar (2), Amhara (3), Oromia (4), Somali (5), Benshangul (6), Southern Ethiopian Peoples’ Administrative Region (SEPA; former Regions 7,8,9,10 and 11), Gambella (12), Harar (13), Addis Ababa (14). Dire Dawa is treated as an independent Region for budgetary purposes, although it is not officially recognized. Administratively, each Region is divided into zones, which in turn are subdivided into districts called woredas.

1/

In fact, annual budgets should be prepared starting at the woreda level, and they should then be combined into zonal budgets, and finally incorporated into a regional budget.

1/

Regular imports are defined as total imports less grain and commodity aid imports.

1/

The trends described in 1992/93 and 1993/94 refer to SDR values since these adjust for the effects of devaluation.

1/

The real effective exchange rate appreciated by about 30 percent during 1989/90-1991/92.

1/

In contrast, subsidies for the export of coffee provided some stimulus to export supply.

2/

Washed coffee constitutes about 15 percent of total exports by volume, the remainder being accounted for by unwashed coffee exports.

1/

Purchases of gold by the NBE from the Ethiopian Mineral Resources Corporation are included in Ethiopia’s export account. These are not normally treated as exports in the balance of payments of most countries, however, so a corresponding debit entry is included in the Ethiopian balance of payments under the item, “short-term capital flows, net.” Thus, gold purchases by the NBE have no effect on the overall balance of payments.

2/

Disbursements under the EU’s STABEX fund are not initially owned by the Government of Ethiopia even after there is an agreement between the donor and the authorities in principle to disburse the funds. At this stage, such funds are not reflected in the net foreign assets of the NBE (being reflected in both foreign assets and foreign liabilities). To maintain consistency with Ethiopia’s monetary survey, therefore, a debit entry is included in the capital account corresponding to the credit entry in the official transfers item, so that the overall balance of payments and the NBE’s reserves are both unaffected. Once the document has been signed by both the EU and TGE representatives, ownership of the funds is transferred to the Government of Ethiopia, and the funds are reflected in the NBE’s net foreign assets as well as in external reserves.

1/

CMEA, Council of Mutual Economic Assistance.

2/

The participating countries were Austria, Belgium, Canada, Finland, France, Germany, Italy, Japan, the Netherlands, Sweden, the United Kingdom, and the United States. The observers were from Australia, Denmark, Israel, Spain, and Switzerland as well as from the African Development Bank, the World Bank, UNCTAD, and the Fund.

3/

A more detailed description of Ethiopia’s exchange and trade system as of December 31, 1992 is provided in the IMF’s 1993 Annual Report on Exchange Arrangements and Exchange Restrictions.

1/

See Chapter VII on the demand equation for foreign exchange at the auction.

1/

See Chapter II, SM/93/102, 5/13/92, for a detailed description of the May 1992 investment code.

1/

During the period covered, the monetary survey takes account of transactions of Eritrean branches of the NBE and the Commercial Bank of Ethiopia (CBE) only through April and May 1991, respectively. Thereafter, their balances have been held constant in the monetary survey of Ethiopia. The monetary survey, however, does not attempt to identify the magnitude of birr notes circulating outside Ethiopia.

2/

From 1991/92 onward, GDP figures used for the calculation of velocity exclude Eritrea. For the purpose of providing comparable estimates of velocity for the full period, it would be necessary to add Eritrean GDP to that of Ethiopia for 1991/92 onward. The velocity figures for Ethiopia and Eritrea together would be 2.43 for 1991/92, 2.67 for 1992/93, and 2.48 for 1993/94.

1/

For example, the STABEX fund has a double signature feature. In Ethiopia’s monetary survey, it is initially treated as both a foreign asset and a foreign liability once the funds have been deposited in the correspondent account. When the formal document has been signed by the TGE and EC representatives, the ownership of the funds is transferred to the TGE and reclassified under other liabilities. The TGE, however, can only utilize these funds if the necessary documents have been submitted to the EC.

2/

Gross base, before provisions, and excludes local investment shares of the CBE.

3/

Loans from the NBE, which were in turn loaned by the AIDB to the state farms.

4/

Based on the latest available figures.

1/

Measured by loans and advances by the CBE.

1/

A real interest rate estimate based on the GDP deflator (factor cost) will give minus 7.6 percent for the same period. Both price indicators have some deficiencies, such as limited coverage and an out-of-date base period for the AARPI, and limited reliability and lack of timeliness for the GDP deflator.

1/

The large difference in outstanding balances of savings and time deposits seems to come from the different nature of the two deposits. Time deposits have a predetermined maturity; withdrawal before the expiry date will cost the depositor the loss of interest relevant to the year in which the withdrawal takes place. In addition, time deposits cannot be immediately withdrawn and usually involve a waiting period of a few days. In contrast, savings deposits can be readily withdrawn within a short period of time and without any cost in interest.

1/

SACCs are also known as Credit Union, Thrift and Credit Co-operatives, or Savings and Credit Associations. An SACC is defined as “a free association of people with a common bond who save and lend money to one another at a low interest rate for productive and provident purposes.” The objectives of SACCs are encouraging thrift among its members and providing services, and not in making profit. Members are paid dividends on shares and interest on savings deposits.

2/

Service cooperatives are in the process of reorganizing as “cooperative societies” under the Agricultural Cooperative Societies Proclamation (issued on February 1, 1994).

1/

The Credit Unit of the NBE monitors and supervises a bank’s credit limits for a borrower, according to certain routine. If the borrower is a public enterprise, the CBE and HSB can lend as much as 10 percent of the sum of working capital, reserves, and profits; the AIDB has to submit a project appraisal when the contemplated loan amount exceeds Br 3 million. If the borrower is from the private sector, the HSB can lent as much as 5 percent of working capital, reserves, and profits; the CBE and AIDB are expected to submit a project appraisal when the contemplated loan amount exceeds Br 1 million and Br 0.5 million, respectively.

1/

The outstanding borrowing prior to the issuance of the proclamation is exempt from these ceilings. For 1989/90-1991/92, AAOR was Br 2,677 million; outstanding balances for direct advances, treasury bills, and government bonds at end-June 1992, were Br 5,145 million (192 percent of AAOR) Br 562 million (21 percent), and Br 3,390 million (127 percent), respectively.

1/

According to Proclamation No. 84/1994, reserve and liquid assets are to be calculated based on the last preceding monthly statement; in practice, the Accounts Department of the NBE monitors these levels on a more frequent basis.

2/

This definition has been in effect since August 17, 1993. Prior to this, all time deposits were deducted from net deposits in order to obtain net current deposits.

3/

As of end-May, the actual reserve and liquidity ratios were 23.1 percent and 46.8 percent, respectively.

1/

Cooperatives used to enjoy the lowest rates (4.5-6 percent), relative to the state-owned (2.5-8 percent) and private sectors (6-10 percent).

1/

The first repayment is due in November 1994.

1/

This recapitalization scheme consists of two stages: (i) the Government will return buildings (Br 70.4 million) as well as inject cash (Br 57.5 million) to the AIDB, which will bring AIDB’s capital level to zero; and (ii) Br 62.5 million will be initially provided in cash during 1994/95, with a remaining Br 187.5 million paid over the following five years, in order to raise the capital base to Br 250 million.

2/

Br 10 million or 8 percent of bank’s risk-weighted assets based on the most recent annual balance sheet, whichever is larger.

3/

Each shareholder’s share not exceeding 20 percent of the total.

1/

In this way, it is assumed that “other items net” pick up the slack on the asset side.

2/

The amount of birr in circulation in Eritrea does not appear at all in Eritrea’s monetary accounts, since such currency notes are not a liability of the Bank of Eritrea, but rather of the National Bank of Ethiopia. Indeed, the BE does not have any information on currency in circulation nor on any changes in birr in circulation that would arise from movements of birr currency across the Ethiopian and Eritrean borders.

3/

In what follows, NFABR is assumed to be comprised of cash for simplicity. Otherwise, depending on its exact composition, deposits would also require adjustments. The noncash component of NFABR reportedly includes items that are still under negotiation between the two banking systems, and is relatively small.

1/

Although the money multiplier would vary considerably.

2/

In an ex ante sense, a further complication would arise; in addition to the assumption on the distribution of currency-in-circulation between the two countries, an estimate on the size and direction of future intra-union monetary-flows would be required, if these adjustments were to be made for the future. For instance, had these accumulated claims been used to extend domestic credits in Eritrea, area-wide broad money could change substantially. In that context, an additional policy question would arise for Ethiopia, i.e. to assess the share of the expansion that is expected to fall upon Ethiopian output.

1/

The transfer could also have derived from a shift of deposits. Focusing on a cash note transfer is a simplifying assumption and should not affect the essence of the argument.

1/

The model specification is log(BRit)= a-b(CDit)+ c(MRt-1)+ d(EDt-1)+et

1/

The variable equals 0 for Auctions 1-4 and 1 thereafter.

2/

NL1 equals 0 for Auctions 1-9 and 1 thereafter. NL2 equals 0 for Auctions 1-22 and 1 thereafter.

  • Collapse
  • Expand
Ethiopia: Recent Economic Developments
Author:
International Monetary Fund
  • CHART 1

    ETHIOPIA DEVELOPMENTS IN THE ADDIS ABABA RETAIL PRICE INDEX, 1983–94

  • CHART 2

    ETHIOPIA DEVELOPMENTS IN THE FOREIGN EXCHANGE AUCTION, MAY 1993 – AUGUST 1994 1/

  • CHART 3

    ETHIOPIA DEVELOPMENTS IN INDICES OF NOMINAL AND REAL TRADE–WEIGHTED EXCHANGE RATES, 1980–94 1/

    (1980=100)