Abstract
This Selected Issues paper assesses the implications of a significant increase in the flow of external financing and grants on real GDP growth in Ethiopia. The paper presents an analysis of the sources of growth during 1991/92–2003/04, as well as an assessment of potential GDP growth. The paper also seeks to assess the historical relationship between foreign aid and the performance of the external sector in Ethiopia to establish whether foreign aid inflows have had an adverse effect on the tradable goods sector in the past.
II. Assessing the Implications for Growth of Achieving the MDGS1
A. Introduction
5. Recent assessments indicate that based on current trends, most MDGs will not be met by most countries, and that while the income poverty goal is likely to be achieved on a global level, African countries will fall well short of this goal.2 Preliminary and partial analysis by the World Bank shows that on current trends, Ethiopia will only achieve the MDG goal for the primary school enrollment ratio.
6. Ethiopia’s PRSP (the Sustainable Development and Poverty Reduction Program–SDPRP) presents sectoral targets and spending programs, that are considered to be consistent with achieving the income poverty MDG, while also making progress towards achieving the other MDGs.3 These programs, however, reveal a significant gap between available resources and those required for achieving these objectives. For instance, the estimated costs of fully implementing the SDPRP during 2002/03-2004/05 (including the cost of the Food Security Program), are 138 percent of 2002/03 GDP, while the government’s proposed medium-term expenditure framework allocates 65 percent of 2002/03 GDP to poverty-related expenditure.
7. According to projections by the Ethiopian authorities, real GDP growth should average 5.7 percent per year to 2015 in order to achieve the income poverty goal of halving the number of people living in poverty. However, based on an assumption of significantly higher external aid flows and implementation of the reforms identified in the SDPRP, the authorities are targeting average annual real GDP growth of about 7 percent. Under this scenario, the authorities assume that external financing and grants would gradually rise from about 11 percent of GDP in 2003/04 to 22 percent by 2015/16. Such higher resource flows would allow per capita poverty spending (in U.S. dollars) to rise from about $20 in 2003/04 to about $78 by 2015/16, while the ratio of public expenditure to GDP would rise from 31 percent in 2003/04 to 42 percent by 2015/16.
8. The purpose of this analysis is to assess the implications for real GDP growth of a significant increase in the flow of external financing and grants. Section B of this paper presents an analysis of the sources of growth during 1991/92-2003/04, as well as an assessment of potential GDP growth, while Section C assesses the implications for achieving average annual real GDP growth of 7 percent.
B. Sources of Growth
9. Reforms aimed at transforming the Ethiopian economy from a centrally planned economy under the Derg regime (1974-1991) to a market-oriented economy were launched by the current government in 1991. Real GDP growth (at factor cost) during 1991/92-2003/04 averaged 4.0 percent per year, while real GDP per capita growth averaged 1.1 percent per year.
10. While this represents a significant improvement on real GDP growth of 2.8 percent per year, compared with the period of the Derg regime, growth remained volatile. For example, the standard deviation of real GDP growth in Ethiopia during 1981–2002 was 6.5 relative to a mean growth rate of 2.8 percent, which was significantly higher than in neighboring countries such as Kenya (2.2), Tanzania (2.4), Uganda (3.6), and Zambia (4.6). Econometric analysis suggests that the volatility of real GDP growth in Ethiopia is largely due to the continued dependence of agricultural production (which accounted for about half of real GDP during 1981-2002) on rainfall.
11. A simple regression framework is used to explore the relationship between the level of real GDP, average annual rainfall, and the terms of trade for the period 1974–2002. The results suggest that rainfall and a trend explain about 94 percent of the variance in the level of real GDP. Fluctuations in average annual rainfall in particular appear to have a substantial impact on real GDP, with a change of 1 percent in average annual rainfall leading to a change in real GDP of 0.3 percent in the next year. While the terms of trade had the correct sign, it was not statistically significant.
Ethiopia: Estimating the Impact of Exogenous Variables on Real GDP
Ethiopia: Estimating the Impact of Exogenous Variables on Real GDP
| Dependent Variable: GDP | ||||
|---|---|---|---|---|
| Method: Least Squares | ||||
| Sample(adjusted): 1975 2002 | ||||
| Variable | Coefficient | Std. Error | t-Statistic | Prob. |
| C | 6.713605 | 0.605934 | 11.07976 | 0 |
| TOT(-1) | 0.02484 | 0.039268 | 0.632572 | 0.5333 |
| RAIN(-1) | 0.317317 | 0.086526 | 3.667315 | 0.0013 |
| T | 0.022484 | 0.002357 | 9.538986 | 0 |
| ADT | -0.001742 | 0.002197 | -0.792908 | 0.4359 |
| R-squared | 0.942486 | Mean dependent var | 9.3372 | |
| Adjusted R-squared | 0.932484 | S.D. dependent var | 0.181176 | |
| S.E. of regression | 0.047077 | Akaike info criterion | -3.1137 | |
| Sum squared residual | 0.050973 | Schwarz criterion | -2.8758 | |
| Log likelihood | 48.59109 | F-statistic | 94.22635 | |
| Durbin-Watson statistic | 1.615758 | Prob(F-statistic) | 0 | |
Ethiopia: Estimating the Impact of Exogenous Variables on Real GDP
| Dependent Variable: GDP | ||||
|---|---|---|---|---|
| Method: Least Squares | ||||
| Sample(adjusted): 1975 2002 | ||||
| Variable | Coefficient | Std. Error | t-Statistic | Prob. |
| C | 6.713605 | 0.605934 | 11.07976 | 0 |
| TOT(-1) | 0.02484 | 0.039268 | 0.632572 | 0.5333 |
| RAIN(-1) | 0.317317 | 0.086526 | 3.667315 | 0.0013 |
| T | 0.022484 | 0.002357 | 9.538986 | 0 |
| ADT | -0.001742 | 0.002197 | -0.792908 | 0.4359 |
| R-squared | 0.942486 | Mean dependent var | 9.3372 | |
| Adjusted R-squared | 0.932484 | S.D. dependent var | 0.181176 | |
| S.E. of regression | 0.047077 | Akaike info criterion | -3.1137 | |
| Sum squared residual | 0.050973 | Schwarz criterion | -2.8758 | |
| Log likelihood | 48.59109 | F-statistic | 94.22635 | |
| Durbin-Watson statistic | 1.615758 | Prob(F-statistic) | 0 | |
12. With the achievement of higher growth during 1991/92-2003/04, the structure of the Ethiopian economy changed noticeably (Figure II.1). The contribution of agriculture to real GDP declined from 57 percent in 1991/92 to 42 percent in 2003/04, and that of services rose from 34 percent to 47 percent. However, the contributions to real GDP by industry and private services (i.e. excluding the public sector) remained essentially unchanged.
Ethiopia: Sectoral Contribution to Real GDP
(In percent of GDP)
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
Ethiopia: Sectoral Contribution to Real GDP
(In percent of GDP)
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
Ethiopia: Sectoral Contribution to Real GDP
(In percent of GDP)
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
13. Furthermore, growth in agricultural production and the services sectors continued to be important for real GDP growth, while growth in the valued-added of industry did not make an important contribution to overall output growth (Table II.2).
Ethiopia: Contribution to real GDP Growth
Ethiopia: Contribution to real GDP Growth
| 1991/92-2003/04 | ||
|---|---|---|
| Agriculture | 1.0 | |
| Industry | 0.5 | |
| Services | 2.5 | |
| Services excluding public sector | 1.3 | |
| Real GDP at factor cost | 4.0 | |
| Consumption | 5.2 | |
| Private | 4.4 | |
| Public | 0.9 | |
| Investment | 0.9 | |
| Private | 0.4 | |
| Public | 0.5 | |
| Resource gap | 0.2 | |
| Exports | 1.3 | |
| Imports | -1.1 | |
| Real GDP at market prices | 6.3 | |
Ethiopia: Contribution to real GDP Growth
| 1991/92-2003/04 | ||
|---|---|---|
| Agriculture | 1.0 | |
| Industry | 0.5 | |
| Services | 2.5 | |
| Services excluding public sector | 1.3 | |
| Real GDP at factor cost | 4.0 | |
| Consumption | 5.2 | |
| Private | 4.4 | |
| Public | 0.9 | |
| Investment | 0.9 | |
| Private | 0.4 | |
| Public | 0.5 | |
| Resource gap | 0.2 | |
| Exports | 1.3 | |
| Imports | -1.1 | |
| Real GDP at market prices | 6.3 | |
14. On the demand side of the economy, growth in consumption expenditure, and particularly private consumption, was the most important source of real GDP growth, while the external sector contributed only marginally to real growth (Table II.2).
15. Reflecting the above, consumption expenditure continued to account for a significant proportion of real GDP (Figure II.2).
Ethiopia: Contribution to GDP
(In percent of GDP)
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
Ethiopia: Contribution to GDP
(In percent of GDP)
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
Ethiopia: Contribution to GDP
(In percent of GDP)
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
16. A growth accounting framework was also applied in order to decompose the growth rate of real GDP into contributions from the accumulation of factor inputs (capital and labor), and improvements in total factor productivity (TFP). The shares of capital and labor are generally derived, either from national accounts, or through econometric estimation. Given data limitations, the shares of capital and labor in real GDP are assumed to be 0.35 and 0.65 respectively, which is consistent with the estimates reported for many developing countries. Of the average annual growth rate of 4.0 percent in real GDP during 1991/92-2003/04, growth in physical capital contributed 1.4 percentage points, labor growth contributed 2.0 percentage points, and the remaining 0.7 percentage points was contributed by TFP (Table II.3).
Ethiopia: Sources of Growth and Potential Real GDP Growth
(In percent)
Ethiopia: Sources of Growth and Potential Real GDP Growth
(In percent)
| 1991/92-2003/04 | |
|---|---|
| Growth accounting | |
| Real GDP at factor cost | 4.0 |
| Capital stock | 1.4 |
| Labor | 2.0 |
| Total factor productivity | 0.7 |
| Potential GDP growth | |
| HP filter | 4.5 |
| Production function | 4.4 |
| Capital-output ratio | 4.4 |
Ethiopia: Sources of Growth and Potential Real GDP Growth
(In percent)
| 1991/92-2003/04 | |
|---|---|
| Growth accounting | |
| Real GDP at factor cost | 4.0 |
| Capital stock | 1.4 |
| Labor | 2.0 |
| Total factor productivity | 0.7 |
| Potential GDP growth | |
| HP filter | 4.5 |
| Production function | 4.4 |
| Capital-output ratio | 4.4 |
17. Potential GDP growth during 1991/92-2003/04 is assessed through the application of three methodologies, namely an HP filter, the production function approach, and the capital-output ratio approach. In the context of the real GDP series, the HP filter derives a trend output rate such that it minimizes a weighted average of the gap between actual output and trend output (Figure II.3). While the principal advantage of this technique is its simplicity, the major shortcoming is that it does not have an economic basis in the sense that the estimated productive limits of the economy are not based on the available factors of production.
Ethiopia: Actual and Potential Output
(In millions of birr)
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
Ethiopia: Actual and Potential Output
(In millions of birr)
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
Ethiopia: Actual and Potential Output
(In millions of birr)
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
18. The production function models output as a function of capital, labor and total factor productivity–the functional form is a Cobb-Douglas production function.
and thus
19. Empirically, the practice is to estimate potential output as the level of output associated with a normal rate of capacity utilization, labor input at the level of the natural rate of unemployment, and total factor productivity (TFP) at its trend level. In practice, this involves the following steps: (i) TFP growth is derived as the difference between the observed real GDP growth and the weighted sum of capital and labor growth; (ii) trend growth rates are computed for labor and TFP; and (iii) potential GDP growth is estimated as the sum of potential TFP growth and the weighted sum of the growth in capital and potential labor.
20. The underlying assumption of the capital-output ratio approach is that developing countries are characterized by excess labor (although there may be shortages in categories of skilled labor), and a lack of infrastructure and capital. The normal capacity of the economy is thus determined by dividing the capital stock by trend productivity, as measured by the trend capital-output ratio. The trend capital-output ratio reflects technology, the composition of capital, and the quality of capital and labor. Generally, it takes a long time to significantly change the composition and quality of capital and labor, and the trend capital-output ratio therefore does not change much in the short run (Figure II.4).
Ethiopia: Capital-Output Ratio
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
Ethiopia: Capital-Output Ratio
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
Ethiopia: Capital-Output Ratio
Citation: IMF Staff Country Reports 2005, 028; 10.5089/9781451812749.002.A002
21. According to the above analysis, potential GDP growth during 1991/92-2003/04 thus amounted to about 4.4 percent per year (Table II.3).
C. Assessing the Authorities’ MDG Growth Scenario
22. The Ethiopian authorities consider that a doubling of external aid flows and determined implementation of the reforms detailed in the SDPRP would significantly enhance Ethiopia’s ability to achieve the MDGs. Specifically regarding the income poverty MDG, the authorities consider that raising the level of external aid and implementation of the identified reforms would allow average annual real GDP growth to rise to a level of 7 percent. Such a growth performance would be consistent with the high case scenario presented in the authorities’ first annual progress report (APR) of the SDPRP, but would constitute a substantial improvement over the growth performance during 1991/92-2003/04 (Table II.4). The authorities’ medium-term projection is premised in particular on a significant increase in agricultural output growth to an annual average of 7.5 percent from 2.2 percent during 1991/92-2003/04.
Ethiopia: Assessing Medium-term Real GDP Growth
Ethiopia: Assessing Medium-term Real GDP Growth
| Actual | Authorities’ Projection | ||||
|---|---|---|---|---|---|
| ICOR | Real value added | ||||
| 1991/92-2003/04 | 2004/05-2020/21 | ||||
| Production approach | |||||
| Real GDP at factor cost | 4.0 | 6.7 | 7.0 | ||
| Agriculture | 2.2 | 7.5 | |||
| Nonagriculture | 5.8 | 6.6 | |||
| Real per capita GDP | 1.1 | 4.2 | |||
| Nominal investment/Nominal GDP at market prices | |||||
| Total | 16.8 | 31.4 | |||
| Private | 9.1 | 15.6 | |||
| Public | 7.7 | 15.8 | |||
| Growth accounting | |||||
| Real GDP at factor cost | 4.0 | 7.0 | |||
| Capital stock | 1.4 | 3.0 | |||
| Labor | 2.0 | 2.3 | |||
| TFP | 0.7 | 1.7 | |||
Ethiopia: Assessing Medium-term Real GDP Growth
| Actual | Authorities’ Projection | ||||
|---|---|---|---|---|---|
| ICOR | Real value added | ||||
| 1991/92-2003/04 | 2004/05-2020/21 | ||||
| Production approach | |||||
| Real GDP at factor cost | 4.0 | 6.7 | 7.0 | ||
| Agriculture | 2.2 | 7.5 | |||
| Nonagriculture | 5.8 | 6.6 | |||
| Real per capita GDP | 1.1 | 4.2 | |||
| Nominal investment/Nominal GDP at market prices | |||||
| Total | 16.8 | 31.4 | |||
| Private | 9.1 | 15.6 | |||
| Public | 7.7 | 15.8 | |||
| Growth accounting | |||||
| Real GDP at factor cost | 4.0 | 7.0 | |||
| Capital stock | 1.4 | 3.0 | |||
| Labor | 2.0 | 2.3 | |||
| TFP | 0.7 | 1.7 | |||
23. To assess the impact of such a scenario on real GDP growth, the staff was guided by the pattern of expenditure detailed in the “Extended PRSP” scenario in the Public Expenditure Review of the World Bank to develop a long-term profile for public recurrent and capital spending. This projects an increase in the ratio of public investment to GDP to an average of 15.8 percent during 2004/05-2020/21 compared with 7.7 percent during 1991/92-2003/04. Furthermore, private sector investment is expected to be positively affected by the implementation of the authorities’ reform program; as a working assumption, it is assumed that the ratio of private to public investment would be the same as that projected under the baseline scenario.
24. Utilizing two approaches, namely the Incremental Capital-Output Ratio (ICOR) approach, and a growth accounting approach, the staff’s calculations show that achieving the authorities’ target for raising the level of average annual real GDP growth to 7 percent would require, in addition to the projected increase in public sector investment, also a substantial increase in private sector investment, as well as in total factor productivity.4
25. The ICOR approach is based on the assumption that economic growth depends on investment as a share of GDP, adjusted by a factor which represents the quality of investment. Thus:
where g is real GDP growth, I is total investment, Y is real output, and μ is the quality of investment, or the incremental capital-output ratio. The ICOR represents the units of additional capital that are required to yield a unit of additional output.
26. Assuming the same ICOR as that implied in the baseline scenario, and applying the projected level of total investment, as described above, yields an average annual growth rate in real GDP at factor cost of 6.7 percent during 2004/05-2020/21. In terms of the growth accounting framework, and accounting for the projected increases in capital and labor, the contribution to real GDP growth from TFP would have to equal 1.7 percentage points in order to achieve average annual growth of 7 percent. The projected capital stock is calculated using the projection for gross investment, while for labor, it is assumed that the labor force continues to grow at the trend growth rate, and that education (as measured by average years of schooling) would increase at an average rate of 5.1 percent per year between 2000 and 2021, a rate that is comparable to historical growth rates.
27. The above conclusion regarding the attainability of the authorities’ target for real GDP growth depends critically on significant progress with the implementation of the authorities’ reform program. This assessment is consistent with empirical research which shows that the effect of aid on growth depends on the quality of institutions and policy.5 Thus, raising average annual real GDP growth to 7 percent over the medium term from 4.0 percent during 1991/92-2003/04 would require, in addition to raising the level of external aid, significant progress with the implementation of key structural reforms in agriculture, private sector development, financial sector development, and external trade.6
References
Arora, V., Bhundia, A., and Bagattini, G., June 2002, “Potential Output and the Sources of Growth, South Africa”, IMF Country Report No. 03/18, January 2003 (Washington: International Monetary Fund).
Burnside, C., and Dollar, D., “Aid, Policies and Growth: Revisiting the Evidence”, World Bank Policy Research Working Paper 3251, March 2004 (Washington: World Bank).
Cohen, D., and Soto, M., “Growth and Human Capital: Good Data, Good Results”, OECD Development Centre, Technical Paper No. 179, September 2001 (Paris: Organization for Economic Cooperation and Development).
Easterly, W., “Can Foreign Aid Buy Growth?”, Journal of Economic Perspectives, Vol. 17, No. 3, summer 2003.
International Monetary Fund, 2004, Global Monitoring Report–Policies and Actions for Achieving the MDGs and Related Outcomes (Washington).
Shiau, A., Kilpatrick, J., and Matthews, M., 2002, “Seven Percent Growth for Mexico? A Quantitative Assessment of Mexico’s Investment Requirements”, Journal of Policy Modeling, Vol. 24, Issues 7-8, November 2002
Tahari, A., Ghura, D., Akitoby, B., and Brou Aka, E., “Sources of Growth in Sub-Saharan Africa”, IMF Working Paper (draft) (Washington: International Monetary Fund).
Vamvakidis, A., and Zanforlin, L., 2002, “The Determinants of Growth: The Experience in the Southern European Economies of Greece and Portugal”, IMF Country Report No. 02/91, April 2002 (Washington: International Monetary Fund).
Prepared by Lodewyk J. F. Erasmus (AFR).
See Draft Global Monitoring Report 2004–Policies and Actions for Achieving the MDGs and Related Outcomes.
See Ethiopia–Sustainable Development and Poverty Reduction Program.
See The Federal Democratic Republic of Ethiopia–Staff Report for the 2004 Article IV Consultation and Sixth Review Under the Three-Year Arrangement Under the Poverty Reduction and Growth Facility.
For details of required reforms, see reference noted in 4 above.
