Growth and Welfare Effects of Macroeconomic Shocks in Uganda

Abstract

The role of external shocks in explaining output fluctuations is contingent upon several factors such as policy stability and effectiveness. African countries were relatively resilient to the 2008 global financial and economic crisis due to the strong policy buffers that were established following sustained periods of prudent policy design and implementation prior to the onset of the crisis. However, these cushions that relatively minimized the impact of the 2008 global recession have weakened. Fiscal and external balances are weaker, debt levels are trending up, foreign exchange reserve levels are lower, and the fiscal space has diminished. This paper focuses on Uganda and seeks to examine, using a computable general equilibrium model, the growth and welfare effects of three macroeconomic shocks: changes in terms of trade, changes in the price of oil, and changes in development assistance inflows. Our analysis reveals three key findings. First, the opposite and often offsetting effects of the three external shocks examined in our paper on the agriculture, industry and the services sectors yields only short-term as opposed to permanent deviations from trend growth in real GDP. Second, the primary channel through which these three external shocks are transmitted to the domestic economy and GDP growth is via changes in the real effective exchange rate and the associated effect on the non-tradable (domestic) and tradable (both import-dependent and/or export-oriented) sectors. Third, household welfare decreases permanently during the entire simulation period for two of the three simulations considered by this study.

Keywords: Sub-Saharan Africa, Uganda, Macroeconomic Shocks, Economic Growth and Household Welfare, Computable General Equilibrium

JEL classification: D58, E62, I32

1

2

1. INTRODUCTIONMacroeconomic volatility has varied impacts on growth and business cycles and such impact is particularly detrimental in developing countries (Loayza and Raddatz 2007). Some of the more studied causes of macroeconomic volatility include fluctuations in terms of trade which are estimated to account for up to 50% of all output fluctuations (Easterly et al. 1993, Hnatkovska and Loayza 2005, Ahmed 2003, and Raddatz 2005). Domestic conditions including policy buffers, efficiency of factor and product markets, and quality of institutions have the potential to mitigate or amplify the impact of the resulting shocks (Caballero and Krishnamurty 2001). Rodrik (1998) argues that social conflicts and domestic conflict management institutions can determine the severity of macroeconomic shocks and the ability of countries to respond to these volatilities. The impact of external shocks on economic output and other outcomes such as poverty is intuitive, particularly for developing countries including in Africa. For these countries in particular, commodities which are prone to price fluctuations account for over 70% of total exports (Raddatz 2008). Between 2007 and 2011, primary commodities including oil, natural gas and metals accounted for over 80% of intra-Africa exports for Algeria, Angola, Mali, Mauritania, Niger, and Nigeria (UNCTAD 2013). Thus, swings in commodity prices can lead to significant changes in fiscal and Balance of Payment balances, and more broadly on macroeconomic stability.

The role of external shocks in explaining output fluctuations is contingent upon several factors such as policy stability and effectiveness. African countries were relatively resilient to the 2008 global financial and economic crisis due to the strong policy buffers that were established following sustained periods of prudent policy design and implementation prior to the onset of the crisis. However, these cushions that relatively minimized the impact of the 2008 global recession have weakened. Fiscal and external balances are weaker, debt levels are trending up, foreign exchange reserve levels are lower, and the fiscal space has diminished. For instance, Africa’s overall fiscal balance including grants has deteriorated from 4.6% of GDP in 2006 to -3.6% in 2011. The external account balance including grants has mirrored similar trends, worsening from 6.3% of GDP in 2006 to -0.6% in 2011. The high fiscal deficits limit the space for counter-cyclical measures in case of new external shocks.

3

This paper focuses on Uganda and seeks to examine the growth and welfare effects of macroeconomic shocks. We identify at least two reasons for pursuing this line of inquiry for Uganda.

First, identifying and quantifying the key external shocks and their importance in explaining output volatility are critical for informing policy responses to smooth out the effects of these shocks as well as structural reforms aimed at reducing Uganda’s exposure and vulnerability to these shocks.

Second, to examine the potential gains in terms of reduced volatility from implementing programs aimed at smoothing the impact of these shocks.

It is apparent that getting to the bottom of the sources of volatility is as important for Uganda as for developing countries especially given the significance of income fluctuations as well as the limited ability to hedge against these fluctuations. Just like any other typical developing country, Uganda possesses shallow financial markets and countercyclical fiscal and monetary policy responses are usually impeded by various structural bottlenecks. This constraints several of efforts aimed at smoothing out the impact of external shocks.

The rest of this paper is organized as follows. Section 2 reviews literature and identifies gaps and Section 3 presents the methodology, data used and simulations. In Section 4 we discuss the findings, while Section 5 concludes with policy recommendations.

2. SOME RELATED LITERATURE

Several studies have been undertaken to analyze the role of external shocks in explaining macroeconomic volatility more generally and fluctuations in output in particular. These studies find that negative external shocks have a significant adverse effects on short- and medium-run growth through their impact on aggregate demand, external and fiscal balances (Collier and Goderis, 2009; Berg et al., 2010). There is some evidence that these effects are asymmetric suggesting that while negative shocks necessarily reduce growth, the positive shocks do not necessarily positively impact on long-run growth, especially in resource-rich countries where institutions are weak (Collier and Goderis, 2007).

4

The impact of external shocks measured in particularly in terms of trade shocks on output volatility has been examined in the literature using various quantitative and multi-sector equilibrium approaches. Kose (2002) finds that world price shocks play an important role in driving business cycles in small open developing economies. His results confirm the results of earlier work by Mendoza (1995) and Kose and Riezman (2001).

Some studies have examined the relationship between terms of trade shocks and variations in in GDP growth using vector auto-regression (VAR) models. Ahmed (2003) applies a VAR model to analyze the sources of short-term fluctuations in the output for six Latin-American countries and finds that changes in the terms of trade and external output only play a moderate role in driving output fluctuations. Raddatz (2008), using a panel VAR model, also find that terms of trade changes have a small effect, explaining an estimated 13% of output volatility for a typical African country. Moreover, oil prices are found to play a relatively more important role as drivers of output fluctuations for oil exporting African countries. Broda (2004) also uses a panel VAR approach to examine the role of exchange rate policies in shielding economies from real shocks. He reports that terms of trade shocks can explain 30% of the real output volatility in the long run in countries characterized by fixed exchange rate regimes against 10% in countries with flexible exchange rate regimes.

Cross-country and panel data analysis has also been used to examine the relationship between terms of trade shocks and GDP volatility. Easterly and Kraay (2000) find a positive relationship between income volatility and terms of trade volatility in a cross-country analysis. Rodrik (1998) interacts terms of trade volatility with openness in a cross country study and finds that this variable has a positive effect on GDP volatility positively. However, Rodrik (2001) finds that the relationship between terms of trade volatility and GNP volatility is positive but insignificant for Latin American economies. Hausman and Gavin (1996) also focus on Latin American countries and their findings reveal that terms of trade shocks have a weaker effect on GDP volatility in countries with a more flexible exchange rate than in countries with a pegged the exchange rate regime, coming findings by Broda (2004). Di Giovanni and Levchenko (2008) use industry-level data and find that the risk content of exports is strongly positively correlated with the variance of terms of trade and that export specialization affects macroeconomic volatility.

5

Another set of studies have examined output volatility in partner countries as a potential determinant of domestic volatility. Ahmed (2003) uses a VAR model that includes the volatility of the aggregate real GDP of the eight largest trading partners as a measure for external shocks. Calderon et al. (2005) includes the standard deviation of the trade-weighted annual growth of the main trading partners. Bacchetta et al., (2009) we use the trade-weighted annual growth of all trading partners as a measure for external shocks. These studies draw the conclusion that output volatility in partner countries has a positive effect on exporters’ GDP volatility. Jansen et. al (2009) examine the role of foreign demand shocks in partner countries in explaining economic volatility in exporting countries. They find that the correlation between trading partners’ cycles is more important in determining exporter’s GDP volatility compared to the volatility of demand in individual export markets.

Collier and Goderis (2009) study the role of structural policies in mitigating the impact on growth from adverse commodity export price shocks and large natural disasters for 130 countries during the period 1963-2003. Their findings reveal that adverse commodity export price shocks and natural disasters substantially affect short-term growth but no find evidence of a long run impact of commodity export price shocks (volatility) on the level of GDP. This external commodity price shocks-growth nexus is found to display asymmetric behavior. While positive price shocks have a positive effect on growth, these effects are not significant. Collier and Goderis (2008) argue that this asymmetry is explained by a country’s productive capacity. In particular, if an economy is operating close to its short-term production potential, sudden upward shifts in export earnings will usually not raise aggregate output. However, sharp and sudden large decreases will normally reduce both export output and demand elsewhere in the economy and these effects will reduce aggregate output. The only exception here is if prices are highly flexible and resource allocation responds swiftly to the resulting price changes. Natural disasters have a negative and significant effect on output, but this effect is modest. Structural policies are found to play a key part in shaping the negative short-term growth effects from shocks. In particular, regulations that delay the speed of firm closure increase the short-term growth loss from adverse price shocks in commodity-exporting countries. Regarding natural disasters, labour market regulations that prevent an efficient re-allocation of workers tend to increase the negative effect on short-term growth from these shocks.

6

Dabla-Norris and Gunduz (2012) develop a vulnerability index to quantify low-income country risks to growth crises arising from external shocks. Two complementary approaches, multivariate regression analysis and a univariate ‘signaling” approach, are applied to aggregate information from a set of macroeconomic and institutional indicators into an ‘early warning’ vulnerability index. The index is then applied to assess vulnerabilities to growth declines in low-income countries during the period 1993-2011. Findings show that country fundamentals, exchange rate regimes, institutional quality, and the size of shocks are important in explaining growth crises in low income countries. Moreover, the sensitivity of crisis risks to varying policy and institutional fundamentals varies across countries. A strengthening of policy and institutional frameworks contributes to a larger reduction in the crisis probability in countries where buffers are initially weak. The study also finds that the vulnerability index has declined significantly following its peak in the early 1990s. This is explained by the improved policy and economic management as well as a more favorable external environment in particular the improvements in terms of trade and debt relief. These factors, Dabla-Norris and Gunduz (2008) argue have contributed to stronger policy buffers and thus reduced vulnerabilities to external shocks. However, these buffers have been increasingly expended to counter the impact of the global crisis leading to elevated growth crises risks in low-income economies to levels well above the pre-crisis years.

3. METHODOLOGY AND DATA

We use a recursive dynamic CGE model (Appendix 1 and 2) for Uganda based on the 2009/10 Social Accounting Matrix (SAM). We draw on a number of strengths from the CGE modeling framework in our analysis. Firstly, the model simulates the functioning of the economy as a whole and tracks changes in economic conditions and how these changes are transmitted through price and quantity adjustments on a range of markets. Secondly, since the basis of the CGE model is a SAM, we are able to discern the effects of the changes in economic conditions on individual sectors of the economy. Thirdly, the link of the model to household survey data enables an assessment of the impacts on the welfare of households. Finally, the recursive dynamic nature of our model implies that the behavior of its agents is based on adaptive expectations, rather than on the forward looking expectations that underlie inter-temporal optimization models. Since a recursive model is solved one period at a time, it is possible to separate the within-period component from the between-period component, where the

7

latter governs the dynamics of the model. The CGE model used in this study is based on a standard CGE model developed by Lofgren, Harris, and Robinson (2002) and adopted to Uganda by Economic Policy Research Center. This is a real model without the financial or banking system (See Table A1). GAMS software is used to calibrate the model and perform the simulations.

3.1 Social Accounting MatrixConsistent with other conventional SAMs, the 2007 Uganda SAM is based on a block of production activities, involving factors of production, households, government, stocks and the rest of the world i. The various commodities (domestic production) supplied are purchased and used by households for final consumption (42 per cent of the total), but also a considerable proportion (34 per cent) is demanded and used by producers as intermediate inputs. Only 7 percent of domestic production is exported, while 11 per cent is used for investment and stocks and the remaining 7 percent is used by government for final consumption.

Households derive 64 per cent of their income from factor income payments, while the rest accrues from government, inter-household transfers, corporations and the rest of the world. The government earns 32 percent of its income from import tariffs – a relatively high proportion, but a characteristic typical of developing countries. It derives 42 percent of its income from the ROW, which includes international aid and interest. The remainder of government’s income is derived from taxes on products (14 percent), income taxes paid by households (6 percent) and corporate taxes (5 percent).

3.2 Productions and commoditiesFor all activities, producers maximize profits given their technology and the prices of inputs and outputs. The production technology is a two-step nested structure. At the bottom level, primary inputs are combined to produce value-added output using a CES (constant elasticity of substitution) function. At the top level, aggregated value added is then combined with intermediate input within a fixed coefficient (Leontief) function to give the output. The profit maximization gives the demand for intermediate goods, labour and capital demand. The detailed disaggregation of production activities captures the changing structure of growth due to the pandemic.

8

The allocation of domestic output between exports and domestic sales is determined using the assumption that domestic producers maximize profits subject to imperfect transformability between these two alternatives. The production possibility frontier of the economy is defined by a constant elasticity of transformation (CET) function between domestic supply and export.

On the demand side, a composite commodity is made up of domestic demand and final imports and it is consumed by households, enterprises, and government. The Armington assumption is used here to distinguish between domestically produced goods and imports. For each good, the model assumes imperfect substitutability (CES function) between imports and the corresponding composite domestic goods. The parameter for CET and CES elasticity used to calibrate the functions used in the CGE model are exogenously determined.

3.3 Factors of productionThere are 6 primary inputs: 3 labour types, capital, cattle and land. Wages and returns to capital are assumed to adjust so as to clear all the factor markets. Unskilled and self-employed labor is mobile across sectors while capital is assumed to be sector-specific. Within the model, producers instantly adjust to changes in rates of returns for factors of production for each sector. The model does not take into account adjustment costs of switching resources between sectors.

3.4 InstitutionsThere are three institutions in the model: households, enterprises and government. Households receive their income from primary factor payments. They also receive transfers from government and the rest of the world. Households pay income taxes and these are proportional to their incomes. Savings and total consumption are assumed to be a fixed proportion of household’s disposable income (income after income taxes). Consumption demand is determined by a Linear Expenditure System (LES) function. Firms receive their income from remuneration of capital; transfers from government and the rest of the world; and net capital transfers from households. Firms pay corporate tax to government and these are proportional to their incomes.

Government revenue is composed of direct taxes collected from households and firms, indirect taxes on domestic activities, domestic value added tax,

9

tariff revenue on imports, factor income to the government, and transfers from the rest of the world. The government also saves and consumes.

3.5 Macro closureEquilibrium in a CGE model is captured by a set of macro closures in a model. Aside from the supply-demand balances in product and factor markets, three macroeconomic balances are specified in the model: (i) fiscal balance, (ii) the external trade balance, and (iii) savings-investment balance. For fiscal balance, government savings is assumed to adjust to equate the different between government revenue and spending. For external balance, foreign savings are fixed with exchange rate adjustment to clear foreign exchange markets. For savings-investment balance, the model assumes that savings are investment driven and adjust through flexible saving rate for firms.

3.6 Recursive dynamicsTo appropriately capture the dynamic aspects of aid on the economy, this model is extended by building some recursive dynamics by adopting the methodology used in previous studies on Botswana and South Africa (Thurlow, 2007). The dynamics is captured by assuming that investments in the current period are used to build on the new capital stock for the next period. The new capital is allocated across sectors according to the profitability of the various sectors. The labour supply path under different policy scenarios is exogenously provided from a demographic model. The model is initially solved to replicate the SAM of 2007.

3.7 Limitations of the modelCGE models have some weaknesses (Thurlow, 2008). The main criticism of the static model is that its core formulation is closely tied to the Walrasian ideal of equilibrium (Dervis et al, 1982). In a pure neoclassical setting, producers and consumers react passively to prices in order to determine their demand and supply schedules. Markets are therefore assumed to clear through the interaction of relative prices, such that equilibrium is achieved in both goods and factor markets. The model accommodates prices in relative terms and therefore cannot adequately address issues related to inflation.

3.8 SimulationsOur analysis is based on a series of scenarios each representing an exogenous change in economic conditions and are compared to a baseline scenario of business as usual. Running scenarios allows us to conduct a sort

10

of controlled experiment of various types of impacts. These impacts are then ascertained in terms of changes in: average sectoral growth patterns, various components of the external account, household welfare, and employment. These changes are then compared to the baseline. This baseline scenario assumes no specific changes to policy and the absence of an external shock. Consistent with the growth patterns in 2010, we calibrate the model to generate about 6.14% for real GDP growth under the baseline for the simulation period. The government finances its activities from domestic and foreign sources in a manner that is designed to be compatible with macroeconomic stability.

We compare the baseline to three simulations: (i) changes in terms of trade (tot); (ii) movements in international oil prices (poil); and (iii) changes in development assistance inflows (faid).

Under the ‘Terms of Trade’ (tot)simulation, the changes in terms of trade that are simulated here are informed by trends in price movements for major export and import commodities between 2010 and 2012. In particular, we simulate a 10% reduction in the average price of Uganda’s exports and a 10% increase in the average cost of imports.1 The key exports considered here include coffee, fish, tea, cotton, and flowers which jointly accounted for an estimated 65% of the total export receipts in 2012. The imports included in the simulation include the following products: petroleum, chemical, manufactured rubber, textiles and vehicles. These products accounted for an estimated 70% of total imports in 2012.

The second simulation (poil) examines the impact of changes in international oil prices. The 2012 World Economic Outlook (WEO) reports that international oil prices increased by 28% per barrel in 2010, 32% in 2011, and 10% in 2012. The 2012 WEO further projects a reduction in international oil prices by 4% in 2013 and we assume a similar reduction for 2014 and 2015.

Our third simulation – ‘Foreign Aid’ (faid) evaluates the impact of the 2012 aid-cut backs. We simulate a 50% reduction in aid in 2012, compared to the 2010 aid levels, following the suspension of development aid equivalent. In line with the FY 2013/14 budget which projects that 80% of public spending

1 Sensitivity analyses in which the changes in terms of trade are varied to 15% and 25% are conducted and the study’s conclusions are not affected.

11

will be financed by domestic resources, further reductions in aid of 75% (compared to the 2010 levels) are simulated for the period 2013-2015.

4. FINDINGS

4.1 Impact on GDP GrowthFigure 1 shows the average real GDP growth rates for the period 2010-2015 under the baseline and the three simulations considered here. The tot and faid simulations report average real GDP growth rates that are lower than the baseline. The baseline assumes no change the policy environment and framework of 2010 is maintained throughout the simulation period and also assumes no external shocks.

base tot poil faid5.525.545.565.585.605.625.645.665.685.705.72 5.70

5.59

5.71

5.65

Figure 1. Average real GDP Growth: 2010 - 2015 (%)

Source: Model simulations and author computations

The ‘poil’ simulations yields average real GDP growth rates similar to the baseline scenario during the simulation period. The similarity in real GDP growth rates for the ‘base’ and ‘poil’ could result from the contrasting impact of the increase in international oil prices (period 2010-2012) versus the project reduction in the same prices during the latter half of the simulations period (period 2013-2015). Thus, where these two opposing factors offset each other, changes in oil prices in this case will not affect trend real GDP growth, as is the case with simulations ‘poil’ and ‘base’ in this case. Two channels in particular could explain the inverse relationship between real GDP growth and international oil prices. First, petroleum products, particularly fuel, comprise core inputs in the production process, and thus changes in oil prices will directly affect productivity. Second, rising

12

oil prices are likely to put pressure on the non-farm wage rate and thus affect growth in the non-agriculture sector. It is plausible that the

2010 2011 2012 2013 2014 20150.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

Figure 2. Real GDP Growth (year-over-year, %)

Source: Model simulations and author computations

Figure 2 illustrates the trends in real GDP growth across the three simulations during the simulation period. Real GDP growth mirrors similar trends across the simulations and also when compared to the baseline. A more thorough assessment of the channels through which the three shocks considered here affects real GDP growth requires examining the trends in sectoral contributions to GDP growth. This impact is illustrated in Figures 3-5.

Figure 3 indicates that trends in the contribution to real GDP growth from the agriculture sector across the three simulations does not differ much and is also similar to the baseline simulation.

13

2010 2011 2012 2013 2014 20150.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

Figure 3. Agriculture - Contribution to real GDP growth (%)

Source: Model simulations and author computations

However, agriculture’s contribution to real GDP growth depicts an increase between 2010 and 2011 before reverting to the path traced by the other simulations for the reminder of the simulation period. This finding could be due to the impact of deterioration of terms of trade on agriculture exports and import substitutable agriculture products. In particular, the deterioration in terms of trade reduces foreign exchange receipts and all else the same, increases the import bill, leading to a depreciation of the local currency. The local currency depreciation makes exports cheaper and imports expensive leading to growth in exports including in the agriculture sector and growth in import substitutable sectors. These events combined could explain the increased contribution of agriculture to real GDP growth between 2010 and 2011. However, the drought driven food shortages and the resulting surge in food inflation experienced during the second half of 2011 appears to have more than offset the impact of terms of trade changes on agriculture sector growth thus leading to a reduction in this sector’s contribution to real GDP growth in 2012.

In the industry sector, two key findings emerge. First, the contribution of this sector to real GDP growth under the ‘faid’ simulation exceeds the baseline scenario and the two other simulations during the period 2011-2014 but seems to revert to ‘trend’ in 2015. Second, the sector’s contribution to real

14

GDP growth decreases markedly between 2010 and 2011 under the ‘tot’ simulation. The first observation (strong contribution of the industry sector to real GDP growth under the ‘faid’ simulation) can be explained by the depreciation of the real exchange rate associated with a reduction in development assistance, and indeed a reduction in external inflows. In particular, real exchange rate depreciation supports export-oriented and/or import substitutable manufacturing as imports become expensive while exports become more costly.

2010 2011 2012 2013 2014 20150.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

Figure 4. Industry - Contribution to real GDP growth (%)

Source: Model simulations and author computations

The second observation (industry sector’s contribution to real GDP growth decreases markedly between 2010 and 2011 under the ‘tot’ simulation) can also be explained the impact of terms of trade changes on export-oriented and import-substitutable industry sub-sectors. However, in this case, the deterioration in terms of trade leads to a depreciation of the real exchange rate via a reduction in foreign exchange receipts.

Figure 5 reveals that the services sector’s contribution to real GDP growth mirrors similar trends across the three simulations.

15

2010 2011 2012 2013 2014 20150.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

5.00

Figure 5. Services - Contribution to real GDP growth (%)

Source: Model simulations and author computations

However, two key results standout. The first key finding here is that the reduction in the sector’s contribution to real GDP growth between 2010 and 2011 is greatest under the ‘faid’ simulation and smallest under the ‘tot’ simulation. Reductions in external inflows, including development assistance contribute to a depreciation of the real exchange rate, all else the same. This tends to benefit, all else the same, growth in export-oriented and import-substitutable sectors. Thus, tradable sectors become more attractive as private investors seek to cash in on the reduced relative costs in the export sectors and/or in the import-competitive sectors as a result of the low real exchange rate and expensive import competition. The upshot of these factors is that private capital will tend to flow towards the tradable sectors including manufacturing and export oriented agriculture and away from the non-tradable sectors such as domestic oriented services. Thus given that the service sector is the least tradable, it tends to suffer most from the reduction in foreign inflows including external development assistance.

The second key finding relates to the contribution of the services sector to real GDP growth which appears to be higher under the ‘poil’ simulation during the period 2012 – 2015. Changes in oil prices have a direct

16

relationship with inflation. Thus, reductions in the international price of oil (as assumed under the ‘poil’ simulation for the period 2013-2015) will likely contribute to growth in the off-farm due to the consequent reduction in costs of production, all else the same.

In conclusion, the impact of the three external shocks considered in this paper impacts the three economic sectors (agriculture, industry and services) in diverse ways and in most cases, these impacts offset each other via the effects on the exchange rate, cost of production, and economic competiveness. This could explain the minimal deviations in real GDP trends between the baseline and the three simulations during the simulation period.

4.2 Impact on Exports and ImportsThe discussion in the preceding section is primarily anchored on the effect of the three simulations on the tradable sectors, in particular the export-oriented and import dependent sectors. In this section, we examine the trends in imports and exports across the three simulations and the possible drivers of these trends. As discussed earlier, Figure 6 also indicates that growth rate of imports decreases across all simulations between 2010 and 2011 followed by a gradual increase during the simulation period post-2012. Changes in the real effective exchange rate are among the key drivers of the trends depicted in Figure 6. Published data confirms that the Uganda shilling depreciated markedly against the US Dollar during the first three quarters of 2011 – leading to a slowdown in the rate of growth in imports – before appreciating 13% during the fourth quarter of 2011 – which explains the reversal in the decline of the growth rate for imports.

2010 2011 2012 2013 2014 20150.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

Figure 6. Growth in Real Imports (year-over-year, %)

Axis Title

17

Source: Model simulations and author computations

Figure 7 shows the trends in the export growth rate and these are also consistent with changes in the value of the local currency against the US Dollar and other major currencies. As would be expected, depreciation in the real effective exchange rate during the first three quarters of 2011 resulted in an increase in the export growth rate with the 2011 Q4 appreciation and subsequent stability of the local currency during 2012 contributing to a slowdown in the growth rate of exports.

2010 2011 2012 2013 2014 20150.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

Figure 7. Real Exports Y.o.Y Growth (%)

Axis Title

Source: Model simulations and author computations

In summary, our findings confirm that the impact of the three external shocks studied here is transmitted to GDP growth through various channels including changes in the real effective exchange rate and its effect on imports and exports.

4.3 Effects on Household Welfare

Figures 8 and 9 illustrate the impact of the three external shocks on household welfare as measure by the equivalent variation and changes in real per capita consumption respectively. Both these measures indicate household welfare declines in 2011 following the on-set of the shocks and but gradually increases during the simulation period post-2012. However, household welfare, as measured by the equivalent variation, remains permanently below ‘trend’ under the ‘tot’ and ‘faid’ simulations. As

18

discussed earlier, the appreciation of the exchange rate post-2011 increases imports and hurts exports as well as the import-substitutable sectors as imports become cheaper. Consequently, the tradable sectors become less attractive owing to the increased costs in the export-oriented and import competitive sectors due to the appreciated real effective exchange rate and less expensive imports. Thus, private capital is driven towards the non-tradable sectors such as construction and domestic-oriented services as opposed to manufacturing including food processing.

2009 2010 2011 2012 2013 2014 2015

-2.00

0.00

2.00

4.00

6.00

8.00

10.00

Figure 8. Equivalent Variation per capita (% of base year consumption spending per capita) - Household

Axis Title

Source: Model simulations and author computations

Moreover, the average growth rates for wages for labor all three labor categories examined here (less-than-secondary, secondary, and tertiary education) are below the baseline growth rates for the ‘tot’ and ‘faid’ simulations. This the reduced investments and the resulting lower productivity in the tradable sectors and lower growth labor remuneration under these two simulations explains the lower household welfare.

19

2009 2010 2011 2012 2013 2014 2015

-3.00

-2.00

-1.00

0.00

1.00

2.00

3.00

Figure 9. Growth in real per-capita Consumption (year-over-year, %) - Household

Source: Model simulations and author computations

In summary, the external shocks examined via the three simulations in this paper affect household welfare at the on-set of the shocks in 2011 and also result in lower welfare under two simulations for the entire simulation period.

5. CONCLUSIONS

The role of external shocks in explaining output fluctuations is contingent upon several factors such as policy stability and effectiveness. African countries were relatively resilient to the 2008 global financial and economic crisis due to the strong policy buffers that were established following sustained periods of prudent policy design and implementation prior to the onset of the crisis. However, these cushions that relatively minimized the impact of the 2008 global recession have weakened. Fiscal and external balances are weaker, debt levels are trending up, foreign exchange reserve levels are lower, and the fiscal space has diminished. Using a computable general equilibrium model for Uganda, this study seeks to examine the growth and welfare effects of three macroeconomic shocks: changes in terms of trade, changes in the price of oil, and changes in development assistance inflows.

Our analysis reveals three key findings. First, the opposite and often offsetting effects of the three external shocks examined in our paper on the agriculture, industry and the services sectors yields only short-term as opposed to permanent deviations from trend growth in real GDP. Second, the primary channel through

20

which these three external shocks are transmitted to the domestic economy and GDP growth is via changes in the real effective exchange rate and the associated effect on the non-tradable (domestic) and tradable (both import-dependent and/or export-oriented) sectors. Third, household welfare decreases permanently during the entire simulation period for two of the three simulations considered by this study.

These findings should be of interest to a diverse audience of policy makers both in Uganda and the region for at least two reasons. First, by identifying and quantifying the impact of key external shocks in explaining output volatility is critical for informing policy responses to smooth out the effects of these shocks as well as structural reforms aimed at reducing Uganda’s exposure and vulnerability to these shocks. Examining the potential gains in terms of reduced volatility from implementing programs aimed at smoothing the impact of these shocks has the potential of informing the prioritization of reforms.

In particular, our findings – and based on the assumptions that we make in this paper – reveal that policy makers should focus on designing and implementing policies that reduce the severity of external shocks on the vulnerable and poor. Such policies as social safety nets and micro insurance including in the agriculture and non-tradable services sub-sectors and construction have the potential of offsetting the permanent effect of external shocks on household welfare. Moreover, due to Uganda’s long standing structural imbalance between imports and exports, policies to diversify economic activity and exports in particular remain critical to ensuring stability of the exchange rate. Exchange rate stability will contribute to the closure of one of the primary channels through which external shocks are transmitted to the domestic economy.

21

APPENDICES Appendix 1: CGE Model Sets, Parameters, and VariablesSymbol Explanation Symbol ExplanationSetsa A Activities ( )c CMN C Commodities not in CM

( )a ALEO A Activities with a Leontief function at the top of the technology nest

( )c CT C Transaction service commodities

c C Commodities ( )c CX C Commodities with domestic production

( )c CD C Commodities with domestic sales of domestic output f F Factors

( )c CDN C Commodities not in CD i INS Institutions (domestic and rest of world)

( )c CE C Exported commodities ( )i INSD INS Domestic institutions

( )c CEN C Commodities not in CE ( )i INSDNG INSD Domestic non-government institutions

Aggregate imported commodities ( )h H INSDNG Households

Parameters

ccwts Weight of commodity c in the CPI cqdst Quantity of stock

changecdwts Weight of commodity c in

the producer price index cqg Base-year quantity of government demand

caica Quantity of c as intermediate input per unit of activity a cqinv

Base-year quantity of private investment demand

'ccicdQuantity of commodity c as trade input per unit of c’ produced and sold domestically

ifshifShare for domestic institution i in income of factor f

'cciceQuantity of commodity c as trade input per exported unit of c’ 'iishii

Share of net income of i’ to i (i’ INSDNG’; i INSDNG)

'ccicmQuantity of commodity c as trade input per imported unit of c’ ata Tax rate for activity a

aintaQuantity of aggregate intermediate input per activity unit itins

Exogenous direct tax rate for domestic institution i

aivaQuantity of aggregate intermediate input per activity unit itins01

0-1 parameter with 1 for institutions with potentially flexed direct tax rates

imps Base savings rate for domestic institution i ctm Import tariff rate

imps010-1 parameter with 1 for institutions with potentially flexed direct tax rates ctq Rate of VAT tax

cpwe Export price (foreign currency) i ftrnsfr Transfer from factor f

to institution icpwm Import price (foreign

currency)aa Efficiency parameter in the CES

activity function CET function share parameter

22

vaa Efficiency parameter in the CES

value-added functionvafa CES value-added function share

parameter for factor f in activity aacc Shift parameter for domestic

commodity aggregation functionmch

Subsistence consumption of marketed commodity c for household h

qc Armington function shift

parameter ac Yield of output c per unit of activity a

tc CET function shift parameter a

a CES production function exponent

Capital sectoral mobility factor vaa CES value-added function

exponentmch

Marginal share of consumption spending on marketed commodity c for household h

acc Domestic commodity aggregation

function exponentaa CES activity function share

parameterqc Armington function exponent

acac Share parameter for domestic

commodity aggregation functiontc CET function exponent

Armington function share parameter Sector share of new capitalCapital depreciation rate

Exogenous VariablesCPI Consumer price index MPSADJ Savings rate scaling factor (= 0 for

base)

DTINSChange in domestic institution tax share (= 0 for base; exogenous variable) fQFS Quantity supplied of factor

FSAV Foreign savings (FCU) TINSADJ Direct tax scaling factor (= 0 for base; exogenous variable)

GADJ Government consumption adjustment factor faWFDIST Wage distortion factor for factor f

in activity aIADJ Investment adjustment factorEndogenous Variables

Average capital rental rate in time period t cQG Government consumption demand

for commodity

DMPSChange in domestic institution savings rates (= 0 for base; exogenous variable) chQH Quantity consumed of commodity

c by household h

DPI Producer price index for domestically marketed output achQHA

Quantity of household home consumption of commodity c from activity a for household h

EG Government expenditures aQINTA Quantity of aggregate intermediate input

hEH Consumption spending for household caQINT Quantity of commodity c as

intermediate input to activity aEXR Exchange rate (LCU per unit of

FCU) cQINV Quantity of investment demand for commodity

GSAV Government savings Quantity of imports of commodity c

faQF Quantity demanded of factor f from activity a

iMPSMarginal propensity to save for domestic non-government institution (exogenous variable)

cQQQuantity of goods supplied to domestic market (composite supply)

aPA Activity price (unit gross revenue) cQT Quantity of commodity

demanded as trade input

cPDDDemand price for commodity produced and sold domestically aQVA Quantity of (aggregate)

value-added

23

cPDSSupply price for commodity produced and sold domestically cQX

Aggregated quantity of domestic output of commodity

Export price (domestic currency) acQXAC

Quantity of output of commodity c from activity a

aPINTA Aggregate intermediate input price for activity a Real average factor priceUnit price of capital in time period t TABS Total nominal absorptionImport price (domestic currency) iTINS Direct tax rate for

institution i (i INSDNG)

cPQ Composite commodity price 'iiTRIITransfers from institution i’ to i (both in the set INSDNG)

aPVA Value-added price (factor income per unit of activity) fWF Average price of factor

cPX Aggregate producer price for commodity fYF Income of factor f

acPXAC Producer price of commodity c for activity a YG Government revenue

aQA Quantity (level) of activity iYI Income of domestic non-government institution

cQD Quantity sold domestically of domestic output ifYIF Income to domestic

institution i from factor fQuantity of exports Quantity of new capital by

activity a for time period t

24

Appendix 2: CGE Model Equations

Production and Price Equations

(1)

(2)

(3)

(4)

(5)

(6)

(7)(8)(9)(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17)

25

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

Institutional Incomes and Domestic Demand Equations

(28)

(29)

(30)

(31)

(32)

(33)

(34)

(35)

(36)

System Constraints and Macroeconomic Closures

26

(37)

(38)

(39)

(40)

(41)

(42)

(43)

Capital Accumulation and Allocation Equations

(44)

(45)

(46)

(47)

(48)

(49)

REFERENCES

27

Ahmed, S. 2003. “Sources of Macroeconomic Fluctuations in Latin America andImplications for Choice of Exchange Rate Regime”. Journal of Development Economics,72: 181-202.

Bacchetta, M., Jansen, M., Lennon, C., and Piermartini, R. (2009) Exposure to External Shocks and the Geographical Diversification of Exports. Newfarmer, R., Shaw, W., and Walkenhorst, P. Breaking into New Markets: Emerging Lessons for Export Diversification. Washington DC, World Bank.

Berg, A., Papageorgiou, C., Pattillo, C., Schindler, M., Spatafora, N., and Weisfeld, H. 2011. “Global Shocks and their Impact on Low-Income Countries: Lessons from the Global Financial Crisis.” IMF Working Paper 11/27

Broda, C. (2004) Terms of trade and exchange rate regimes in developing countries. Journal ofInternational Economics 63[1], 31-58.

Caballero, R.J. and Krishnamurthy, A. 2001. “International and Domestic CollateralConstraints in a Model of Emerging Market Crises.” Journal of Monetary Economics48(3): 513-548

Calderon, C., Loayza, N., and Schmidt-Hebbel, K. (2005) Does openness imply greater exposure? World Bank Policy Research Working Papers 3733. Washington DC, World Bank.

Collier, P. and Goderis, B. 2009. “Structural Policies for Shock-Prone Developing Countries.” University of Oxford, CSAE WPS/2009-03

Dabla-Norris, E. and Gunduz, Y. B. 2012. “Exogenous Shocks and Growth Crises in Low-Income Countries: A Vulnerability Index”. IMF Working Paper 12/264

Di Giovanni, J. and Levchenko, A. (2008) Trade openness and volatility. Review of Economics and Statistics.

Easterly, W., Kremer, M., Pritchett, L., and Summers, L. 1993. “Good Policy or GoodLuck?: Country Growth Performance and Temporary Shocks.” Journal of MonetaryEconomics 32(3): 459-483.

28

Easterly, W. and Kraay, A. (2000) Small States, Small Problems? Income, Growth, and Volatility in Small States. World Development 28[11], 2013-2027.

Hausmann, R. and Gavin, M. (1996) Securing stability and growth in a shock prone region: the policy challenge for Latin America. Inter-American Development Bank Working Paper 315.

Hnatkovska, V., and N. Loayza. 2005. “Volatility and Growth,” in Managing Volatility,edited by Joshua Aizenmann and Brian Pinto. Cambridge University Press. 65-100.

Jansen, M., Lennon, C., and Piermartini, R. 2009. “Exposure to External Country Specific Shocks and Income Volatility.” Economic Research and Statistics Division, World Trade Organization. Geneva

Kose, M. A. (2002) Explaining business cycles in small open economies: [`]How much do worldprices matter?'. Journal of International Economics 56[2], 299-327.

Kose, M. A. and Riezman, R. (2001) Trade shocks and macroeconomic fluctuations in Africa. Journal of Development Economics 65[1], 55-80.

Loayza, Norman and Claudio Raddatz. 2007. The Structural Determinants of External Vulnerability. The World Bank Economic Review, 21(3): 359–387. Oxford University Press

Mendoza, E. G. (1995) The terms of trade, the real exchange rate and economic fluctuations.International Economic Review 36, 101-137.

Raddatz, C Claudio. (2005). “Are external shocks responsible for the instability of output in low income-countries?” World Bank, Policy Research Working Paper 3680.

Raddatz, Claudio. 2008. Have External Shocks Become More Important for Output Fluctuations in African Countries? World Bank.

Rodrik Dani. 1998. Where Did All The Growth Go? External Shocks, Social Conflict, and Growth Collapses. John F. Kennedy School of Government. Harvard University

Rodrik, D. (2001) Why is there so much economic insecurity in Latin America ? Cepal Review 73, 7-29.

29

30

ENDNOTES

31

i The first SAM for Uganda was developed in 2002 by the Uganda Bureau of Statistics and later

updated by the International Food Policy Reserach Institute in 2007. The 2007 SAM reflects the

current production structure of Uganda which has not changed significantly over the past four

years.