The Arithmetic of Debt Sustainability and Its Fiscal policy Implications: the Case of Ethiopia

 

 

 

 

 

Haile Kebret (Ph.D.)

EEA/EEPRI

 

 

 

 

 

 

 

 

 

Working Paper No. 1/05

EEA/Ethiopian Economic Policy Research Institute

March 2005

Addis Ababa

 

 

 

(Preliminary Draft, do not Quote)

 

 

 

 

 

 

 

 

 

 

 

Abstract: The objective of this paper is to examine the sustainability of Ethiopian public debt and to infer its fiscal policy implications. It evaluates debt sustainability using conventional techniques of stationarity and co-integration. It further evaluates the fiscal policy implications of debt relief under different donors’ behavior and growth scenarios. The results suggest that while conventional tests indicate that the Ethiopian debt is sustainable but these tests are inadequate and do not address the fiscal implications of the burden of debt and its inter-temporal trade offs. When such issues are incorporated, the opportunity cost of the debt is significant.

 

 

 

 

 

 

 

 

 

 

 

1.       Introduction

 

 

The interaction between government revenue and expenditures and their net inflows that determine the growth of the accumulated debt stock has been of serious macroeconomic concern in recent years. Among the main reasons for this concern are the following: First, the amount of debt that has been accumulated by some poor countries is huge relative to the size of their economies, as measured relative to their GDP; second, the recent economic growth performance of these countries has been modest, at best, to make the accumulated debt sustainable; third, even if they were willing to pay, the opportunity cost of doing so would have severe socio-economic and possibly political consequences in these countries; and, fourth, due to the above reasons the international pressure on lenders to grant debt relief has been mounting in recent years, spearheaded by institutions like the Jubilee 2000.

 

 The above factors have prompted the donor community in general and the multilateral institutions in particular, to design a scheme whereby poor countries that fulfill certain conditions will be granted a debt relief under what is known as the HIPc initiative[1].  Including Ethiopia, about 27developing countries, most of them in Africa, have qualified to be covered under this initiative. As it is well known by now, detailed evaluation of the initiative and the conditions under which these countries qualified for such a grant will not be made here. Suffice it to say that, being poor with a huge accumulated debt and carrying out prescribed policy reforms that are deemed necessary for growth and poverty reduction are the general prerequisites to qualify. Hence, the expected outcomes of such debt relief are adequate economic growth performance that would enable countries to achieve a sustainable debt burden following a partial debt relief and at the same time to reduce poverty in their respective countries via what are termed pro-poor policies.     

Accordingly, owing to its huge debt burden, severe poverty and willingness to carry out the prescribed reforms, Ethiopia became eligible for debt reduction under the HIPC initiative and reached this year (2004) the final stage or what is termed the ‘completion’ point. According to IMF figures, its total nominal debt stock at the end of fiscal year 2002/2003 was about $6.8 billion (or $4.5 billion in net present value (NPV) terms). This debt is huge relative to the size of the economy and the performance of exports. That is, this constitutes about 100% of GDP or about 246% of exports. Even after the scheduled debt relief under the initiative is applied, the IMF projects that Ethiopia’s outstanding debt in NPV terms will be reduced to $3.9 billion or to about 90% of the current GDP in nominal terms.

  

The most important rationale of the HIPC initiative is that once these countries are granted the partial debt relief, they will achieve a sustainable debt burden. The objective of this paper is, therefore, to (a) examine whether the existing stock of total (both domestic and foreign) debt is sustainable during the post-HIPC era, and (b) to compute the fiscal path that traces debt sustainability under alternative growth scenarios and its policy implications.

 

 The remainder of the paper is organized as follows.  Following this brief introduction, Section Two highlights some basic macroeconomic attributes and outlines the historical evolution of debt in Ethiopia. Section Three presents the models that will be used in analyzing debt sustainability and the primary government budget balance that is compatible with a sustainable debt. This section will also present the summary of the results related to debt sustainability. The results and the policy implications of the findings regarding the required primary balance to ensure debt sustainability are discussed in Section Four. A brief summary and conclusions of the study are presented in Section Five.

 

 

2. Some Basic Macroeconomic Attributes and Evolution of Debt in Ethiopia

 

2.1 Some Basic Macroeconomic Attributes

In terms of broadly defined economic development, Ethiopia lags behind in almost every measure. Socio-economic development could be examined in absolute terms, relative to other countries or in terms of improvements overtime (dynamic changes). Regardless of the approach one takes, Ethiopia is one of the least developed countries in the world and its economic structure has not changed much over the years. Both broad social indicators and narrow economic aggregates, support this observation. For instance, both the structure of the economy and the provision of social services indicate that the level of development of the country is at a very low level. Using the conventionally used measure of poverty ( 1$ a day),  about 44 per cent of Ethiopia's population falls under the poverty line, which is close to the African Average but very low by  LDCs standards. According to World Bank Development indicators (2004), if one uses the $2 a day measure, the percent of population that falls under the poverty line exceeds 80%.

 

Despite some improvements in the overall macro-economic performance in the last decade (GDP growth averaging about 5%), the majority of its people still live in abject poverty. With a GNP per capita of US $100 (compared to an average of 480 in Sub-Saharan Africa and 520 for all LDCs), a life-expectancy at birth of 43 years (relative to 51 and 63 years in Sub-Saharan Africa and all LDCs, respectively), with only 26 per cent of its population having access to safe water (compared to 47 and 74 per cent for Sub-Saharan Africa and all LDCs, respectively) Ethiopia is very poor by any standard. Due to the above and other socioeconomic indicators, Ethiopia ranks bottom even among the least developed countries according to the UNDP’s recent Human Development Indices.

 

Further, the dominant sector of the economy is traditional agriculture while industry which is believed to be the engine of economic growth is at its rudimentary level. All the remaining sectors are also weak. Consequently, the ability of the economy to mobilize resources to save and invest is limited. And resource limitations coupled with inefficient use of the meager available resources have made the economy incapable of fully financing its recurrent and capital expenditures. Therefore, the economy is in essence dependent on foreign assistance and loans both to finance its food deficiency and other development expenditures.

     

In terms of specific macroeconomic aggregates, as noted in Appendixes 1 and 2, the share of saving in GDP is very low both in absolute terms and relative to investment. This indicates the subsistence nature of the economy and the attendant resource gap (ranging between 9 to 20% of GDP in the last ten years). Such a resource gap in turn suggests that the country is dependent on external sources to finance this gap. Consequently, the Ethiopian economy has had a persistent deficit in its balance of payments and accumulated an external debt that is huge relative to its GDP. The accumulation of such a huge debt implies that in the absence of debt relief or an export boom its ability to service its debt (which is at times as high as 40 per cent of its exports) will render any meaningful domestic investment effort.     

 

In addition to the saving-investment gaps noted above, the country’s external sector and its fiscal balance are also weak. In particular, the economy has suffered a consistent budget deficit and balance of payments crisis over the years. For instance, the size of the budget deficits (excluding grants) and the current account deficit (excluding transfers) averaged 8.8 and 8.4% of GDP, respectively, between 1995/96 and 2002/2003. Due to all the above resource shortfalls the country has been dependent on external financial flows in the form of loans, grants and aid. Consequently, as noted above, the country has accumulated a huge amount of both external and internal debt over the years

 

2.2. Historical Evolution of Ethiopian Debt

As noted above, Ethiopia is one of the highly indebted countries, even by the standards of HIPCs of Sub-Saharan Africa.  Even though slightly decreased in recent years, owing to some cancellation and rescheduling, and mainly due to the granted debt relief amounting to US$1.3 billion (in NPV terms) under the HIPC initiative, the remaining balance is still high relative to GDP. The largest share of this debt is owed to the World Bank Group (IDA). Specifically, out of the total stock of debt outstanding, about 65 per cent is owed to the IDA group of creditors.

 

 To put it in a historical context, the size of the debt and its composition has changed since the mid 1970s. During the Imperial regime the size of the debt was modest. The magnitude of the debt in 1975 when the Imperial regime fell was only US$371 million. But by the end of 1991 (the time the present government took power) it reached US$8790 million. More than half (US$4744 million or 54%) of the total debt was contracted for defense purposes. Consequently, the major share (76.4 per cent) of the debt was owed for bilateral creditors in which the Former Soviet Union alone accounted for about 78 per cent of the total bilateral debt. In contrast to the composition of the present debt, the share of Multilateral Institutions in the total debt was only 16.8 per cent during the previous regime (Teklu, 2000).

 

Clearly, the debt contracted until the 1990s was largely used for defense purposes and helped neither in improving the productive capacity of the economy nor in alleviating poverty. The macroeconomic performance indicators attest to this observation as reflected in negative growth rate in GDP per capita, huge external imbalances and very low Human Development Indices during the period.

 

To cope with its unsustainable debt, Ethiopia has been engaged in negotiations with its creditors since 1992. Consequently, a total of US$ 372.89 million was agreed to either be canceled (US$ 101.60 million) or rescheduled (the remaining US$ 271.29 million). In addition to these ‘London Terms’ or Enhanced Torornto Terms’, the second round negotiations with the Paris Club Creditors for debts contracted before 1989 resulted in debt reduction on ‘Naples Terms’. Accordingly, a 67 per cent reduction on the net present value of the eligible debt was applied, which resulted in a total debt relief of US$164.8 million (constituting 24.3 per cent cancellation and 75.7 per cent rescheduling). Hence, the total debt relief obtained through the two negotiations amounted to US$537.7 million (consisting of 26.3% cancellations and 73.7% rescheduling). Owing to the HIPC initiative, recent IMF figures (IMF, 2004) show that to date Ethiopia has benefited a debt relief amounting to US$1275 million in NPV terms.

The questions that this paper poses are then, (1) after all the debt relief is granted, will the Ethiopian debt be sustainable?  (2) Will this sustainability depend on extending concessional (subsidized) loans beyond the HIPC era? And (3) what are the fiscal and social (such as poverty reduction) implications of attempting to achieve debt sustainability?  These are crucial questions for a country like Ethiopia whose economic capacity to finance basic social provisions is limited and has been dependent on foreign financial flows to meet even its basic needs.

 

3. Model and Estimation

3.1. Modeling sustainability

The standard formulation of debt sustainability starts from the basic government budget constraint which could be written as:

 

Dt+1 = (1+ρ) D t  + Gt  -Rt                                                             (1)

 

Where D is the level of outstanding public debt, ρ is the real interest rate, G and R are real government expenditure and revenue including seignorage, respectively.

Solving for Dt and taking expectations, (1) becomes:

 

Dt  = -E Σj=0 (1+ ρ)-(j+1)(Gt+j -R t+j )+ limj→∞ Et (1+ ρ)-(j+1)  Dt+j+1               (2)

 

(2) is the conventional inter-temporal government budget constraint which simply states that the outstanding debt in period t must equal the expected sum of the discounted value of budget deficit and the limit term which accounts for the discounted value of the debt in some future period.  

 

In the recent research literature, testing for sustainability of debt proceeded along two lines: one focusing on the flow and the other on the stock components of equation (1). The first approach (for instance used by Hamilton and Flavin (1986), Kreamers (1988), Wilcox(1989), Haug (1991) and Crosetti and Roubini (1991)),  focused on testing whether the limit term converges to zero or not. This has been examined using a unit root test to see whether the discounted debt stock is stationary or not whereby stationarity of the series is interpreted as indication of sustainability.

 

The second approach starts from the proposition that for the stock of debt to converge to zero, the flow or the budget balance must on average be zero. This suggests that the necessary and sufficient condition for debt sustainability is for government revenue and expenditures to be co-integrated.  As in Trehan and Walsh (1988, 1991), Hakkio and Rush (1991), Arghyrou (2003) the typical model specified in such analyses takes the following form.

 

Rt = α + βG t + ut                                                          (4)

 

Where R and G are as defined above, u is a white-nose error term and α, β are coefficients. 

Even though in principle equation (4) could be estimated in many ways, Arghyrou (2003, p. 6) favors using Dynamic OLS (DOLS). He argues that DOLS “is asymptotically equivalent to Johansen’s (1988) maximum-likelihood estimator and is known to have a superior performance in small samples”. The main advantage of the Stock and Watson (1993) or what is known as the DOLS model is that because of the lags and leads that are included, it captures any feed back the dependent variable might have on the independent variable(s) and hence ensures consistency of estimates. Accordingly, the usual equation estimated including in this study takes the following form.

                        k

Rt  =  α + βG t  + Σ  γi ΔGt-i +u t                                                                  (5)

                            i=-k

 

A tests for the existence of co-integration (or there lack of) between Rt and Gt indicates whether a given debt is sustainable or not. That is, if the two flow variables are co-integrated, a debt is said to be sustainable.  Alternatively, once it is established that the variables are co-integrated, sustainability could be further tested using an Error-Correction formulation and checking the significance of the error correcting term.

                    m                        j

Δ Rt =  δ +  Σ Ψi ΔRt-n + Σ  γi ΔGt-n +kΦ t-1 + vt                           (6)

                   t=1                    n=1

 

Where Φ is the error-correcting term, δ, Ψ, γ, and k are respective coefficients, and Δ is first difference operator.

Hence, if k in (6) is significant it suggests that debt is sustainable otherwise it indicates lack of equilibrium and, therefore, un sustainability of debt.

 

3.2. Estimation Results

In this study, both equation (5) and (6) are estimated for Ethiopia for the period 1965 to 1998. In short, as could be seen from the various test-results in appendixes 3 to 6, the findings could be summarized as follows. First, real government revenue and government expenditures are I(1) variables. Second the error term is stationary. Third, estimating (6) suggests that the error-term is significantly different from zero with a plausible speed of adjustment of abut 68% per year. All these results suggest that according to these findings, the Ethiopian debt is sustainable.

 

The weakness of such results in addressing debt sustainability issues is that they don’t address the time frame in which such sustainability will be achieved and do not take into account the fiscal policy requirements to achieve such sustainability. That is, they only focus on the co-movements of government revenue and expenditure to determine sustainability with the implicit assumption that there will be no shock that affects the behavior of the variables in the future. In short, they take the statuesque of the budget structure as given for the indefinite future[2]. But as is well known, most government expenditures in countries like Ethiopia are financed from foreign sources. Therefore, an increase or a decrease of such assistance will have an impact on the fiscal policy stance. In particular, one could envisage a scenario during the post-HIPC era in which access to subsidized loans or aid will be reduced once debt relief is granted. The relevant question to ask is, then, will Ethiopia be able to sustain its debt in the absence of such assistance? What will be the degree of indebtedness and the required fiscal policy path under different growth and foreign assistance scenarios that would ensure debt sustainability? The next sub-section attempts to address these questions.

    

3.1. Debt Sustainability, Fiscal Policy Path and Debt Relief

 

To address the above questions, this study follows the model developed by Edwards (2002) to examine the required fiscal policy path to achieve sustainability of the total public sector debt.  His model is suited to project the fiscal stance which is consistent with a sustainable public sector debt during the post-HIPC era.

Edawards (2002) started from the basic conventional debt accumulation equation, which states that the changes to the accumulated debt (ΔDt ) at any point in time is equal to the interest payment on foreign debt (r*DF t-1)  and domestic debt (rDD t-1), plus the primary government balance less the change in the monetary base (ΔBt ) , which is used as a proxy for seignorage revenue.  That is,

 

ΔDt  = {r*DFt-1 + rDDt-1 } + pb t - ΔBt                                    (7)

 

r*, r are nominal interest rates on foreign and domestic debt respectively. And DD could be interpreted as commercial debt, be it foreign or domestic while DF refers to foreign debt obtained on concessional (or subsidized) terms. 

 

The variable of interest in (7) is the government primary balance (pb t ). That is, what is the primary balance which is consistent with a sustainable debt burden after the debt forgiveness is granted? Edwards (2002, p. 5) defines debt sustainability as “a situation where increases in each type of debt are in line with the pace at which national and international creditors desire to accumulate government-issued securities”.  And since the flow of both domestic and foreign loans have an upper  limit, they are assumed to behave as follows: During the post- HIPC era, international donors will increase concessional loan by an amount equal to or less than θ, and domestic creditors are willing to increase their credit by an amount equal to β. As an upper limit, creditors are assumed to increase their lending by an amount equal to the real growth rate of GDP (g) and the dollar inflation (target) rate (π*). That is, the limits of θ and β are:

 θ ≤ (g +π*) ; and  β ≤ (g +π*).                                     (8)

Given the above basic relationships, the dynamic path of the sustainable primary government balance could be written as:

 

{ pb t /Y t) = [{θ –r*}(DF0 /Y0)e(θ-g- π*)(t-1) +

{β-rt}(DD0 /Y0) e(β -g- π*)(t-1)]

                             [1/(1+g+ π*)]-( g + π)(B0 / Y0).                    (9)

 

Similarly, the steady-state sustainable primary balance could be written as follows[3]:

 

{ pb  /Y ) = {g+ π*–r}(DD0 /Y0)         [1/(1+g+ π*)]+( g + π)(B0 / Y0).     (10)       

 

DF0 /Y0 is the initial ratio of the face value of concessional loan to GDP

DD0 /Y0 is the initial domestic debt to GDP ratio

π is the target rate of domestic inflation

B0 / Y0 is the initial ratio of base money to GDP

t0 should be interpreted as the time following the period after all the HIPC initiative debt reductions are carried out. 

 

Clearly, the sustainable primary balance that is consistent with a sustainable debt is determined by both initial ratios of domestic and foreign debts to GDP, nominal domestic and foreign interest rates, domestic and foreign inflation rates, the rate of growth of real GDP, and the sustainable increases in both foreign and domestic debt (θ and β).

             

Given the above basic relationship between government primary balance and debt outstanding, it is possible to conjure up various scenarios regarding the likely behavior of the determinants of debt sustainability.  Among others, just to name a few, it is possible to consider different credit flows from the donor community, variations in GDP growth rates, changes in both foreign and domestic interest rates[4] and inflation rates, and changes in the domestic exchange rate which may affect the domestic inflation rate if there is a substantial pass through to the domestic economy.

 

As a first step, this study is limited to considering the impact of different assumptions on the flow of concessional credit once the HIPC initiative ran its course under plausible different GDP growth scenarios[5].

In particular, the study will examine the following issues.

(1)What will be the dynamic path of sustainable primary balance to GDP ratio if:

(a) the donor community decreases its credit facility to zero in the post-HIPC era once              debt forgiveness is granted?

(b) The flow of concessional loans in the post-HIPC era is not zero, but

 some new partial funding is still forthcoming? and

(c) The flow of foreign financial flows continues at a rate that prevails today?

(2) Under the above three scenarios, that range from the least favorable to the most favorable, the primary balance required to achieve debt sustainability will be examined for different plausible GDP growth rates.

(3) Further, the evolution of Ethiopia’s debt under the first two scenarios will be examined to evaluate the speed at which it converges to a steady state. The third case will not be computed since it is trivially the initial value by construction.     

 

4. Dynamic Path of a Sustainable Primary Balance 

The above outlined scenarios are evaluated using the recent Ethiopian data. The parameter values used to evaluate the alternative scenarios are summarized below. It has to be noted that some of the parameters chosen are on the low side due to the volatility of the values and owing to expectations that their future values will be less than what they are today.

Summary of parameters and values used for simulation

Variable

Value

Explanations

Pb t /Y t

…

Primary balance to GDP (to be computed).

θ 

…

Values vary depending on assumptions. 

r*

0.03

The approximate interest rate for concessional laons

DF0 /Y0

0.8

Ratio of foreign debt to GDP

g

 

Different growth rates (ranging from 2 to 10%) are used to reflect its volatility  

π*

0.025

Since most debt is denominated in US$ and inflation in the US averaged around 2.5% in recent years.

β

g +π*

Assumed a constant domestic debt burden equal to this rate.

rt

0.085

The recent commercial lending rate in Ethiopia is about 8.5%

DD0 /Y0

0.4

The domestic debt is about 40% of GDP.

π

0.085

Since this is dollar denominated target domestic inflation, it is taken as a sum of average depreciation rate of the Birr (6%)+US inflation (2.5%).

B0 / Y0

0.2

Recent ratio of base money to GDP.

 

Using the above parameters, the first case considered is case A, in which it is assumed that in the post HIPC era, no additional new funding will be forthcoming once the debt relief under the HIPC initiative is completed. This means, it is assumed that θ=0 in (9) above.

                       

The results obtained using the above parameters in the model are reported in Table 1 below for different growth rates. It has to be noted that even though the average GDP growth rate in Ethiopia has been around 5.5% and its future growth rate is projected by the IMF to be around 6%, the yearly growth rates have been very volatile ranging from -4 to 12 % in the last few years. Hence, alternative growth rates that range from 2 to 10% are used to capture this volatility.

It has to be also noted that a negative primary balance implies that the government has to run a budget surplus to achieve a sustainable debt while a positive primary balance indicates that the government could afford to incur a deficit and yet maintain a sustainable debt.

Looking at Table 1, it is clear that for any GDP growth rate less than 5%, the government has to run a primary budget surplus for, at least, more than 10 years to maintain a sustainable debt. Even at a GDP growth rate of 5% and by the 10th year, the amount of deficit compatible with a sustainable debt is only 0.5% of GDP.  It has to be noted that the budget deficit in Ethiopia averaged about 9% of GDP during the last 10 years. It is not, therefore, difficult to see how daunting a task this will be to maintain a surplus and even to only incur such a small deficit. The possible negative implications of such envisioned budgetary discipline will be noted later.   It should further be noted that even the steady state primary balance required for debt sustainability is only about 5% of GDP, assuming an average GDP growth rate of 10%.    

 

Table1     Debt Sustainability Case A (No availability of new concessional loans)

Year

                                 Alternative Growth Rates

 

2%

3%

4%

5%

6%

7%

10%

1

-1.73

-1.11

-0.50

0.10

0.69

1.27

2.99

2

-1.70

-1.08

-0.46

0.15

0.75

1.35

3.09

3

-1.70

-1.05

-0.42

0.20

0.82

1.42

3.19

4

-1.73

-1.01

-0.37

0.26

0.88

1.49

3.29

5

-1.64

-0.98

-0.33

0.31

0.94

1.56

3.38

6

-1.62

-0.95

-0.29

0.36

1.00

1.63

3.46

7

-1.59

-0.92

-0.25

0.41

1.05

1.69

3.54

8

-1.57

-0.89

-0.21

0.45

1.11

1.75

3.62

9

-1.55

-0.86

-0.17

0.50

1.16

1.81

3.69

10

-1.53

-0.83

-0.13

0.55

1.21

1.87

3.76

Steady state

  0.57

1.16

1.75

2.33

2.90

3.47

5.12

 

Admittedly, the above scenario that assumed no new concessional loans will be available in the post-HIPC era might be unrealistic. Instead a more plausible assumption might be that even after the debt forgiveness under the HIPC initiative is completed, there will be a flow of new subsidized loans albeit relatively less than before. Accordingly, the following exercise considers what the dynamic path of the primary balance will be assuming that the new loan will be, θ =(g/2+π*). The results of using this value for θ in the model (with all other variables taking the previous value) are reported in Table 2.

 

Table 2 Debt Sustainability CASE  B  (partial availability of concessional loans)

Year

                                        Alternative Growth Rates

 

2%

3%

4%

5%

6%

7%

10%

1

0.95

1.92

2.88

3.82

4.74

5.66

8.32

2

0.95

1.91

2.85

3.78

4.69

5.58

8.17

3

0.95

1.90

2.83

3.74

4.64

5.51

8.02

4

0.95

1.89

2.81

3.71

4.58

5.44

7.88

5

0.94

1.88

2.79

3.67

4.53

5.37

7.74

6

0.93

1.87

2.77

3.64

4.49

5.31

7.61

7

0.93

1.86

2.75

3.61

4.44

5.24

7.49

8

0.93

1.85

2.73

3.58

4.39

5.18

7.38

9

0.92

1.84

2.71

3.55

4.35

5.12

7.27

10

0.92

1.83

2.69

3.52

4.31

5.06

7.16

Steady State

0.57

1.16

1.75

2.33

2.90

3.47

5.12

 

 

As is clear from Table 2, Under scenario B,  the government will be able to incur a deficit under all growth scenarios even though it is only if the economy grows at 10% every year that the ratio of the primary balance to GDP will get closer to what prevailed in recent years. But even in the steady sate (last row of Table 2), the equilibrium primary balance is lower than what is historically observed in the presence of grants and concessional loans. 

 

To cover alternative scenarios, the third case considers, case C, in which it is assumed that donors will continue providing financial aid at the prevailing rate. In 2002, Africa Development Bank (2003/2004) Report indicates that the ratio of aid flows to GDP for Ethiopia from all donors was around 20% of GDP. This ratio is the closest  to the assumption that θ =(g+π*) in this model’s formulation.

The results obtained after simulating the model using the above value for θ are reported in Table 3, below. 

Table 3 Debt Sustainability CASE  C (Financial Flows continue at existing rate)   

Year

                                                       Alternative Growth Rates

 

2%

3%

4%

5%

6%

7%

8%

9%

10%

1

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

2

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

3

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

4

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

5

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

6

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

7

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

8

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

9

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

10

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

Steady-State

1.72

3.06

4.38

5.68

6.96

8.21

9.45

10.67

11.88

 

It has to be noted that, considering this scenario is interesting in illustrating what debt sustainability will look like if the existing state of dependency continues. But, it has to also be noted that it is unrealistic to assume that donors will continue assisting countries like Ethiopia indefinitely. Instead a more plausible assumption is that donors are likely to reduce their aid flows, if nothing else for the simple reason that as dependency continues what is named “Aid Fatigue” will set in before long if it has not already. Further, the primary motive of the HIPC initiative is that once the indebted countries received debt forgiveness, they will maintain a sustainable debt in the future on their own, without resorting to concessional loans.

 

The next issue to be addressed is the evolution of the concessional loan over time under the three scenarios considered for alternative real GDP growth rates. As is evident from Table 4, the yearly decline in the ratio of subsidize loans to GDP is very gradual. For instance, for any GDP growth rate below 5%, it takes about ten years to bring the ratio of debt to GDP to about 50%. Even in the more realistic growth (at least in historical terms) rate of 5-to 6%, it takes about seven years to bring the ratio of debt to GDP to about 50%. 

 

 

 

 

 

  Table 4 Evolution of Debt - Case A

Year

               Alternative Growth rates

 

2%

3%

4%

5%

6%

7%

10%

1

80

80

80

80

80

80

80

2

76.48

75.72

74.97

74.22

73.48

72.75

70.60

3

73.11

71.67

70.25

68.86

67.49

66.16

62.30

4

69.90

67.83

65.83

63.88

61.99

60.16

54.98

5

66.82

64.20

61.68

59.27

56.94

54.71

48.52

6

63.88

60.77

57.80

54.98

52.30

49.75

42.82

7

61.07

57.51

54.16

51.01

48.04

45.24

37.79

8

58.38

54.44

50.76

47.32

44.13

41.14

33.35

9

55.81

51.52

47.56

43.90

40.53

37.41

29.43

10

53.36

48.77

44.57

40.73

37.23

34.02

25.97

Steady State

          0

0

0

0

0

0

0

 

 

The second scenario considered (which assumes that new additional subsidized loans will be available) shares with the firs case in that the decline in the share of debt to GDP ratio is gradual.  Due to the accumulation of new loans, even in the best case scenario of a 10% real GDP growth rate the share of debt to GDP will remain above 50% after 10 years. But note that in both cases, the ratio of debt to GDP will be zero at the steady state. It is also important to note that, under the scenarios considered, the time frame in which the debt to GDP ratio will converge to zero takes a long time. For instance, under scenario A, the debt to GDP ratio will range from 7 to 0.5% of GDP for respective growth rates ranging from 2 to 10% after 50 years. On the other hand, for scenario B, since new loans are also added, in 50 years, the ratio of debt to GDP will only decline in the range of 7 to 49% for growth rates ranging from 2 to 10%.

Table 5         EVOLUTION OF DEBT - CASE B

Year

                                        Alternative Growth Rates

 

2%

3%

4%

5%

6%

7%

10%

1

80

80

80

80

80

80

80

2

79.20

78.81

78.42

78.02

77.64

77.25

76.10

3

78.42

77.64

76.86

76.10

75.34

74.59

72.39

4

77.64

76.48

75.34

74.22

73.11

72.03

68.86

5

76.86

75.34

73.85

72.39

70.95

69.55

65.50

6

76.10

74.22

72.39

70.60

68.86

67.16

62.30

7

75.34

73.11

70.95

68.86

66.82

64.85

59.27

8

74.59

72.03

69.55

67.16

64.85

62.62

56.38

9

73.85

70.95

68.17

65.50

62.93

60.46

53.63

10

73.11

69.90

66.82

63.88

61.07

58.38

51.01

Steady state

0

0

0

0

0

0

0

 

 

The evolution of debt under scenario C in which it is assumed that financial flows will continue indefinitely, of course suggests that the rate of indebtedness will continue at the initial rate with no change in the future since by construction the flow is equivalent to what exists at the initial time period.

 

It is worth noting that the above analysis is not to suggest that accumulating debt is always necessarily bad. But what matter are the size of the debt relative to the size of the economy since servicing the debt has significant opportunity costs and possibly adverse social consequences, and the purpose for which it is used. And since we are mainly examining the debt accumulated to finance the to date accomplished economic activities, it is clear that an optimal benefit has not been derived from the accumulated debt if the attendant performance of the economy is the appropriate yardstick to measure it with.

 

5. Summary and conclusions

 

 

This paper attempted to examine debt sustainability in general and in the post-HIPC era, in particular. Starting from conventional estimates of debt sustainability, it further examined what the sustainable primary balance will be under different scenarios of donor behavior in terms of allowing access to subsidized loans once the HIPC initiative is over. The findings could be summarized as follows.

 

First, the simple primary balance flow based tests suggest that the Ethiopian debt is sustainable. The unit root and co-integration tests seem to indicate that given the attendant flow of revenue and expenditures, the current outstanding debt is sustainable.

 

Second, when the primary balance that is compatible with a sustainable debt is computed, it suggests that the fiscal policy effort required to achieve this sustainability is daunting.  This is because the computed required primary balance under different economic growth scenarios that are consistent with a sustainable debt are much lower than what is historically observed.

 

Third, examining the evolution of the debt under the two scenarios considered suggest that the decline in the debt to GDP ratio is gradual in all the alternative growth rates considered. 

 

The important contribution of this exercise is that it helps us gauge the extent to which the existing fiscal structure exhibits some equilibrating characteristics as indicated in the first set of estimations. It also highlights the degree of fiscal effort required to achieve a primary balance that is compatible with a sustainable debt. The important message that comes from this study is that even under the more optimistic scenario that the donor community will continue to offer concessional or subsidized loans even after the end of the HIPC initiative, the public sector primary balance required to maintain a sustainable debt is going to be constraining even under the more optimistic scenario of high GDP growth.

 

The impact of such a constraint on an economy that has a big public sector relative to GDP, high unemployment rate, high incidence of poverty, low tax base is not difficult to imagine. The alternative growth rates considered seem realistic, at least based on the recent growth experience of the economy. And therefore, if the assumptions regarding the flow of new loans in the post-HIPC era materialize, the required tight fiscal policy stance implied by the above results will likely have a significant adverse effect on national efforts to address the multidimensional socio-economic development concerns of the country that range from multifaceted issues such as poverty reduction, in general, and specific social provisions (such as health and education), in particular.           

 

 

 

 

 

 

 

 

                                    References

  

African Development Bank (2003/2004), African Economic Outlook.

 

Allen, T. And D. Weinhold (2000), "Dropping the Debt for the New Millennium:

Is It Such a Good Idea?" Working Paper Series, No. 00-09, London School of Economics.

 

Andrews D., A.R. Boote, S.S. Rizavi, and S. Singh (2004), “Debt Relief for Low-        Income Countries: The Enhanced HIPC Initiative”, IMF Pamphlet Series,

 No. 51.

Arghyrou, M. (2003), “ Debt Sustainability, Structural Breaks and Non-linear

 adjustment: A testing Application to Greek fiscal policy”, Department of

 Economics and Finance, Brunel University.

 

Bangura, Sheku. D. Kitabire and R. Powell (2004), “External Debt Management in

 Low-Income Countries” IMF, Working Paper No. 00/196.

Boote, A. and K. Thugge (1997), Debt Relief for Low-Income Countries: The

HIPIC Initiative, International Monetary Fund, Pamphlet Series No. 51.

 

Boote,  A. R and Kamau T., (2004), “Debt Relief for Low-Income Countries and the               HIPC Initiative”,  IMF, Working Paper Series No. 97/24.

Cohen, Daniel (2000), The HIPIC Initiative: True and False promises. OECD

Development Centre, Technical Paper No. 16.

 

Corsetti, G. and Roubini N. (1991), “Fiscal deficits, public debt and government

 Solvency: Evidence from OECD countries”, Journal of Japanese and

 International Economies, 5, PP. 354-80.

 

Easterly, William (2000), "Growth Implosions and Debt Explositions: Do

Growth Slowdons Cause Public Debt Crises?" World Bank (October, unpublished paper)

______________________, How Did Highly Indebted Countries Become

Highly Indebted? Reviewing Two Decades of Debt Relief" World Bank (June, unpublished paper).

 

Edwards, S. (2002), “Debt Relief and Fiscal Sustainability,” NBER Working Paper

 Series 8939.

 

Government of Ethiopia, "Interim Poverty Reduction Strategy paper 2000/01-

2002/03", September 2000, Addis Ababa, Ethiopia.

 

Gunter, Bernhard G (2001), “Does the HIPC Initiative Achieve its Goal of Debt

 Sustainability?”. WIDER, Discussion Paper No. 2001/100 (September)

 

Hakkio, C. S. and Rush M. (1991), “Cointegration and Government Borrowing

 Constraints: Evidence for the United States”, Journal of Business &

 Economics Statistics”, 9, PP. 429-445.

 

Hamilton, J. and Flavin M. A. (1986), “On the Limitations of Government Borrowing:

 A Framework for Testing,” American Economic Review, 76, pp. 808-819.

 

Hung, A. A. (1991), “Cointegration and Government Borrowing Constraints:

 Evidence for the United States”, Journal of Business & Economics

 Statistics”, 9, PP. 97-101.

 

Jakob, E. C. (2004), “Domestic Debt Markets in Sub-Saharan Africa”  

IMF, Working Paper No. 04/46.

Kremers, J. J. (1988), “Long-Run limits on the US federal debt”, Economic Letters,

 28, pp. 259-262.

 

MOFED (2002), “”Ethiopia: Sustainable Development and Poverty Reduction

Program”, Federal Democratic Republic of Ethiopia (FDRE), Ministry of Finance and Economic Development MOFED).

  

Peterson, S. B. (2001), "Financial Reform in a Devolved African Country:

Lessons from Ethiopia", Public Administration and Development, Vol. 21 (pp. 1-18).

Roubini, Nouriel (2001), “Debt Sustainability: How to Assess Whether a Country is

Insolvent”, Stern School of Business, New York University (December 20).

 

 Stock, J. and Watson. M. (1993), “A Simple Estimator of Cointegrating Vectors

 in Higher Order Integrated System”, Econometrica, 5, pp. 1035-1056.

 

Teklu, Tefera (2000), External Resource Mobilization and External Debt

Situation", Paper presented at a "Symposium for reviewing Ethiopia’s

Socio-Economic Performance 1991-1999". Inter-African Group.

 

Trehan, B. and Walsh C. E. (1988), “Common Trends, The Governemnt’s Budget

 Constraint and Revenue Smoothing”, Journal of Economic Dynamics and

 Control, pp. 425-444.

Wilcox, D. (1989), “The Sustainability of Government Deficits: Implications of the

 Present-Value Borrowing Constraint,” Journal of Money, Credit and Banking,

 21, pp. 291-306.

Yan, S. (2004), “External Debt Sustainability in HIPC Completion Point Countries”

IMF, Working Paper No. 04/160.
 

 

 

 

Appendices

 

Appendix 1 Selected Macroeconomic Aggregates -% GDP (unless otherwise indicated)

                                                       (1962/65-2002/03)

Aggregate

‘62/3-66/7

‘67/8-7/72

‘72/3-76/7

‘77/8-81/2

‘82/3-86/7

‘87/8-92/3

‘92/3-99/0

2000/1

2001/2

2002/3

Real GDP growth rate (%)

4.7

4.0

1.3

2.3

3.7

-0.01

5.7

7.7

1.2

-3.8

Investment

13.5

12.6

9.7

11.0

14.3

13.4

15.9

17.8

20.5

21.2

Saving

11.4

11.0

9.0

4.7

6.5

7.1

5.3

3.1

2.5

1.8

Exports+Imports

24.1

22.1

26.5

29.1

26.0

20.2

37.8

29.5

34.9

34.3

Inflation (%)

 

1.7

11.4

10.7

3.4

11.8

3.8

-5.2

-7.2

15.1

Export as % of Import

83.6

86.6

95.8

53.6

53.7

52.3

56.4

29.9

25.1

25.4

Government Revenue

9.8

11.0

16.3

18.5

18.8

13.6

12.6

18.8

20.1

19.6

Government Expenditure

12.2

13.1

21.2

25.1

29.8

23.5

17.2

28.4

32.2

34.8

Source: Computed based on (MOFED) and CSA data (various years) and IMF (2004) estimates.

 

 

Appendix 2 Selected Financial Aggregates (as % of GDP unless otherwise indicated)

Aggregate

1995/6

1996/7

1997/8

1998/9

1999/0

2000/1

2001/2

2002/3

Broad Money growth %

11.6

3.4

12.7

5.9

14.0

9.5

12.3

10.4

Resource gap

-9.9

-9.1

-9.4

-14.9

-17.0

-14.7

-18.0

-19.4

C/A balance*

-5.4

-3.0

-6.5

-5.6

-11.2

-9.7

-12.9

-12.8

Budget Deficit*

-8.5

-6.0

-7.2

-12.2

-15.1

-9.6

-12.1

-15.3

Domestic debt

32.2

28.6

29.0

31.2

42.2

37.4

39.8

39.1

External Debt

151

79.9

78.4

82.8

86.5

86.3

109.8

98.7

Source: IMF (2001 and 2004).

*Excluding transfers and grants, respectively..

 

     

Appendix 3 Sustainable and Steady-state Primary Balance and Debt to GDP

 ratios Under Alternative Scenarios.                                

 

Dynamic Path

Steady-state sustainable primary balance to GDP ratio

Stationary Susidized  debt to GDP ratio

Stationary domestic debt to GDP ratio

Case A:   θ=0

{ pb t /Y t) = [{ –r*}(DF0/ Y0)e-(g+ π*)(t-1) +{g+ π*-r}(DD0 /Y0) [1/(1+g+ π*)] +

( g + π)(B0 / Y0)

{ pb t /Y t) = (g+ π* -r)  (DD0/Y0)[1/(1+g+ π*)] +

( g + π)(B0 / Y0)

(DF/Y) =0

(DD/Y) =

(DDo/Yo)

Case B:  θ=(g/2+ π*)

{ pb t /Y t) = [{ g/2+π*–r*}(DF0/ Y0)

e-g/2(t-1) +{ g+ π*-r}(DD0 /Y0) [1/(1+g+ π*)]+( g + π)(B0 / Y0)

{ pb t /Y t) = (g+ π* -r)  (DD0/Y0)[1/(1+g+ π*)] +

( g + π)(B0 / Y0)

(DF/Y) =0

(DD/Y) =

(DDo/Yo)

Case C:

θ=(g+ π*)

{ pb t /Y t) = (D0/ Y0)[( g+ π*)-r*

( DF0/D0) -r}(DD0 /D0)]

[1/(1+g+ π*)]+( g + π)(B0 / Y0)

{ pb t /Y t) = (D0/ Y0)[( g+ π*)-r*( DF0/D0) -r}

(DD0 /D0)][1/(1+g+ π*)] + ( g + π)(B0 / Y0)

(DF/Y) =

(DFo/Yo)

(DD/Y) =

(DDo/Yo)

 

 

 

 

Appendix 4 Unit Root Tests for Real Government Expenditures

PP Test Statistic

-4.009161

    1%   Critical Value*

-3.6422

 

 

 

    5%   Critical Value

-2.9527

 

 

 

    10% Critical Value

-2.6148

 

*MacKinnon critical values for rejection of hypothesis of a unit root.

Lag truncation for Bartlett kernel: 3

   ( Newey-West suggests: 3 )

Residual variance with no correction

1191402.

Residual variance with correction

1102528.

Phillips-Perron Test Equation

 

Dependent Variable: D(LRTGE,2)

 

Method: Least Squares

 

Date: 10/20/04   Time: 17:01

 

Sample(adjusted): 1967 1999

 

Included observations: 33 after adjusting endpoints

 

Variable

Coefficient

Std. Error

t-Statistic

Prob. 

 

ΔLRTGE(-1)

-0.712142

0.175159

-4.065686

0.0003

 

C

276.7239

204.3546

1.354136

0.1855

 

R-squared

0.347778

    Mean dependent var

42.16324

 

Adjusted R-squared

0.326738

    S.D. dependent var

1372.503

 

S.E. of regression

1126.174

    Akaike info criterion

16.94973

 

Sum squared resid

39316275

    Schwarz criterion

17.04043

 

Log likelihood

-277.6706

    F-statistic

16.52980

 

Durbin-Watson stat

1.928091

    Prob(F-statistic)

0.000304

 

 

 

 

 

Appendix 5 Unit Root Tests for Real Government Revenue

PP Test Statistic

-5.357541

    1%   Critical Value*

-3.6171

 

 

 

    5%   Critical Value

-2.9422

 

 

 

    10% Critical Value

-2.6092

 

*MacKinnon critical values for rejection of hypothesis of a unit root.

 

Lag truncation for Bartlett kernel: 3

   ( Newey-West suggests: 3 )

 

Residual variance with no correction

0.028965

 

 

 

 

Residual variance with correction

0.026547

 

 

Phillips-Perron Test Equation

Dependent Variable: ΔLRTGR

Method: Least Squares

Date: 10/20/04   Time: 17:46

Sample(adjusted): 1963 1999

 

Includeed observations: 37 after adjusting endpoints

Variable

Coefficient

Std. Error

t-Statistic

Prob. 

Δ LRTGR(-1)

-0.904109

0.168006

-5.381400

0.0000

C

0.034584

0.029405

1.176132

0.2475

R-squared

0.452778

    Mean dependent var

0.001811

Adjusted R-squared

0.437143

    S.D. dependent var

0.233241

S.E. of regression

0.174986

    Akaike info criterion

-0.595678

Sum squared resid

1.071709

    Schwarz criterion

-0.508602

Log likelihood

13.02005

    F-statistic

28.95946

Durbin-Watson stat

1.749698

    Prob(F-statistic)

0.000005

  

   Apendix 6  Static Equation

Dependent Variable: LRTGR

Method: Least Squares

Date: 10/21/04   Time: 11:12

Sample(adjusted): 1967 1998

Included observations: 32 after adjusting endpoints

Variable

Coefficient

Std. Error

t-Statistic

Prob. 

C

5.666999

1.044749

5.424269

0.0000

LRTGR(-1)

0.270089

0.135226

1.997323

0.0564

LRTGE

6.83E-05

1.54E-05

4.447672

0.0001

Δ LRTGE(-1)

6.04E-05

1.97E-05

3.067002

0.0050

Δ LRTGR(-2)

0.218345

0.140820

1.550528

0.1331

Δ LRTGE(1)

3.76E-05

1.49E-05

2.520137

0.0182

R-squared

0.929582

    Mean dependent var

8.429470

Adjusted R-squared

0.916040

    S.D. dependent var

0.317083

S.E. of regression

0.091877

    Akaike info criterion

-1.769364

Sum squared resid

0.219478

    Schwarz criterion

-1.494539

Log likelihood

34.30983

    F-statistic

68.64483

Durbin-Watson stat

1.393231

    Prob(F-statistic)

0.000000

 

 

Appendix 7 Error –Correction Model

Method: Least Squares

Date: 10/29/04   Time: 10:19

Sample(adjusted): 1969 1998

Included observations: 30 after adjusting endpoints

Variable

Coefficient

Std. Error

t-Statistic

Prob. 

 Δ LRTGE

4.89E-05

1.75E-05

2.795519

0.0100

Δ LRTGE(-1)

4.68E-05

2.43E-05

1.925927

0.0660

Δ LRTGE(1)

3.64E-05

1.50E-05

2.430339

0.0229

Δ LRTGE(-2)

-5.03E-05

1.93E-05

-2.606061

0.0155

Δ LRTGR(-1)

0.284700

0.191816

1.484234

0.1508

    Φ(-1)

-0.684511

0.234643

-2.917240

0.0075

R-squared

0.698913

    Mean dependent var

0.033998

Adjusted R-squared

0.636187

    S.D. dependent var

0.158785

S.E. of regression

0.095774

    Akaike info criterion

-1.676788

Sum squared resid

0.220145

    Schwarz criterion

-1.396549

Log likelihood

31.15183

    F-statistic

11.14226

Durbin-Watson stat

2.052731

    Prob(F-statistic)

0.000012

 

 

 

 

 Appendix 8 Case A (With reduced domestic and foreign interest rates)

Year

                             Alternative Growth Rates

 

2%

3%

4%

5%

6%

7%

10%

1

0.3775

0.9730

1.5610

1.5610

1.5610

1.5610

1.5610

2

0.3927

0.9956

1.5908

1.5981

1.6054

1.6127

1.6343

3

0.3927

1.0178

1.6199

1.6343

1.6485

1.6626

1.7040

4

0.3775

1.0397

1.6485

1.6696

1.6903

1.7108

1.7703

5

0.4375

1.0613

1.6765

1.7040

1.7309

1.7573

1.8334

6

0.4522

1.0826

1.7040

1.7376

1.7703

1.8022

1.8934

7

0.4667

1.1035

1.7309

1.7703

1.8085

1.8456

1.9504

8

0.4810

1.1242

1.7573

1.8022

1.8456

1.8875

2.0047

9

0.4952

1.1445

1.7832

1.8334

1.8816

1.9279

2.0563

10

0.5093

1.1645

1.8085

1.8637

1.9165

1.9670

2.1054

Steady state

1.9086

2.4896

3.0634

3.6302

4.1903

4.7438

6.3667

 

 

Appendix 9 Case B (With reduced domestic and foreign interest rates)

 

                           Alternative Growth Rates

Year

2%

3%

4%

5%

6%

7%

10%

1

3.0569

4.0062

4.9413

5.8628

6.7710

7.6662

10.2778

2

3.0455

3.9836

4.9041

5.8077

6.6947

7.5657

10.0870

3

3.0455

3.9613

4.8677

5.7539

6.6207

7.4686

9.9056

4

3.0569

3.9394

4.8320

5.7015

6.5489

7.3749

9.7330

5

3.0119

3.9178

4.7969

5.6503

6.4791

7.2844

9.5688

6

3.0009

3.8966

4.7626

5.6005

6.4115

7.1970

9.4126

7

2.9901

3.8756

4.7290

5.5518

6.3459

7.1127

9.2641

8

2.9793

3.8550

4.6960

5.5044

6.2822

7.0312

9.1228

9

2.9687

3.8347

4.6637

5.4581

6.2203

6.9525

8.9884

10

2.9581

3.8146

4.6320

5.4130

6.1603

6.8766

8.8605

Steady state

1.9086

2.4896

3.0634

3.6302

4.1903

4.7438

6.3667

 

Appendix 10  Case C (With reduced domestic and foreign interest rates)

Year

                           Alternative Growth Rates

 

2%

3%

4%

5%

6%

7%

8%

9%

10%

1

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

2

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

3

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

4

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

5

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

6

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

7

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

8

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

9

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

10

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

Steady-State

3.82

5.14

6.44

7.72

8.98

10.22

11.44

12.65

13.83

 


[1] For detailed discussion of the HIPc initiative and conditions attached to it see Boote and Thugge (1997) and Cohen(2000).

[2] Some Studies (Arghyrou, 2003, for instance) attempted to address such issues by introducing non-linearity in the budget structure.

[3] Alternative formulations of equations (9) and (10) for different scenarios are presented in Appendix 3.

[4] Results for lower domestic and foreign interest rates are reported in Appendices 8 to 10.

[5] Consideration of alternative scenarios and other extensions that include different parameter values and policy changes such as exchange rate devaluation are in progress in a separate paper.