Determinants of Budget Deficit in Ethiopia

You might also like

Download as doc, pdf, or txt
Download as doc, pdf, or txt
You are on page 1of 29

DETERMINANTS OF BUDGET DEFICIT IN

ETHIOPIA

A SENIOR ESSAY

By: EFA GOBENA

Advisor: Mr. G.V. Chandra Sekhar

In Partial Fulfillment of the Bachelor of Arts Degree in Economics

Submitted to:

DEPARTMENT OF ECONOMICS,
COLLEGE OF BUSINESS AND ECONOMICS,
ARBA MINCH UNIVERSITY

June, 2011
Arba Minch, Ethiopia

CHAPTER ONE
INTRODUCTION
1.1 Background of the study
Ethiopia’s growth in the last decade, and in particular in the second half of the decade,
has been appreciated by international observers, including the continental economic
Institutions such as the Economic Commission for Africa (ECA, 2007) and also ‘The
Economist’ magazine in its July 2008 issue. The government has began to openly
argue that Its success is fundamentally related to its rejection of the ‘neo-liberal’
economic policy which is usually referred as the ‘Washington Consensus’ or
Structural Adjustment Policies (SAPs). The government also attributes this growth
success to its embrace of the idea of a ‘developmental state’, the latter usually
referring to a policy of public sector intervention the economy both through policy
and active investment.
This issue surrounding budget deficit are not certainly new, but the economic
development of the past decades has led to renewed interest in the fiscal
themes .Government budget deficit is a perennial topic of debate among policy
makers. Traditional view of government deficit taken by by most economist said that
when government runs a budget deficit and issues deficit. It reduces national saving
which in turn lead to lower investment and a larger foreign debt. This view concludes
that government debt places burden on future generations. While Ricardians view of
government deficit stress that budget deficits merely represent a substitution of future
taxes for current taxes. Budget deficit of government may affect a nation’s role in the
world economy (Mankiw 7th Edition 2007).
In recent years government spending LDC’S has increased significantly. This is
mainly attributed to the fact that the governments of LDC’S are involved in social,
political and economic affairs. Meanwhile revenues do not grow rapidly in the same
proportion as expenditure due to narrow taxes basis and inefficient tax collection of
systems resulting in a budget deficit. In industrial countries, too, the failure of market
mechanism in 1930’s (during great depression) calls for government intervention in
the economy paving the way for increased government expenditure in those countries.
Despite the consensus on the need to reduce budget deficit, LDC’S offer a wide range
of expenditure with budget deficit. From country to country budget led to high and
variable inflation with crowding out of investment and growth .while in some
countries moderately high budget deficit seem not to generate macroeconomic
imbalance at all. The explanation for this different outcome is that different
governments adopt different techniques of financing their deficits. Thus the
consequences of fiscal deficit depend on the way they are financed (William
easterly .files.wordpress.com/2010/08 and Mankiw, 2007)
Ethiopia faces big budget deficit. The combined effect of a very rapid increase in
expenditure and a relatively slower increase in revenue is obviously increase
government’s budget deficit. The budget deficit including grants increased from a
little less than one billion birr in 1996/97 to over 6 billion in 1999/00. The fast
increase in the budget deficit has started in 1997/98 when it more than doubled from
its previous year level to 2.1 billion birr again doubling to 4.5 billion birr in 1999/00.
Budget deficit excluding grants increased from about 2.5 billion birr in 1996/97 to
over 7.7 billion birr in 1999/00. As a proportion to GDP the budget deficit including
grants increased from 2.4% in 1996/97 to 4.5%%, 9.3% and 11.5% respectively in the
following years. the deficit excluding grants rose to as high as 15% of GDP in
1999/00 from its level of 6% in 1996/97 increasing to 7.5% and 13% in the following
two years. this is a significant and certainly unsustainable increase in the country’s
budget deficit and compares unfavorably with the average level of deficit of both the
early years of a new government (4% including grants) and the last years of the
Dergue (8% including grants. in general budget deficit is increase from time to time
(Befekadu Degefe, Berhanu Nega, Getahun Tafesse, 200/01).
As part of the structural reforms agreed with multilateral financial institutions
(MLFIS) the government of Ethiopia was committed to significantly decrease or over
abandon domestic borrowing to finance its deficit. This was considered as prudent
strategy both to curb inflation and even more importantly to avoid crowding out
private investment. However, given the attitude of donors towards the war with
Eritrea and even more their attempt to force the country to accept a peace deal it was
not ready to accept with a treat of cutting off aid the government had option but to
resort to domestic sources to finance the war. This is clearly reflected in the war, the
government financed its deficit. On the side of monetary policy, the government
accepted the advice of the MLFIS and practiced tight monetary policy until 1996/97
where money supplies were increasing at less than the rate of growth of nominal
GDP. Policy changed, however, beginning with the conflict and money supply started
to increase rather considerably. In 1997/98 narrow money increased by 10.7% and
broad money by 12.7% compared with the rate of growth of nominal GDP of 8.1 %
( Befekadu Degefe, Berhanu Nega, Getahun Tafesse, 200/01).

1.2 statements of the problem


Attaining a sound macroeconomic balance has become a priority of industrial and
developing countries’ economies in the measurement of government success. Ethiopia
has a long economic history characterized by macroeconomic imbalance, one of
which is the gap between government spending and revenue.
It is perceived that a larger deficit always signifies a more expansionary fiscal policy.
But that is not always true. Because during a recession as GDP falls, the governments
most important sources of tax revenue; income taxes, corporate taxes and payroll
taxes all shrink because firms and people pay lower taxes when they earn less(Liam
J.BAUMOL and Alan S.Blinder 1998)
Hence, the study attempts to observe the determinants of budget deficit in Ethiopia.
Among the determinants of budget deficit this study relies on seven important factors
which include real GDP annual growth rate, inflation, real exchange rate, money
supply real interest rate, share of agriculture in the GDP and real per capita GDP.
Study also analyzes the significance of each determinant of budget deficit and their
implication on the Ethiopia economy. Finally, the study attempts to find out the
significant determinants of budget deficit in Ethiopia.
1.3 objectives of the study
1.3.1 General objectives
The main objective of this study is to investigate the determining factors of budget
deficit in Ethiopia.
1.3.2 Specific objectives
To assess the effect of the real GDP annual growth rate on budget deficit.
To assess the effect of the real exchange rate on budget deficit.
To assess the effect of the share of agriculture in the GDP on budget deficit.
To assess the effect of the inflation on budget deficit.
To assess the effect of the domestic interest rate on budget deficit.
To assess the effect of the broad money on budget deficit.

1.4 hypotheses
The following relationships are hypothesized based on the earlier studies (conducted
by Combes and Saadi-Sedik (2006) and tested with reference to the time series data.
 There is positive relationship between real GDP annual growth rate
and budget deficit.
 There is positive relationship between real per capita GDP and budget
deficit.
 There is negative relationship between broad money supply and budget
deficit.
 There is negative relationship between the inflation and budget deficit.
 There is positive relationship between the real interest rate and budget
deficit.
 There is negative relationship between the share of agriculture from
GDP and budget deficit.
 There is positive relationship between the real exchange rate and
budget deficit.

1.5 sources of data and method of analysis

1.5.1 Source of data


The study entirely depends on secondary data collected from different sources such as
National Bank of Ethiopia (NBE), Minister of Finance and Economic Development
(MOFED).
1.5.2 Methodology of analysis
To meet the objectives of the study the descriptive and econometric methods of
analysis was used. OLS is undertaken to clearly see the determinants relationship with
budget deficit in Ethiopia. The descriptive methods such as graphs, percentages are
used to show the trend of budget deficit, revenue and expenditure. In general study
used both descriptive and econometric methods of data analysis.
1.6 scope of the study
In this study, the determinants of budget deficit were studied. The study covers the
time period from 1974/75 up to 2009/10. It is chosen because it can explain the
budgetary deficit trend experienced in Ethiopia. It tries to give an emphasis on the
trend and condition of budget deficit during the two regimes, the Dergue regime and
present (FDRE) regime.
1.7 significance of study
The study has been conducted on the determinants of budget deficit at national level.
The study is significant in the context of the following:
 Provides information about the determinants of budget deficit.

 Shows the proper measures taken to reduce budget deficit.


 Finally, this study may also give a clue for further study about determinants of
budget deficit in the future.

1.8 limitation of the study


Due to errors in the collected data the problem of non-stationary raises then the
researcher used growth rate to overcome this error of data manipulation.
1.9 organization of the study
The study is organized in to four chapters. The first chapter of study deals with
background of the study, statement of the problem, objective of study, hypothesis,
significance of the study, scope of the study, method of data collection and data
analysis, limitation of the study and organization of the study. The second chapter
discusses the theoretical literature on budget deficit, means of financing budget
deficit. In addition review of the empirical literature is dealt. The third chapter deals
with descriptive analysis of budget deficit and its components such as government
revenue, government expenditure during the Dergue and the present government.
Chapter four deals with the discussion and review of econometrics issue. Under the
econometrics issue: specification of the model, tests and interpretation of the results.
Finally concluding remarks and policy implication of the study are discussed in
chapter five.

CHAPTER TWO
LITERATURE REVIEW
2.1 Theoretical literature reviews
2.1.1 Government Budget deficit
Analysis of budget deficits remained peripheral in most literature on public finance.
Most of the literature prior to the 1950s focused on separate discussion of the revenue
and expenditure side of government budget where the main occupation of the analysis
is to explain the role of fiscal policies in resource allocation, stabilization, income
distribution and economic growth, and devoted little explanation for budget deficit.
Most of the literature of that period was based on developed countries and mere
discussion of budget deficit was ascribed to developed financial markets which made
deficit financing easy with little impact on macroeconomic variables. Moreover,
Keynesian conception of budget deficit is that of an average balanced budget deficit
over a business cycle, i.e. surplus during boom and deficit during recession as a norm
of fiscal behavior, hence fiscal deficit accordingly was not much of the problem over
a period. Even if analysis of budget deficit got consideration since the 1960s, a
significant number of studies that model the determinants of budget deficit explicitly
were still limited (MANKIW, 2007).

Budget deficit is related to activities of the government. The early mercantilist school
of thought support active government participation. This view was opposed by
classical economist who believed that market mechanism itself can regulate the
economy and therefore government should limit its intervention in the economic
sphere. However in the 1930s (due to great depression) economists started to question
the working of the market mechanism. They recognized that there are some instances’
where the free market operation suggested by the classical does not work correctly.
Models of development in the late 1940s and 1950s focused on market failures in
LDC’S and advocated large government involvement.
Thus all programs taken in those countries during that period like, the big push
balanced growth, redistributions with growth and all the rest suggest more, not less
government involvement. However, large government interventions on the other hand
mean increase in government expenditure.
2.1.2 Concept of budget deficit
Any attempt to assess budgetary impact on macroeconomic variables such as money
supply, balance of payment, public debt and aggregate economic activity requires a
specific measure of the budget deficit. Several concept of budget deficit, however,
seem to be in circulation. Notwithstanding this, a deficit has tended to be viewed as
summary of government receipt and payment in a single budget year. In this context
the conventional deficit is defined as follows;
Fiscal deficit as convectional defined on a cash basis, measure the difference
between total government cash outlay, including interest outlays, but excluding
amortization payment on the outstanding stock of public debt and total cash receipt,
including tax and non-tax revenue and grant but excluding borrowing proceeds. In
this manner budget deficit reflect the gap to be covered by net government borrowing
including borrowing from central bank.
The most widely used purpose-oriented deficit measures are the current account
deficit, the structurally adjusted deficit, the primary deficit and the operational deficit.
2.1.2.1 Current account deficit
The current account deficit of the government is the excess of non-capital
expenditures over non-capital revenue. It depicts government dissaving and could be
used as a measure of the extent to which government exercises prudence in its
financial management.
However, useful the current deficit may be the problems surrounding its calculation
are intractable. A case in point is the treatment of depreciation in enterprise account
on accrual basis, which contradict public sector accounts that tend to be first available
on the basis.
Moreover, the accounting treatment of investment in human capital as current outlay
despite its importance in explaining growth is at odds with the economic treatment.
Furthermore, the recurrent component of any capital project is likely to manipulated
in many ways to give different picture of government saving. Finally in the context of
economic adjustment, the difficulty of separating adjustment induced effects on the
budget from those introduced by external shocks limits the use of the current account
deficit as a measure of fiscal stance.

2.1.2.2 Cyclically Adjusted Deficit/Trend Deficit


That the budget and aggregate demand influence each other hardly needs evidence.
The cyclically adjusted deficit cleans the budget from the effect of business cycles. It
is used by practitioners as a tool to assess the adequacy of the stance of fiscal policy.
In practice an assessment of the stance of fiscal policy involves three ableit simple
steps.
First, definition of the cyclically adjusted budget (CAB).This involves developing a
budget profile that allow for the impact of the economy on the budget. It should be
noted that the choice of base year when the budgetary position was consistent with a
satisfactory level of economic performance is central to the CAB. Thus; the CAB can
be formalized as follows.
CAB=g*PGDP - t*AGDP…………………………1
Where CAB cyclically adjusted budget
PGDP = potential GDP
AGDP = actual GDP
g = base year ratio of government expenditure to GDP
t = base year ratio of revenue to actual GDP
Note that relating expenditure and taxes to potential GDP and actual GDP
respectively allows the budget to vary with the cyclical position of the
economy.i.e.the deficit will rise during recession and fall during a boom.
Second, calculating the cyclical effect of the budget. This is accomplished by
comparing the actual deficit with the cyclically neutral budget. Thus,
CEB= (G-T)-CAB…………………………….2
Where CEB –cyclical effect of the budget
G- Government expenditure
T-government revenue
Whether fiscal policy was expansionary or contractionary is indicated by the sign of
the CEB.A positive CEB points to the former while a negative one indicates the latter.
Third, assessing stance of fiscal policy. The appropriateness of fiscal policy is
evaluated buy comparing the CEB with the cyclical situation. For example, an
expansionary CEB would be considered counter cyclical if the economy was in a
slump and procyclical if the economy was in a boom.
Sometimes the fiscal impulse, i.e. the change in the impact rather than the impact
itself is used to assess the adequacy of fiscal stance. The fiscal ”impulse” is measured
by the change in CEB.A positive movement of the CEB indicates fiscal injection
while a negative change shows a withdrawal of fiscal stimulus. Therefore using the
fiscal impulse was to get around the problem of choice of a base year.
2.1.2.3 Primary and operational deficits
Recall that the CAB includes an important non-discretionary variable; interest
payment which usually depends on previous deficit. The primary deficit (non-interest
deficit) is a measure of the fiscal deficit adjusted for interest payments. It is computed
by subtracting net payments by the government from total government expenditure.
The implication is that the primary balance should eventually turn surplus to provide
at least in part where with for interest on current debt.
This, however, is not required in a situation where government revenue and GDP
grow faster than real interest rate. What are required are similar growth rates for
interest payments, debt and GDP so that they remain constant in relative terms. In any
case, it is difficult for debt and GDP to grow faster than the real interest rate.
One basic deficit with the primary deficit is that it does not exclude the part of interest
payment including inflation. This is particularly important in high inflation situations.
To alleviate this problem the effect of inflation on interest payment should be getting
an inflation –corrected deficit known as operational deficit.
The economic underpinning of the operational deficit is that inflation-induced interest
payments are similar to amortization in that they are a return of capital rather than a
return to it and don’t affect aggregate demand in real terms. Thus the question is
whether inflation generated interest payments are used to buy new bonds or to finance
consumption.
Although, the practical significance of this deficit measure is clear, it has
macroeconomic deficiency by correcting the deficit for the impact of inflation on it.
The ability to assess the impact of the deficit on inflation is lost (Blejer and Cheasty,
1991).Nevertheless; it is useful as a guide for practitioners, because it brings out the
size of over statement of the fiscal imbalances given by the conventional deficit
during high inflation and high nominal interest rates.
The different deficit measures outlined above are used for different purposes. One can
therefore use any of them depending on the purpose to serve. In particular where an
economic adjustment program is underway, it would be useful to measure the fiscal
deficit from different angles.
2.1.3 Causes of budget deficit
In recent years government spending in LDC’s has increased significantly. This is
mainly attributed to the fact that governments of LDC’s are involved in social,
political and economic affairs. Mean while, revenues do not grow rapidly in the same
proportion as expenditures due to narrow tax bases and inefficient tax collecting
systems resulting in a budget deficit. In DCs too, the failure of market mechanism in
the 1930s (during great depression) calls for government intervention in the economy
paving therefore increasing government expenditure in those countries.
Changes in the general price level, income and population size determine growth of
government spending while tax administration and income determine the growth of
government revenue and the imbalances growth of the two(spending over revenues)
causes budget deficit(http://swbplus.bsz-bw.de/bsz118062182kap.pdf). While
Shibeshi (1992) has identified fiscal indiscipline aligned with structural variables as a
major factor underlying the expansion of fiscal deficit.
Taylors (1988) on the other hand tried to explain the reasons why government is open
to high fiscal deficits by forwarding three arguments. First, a government deliberately
favors high spending and lower taxes for political reasons to make there governance
legitimate and this view are known as the view of political deficit. Second, it is argued
that structural factors like fall in term of trade (TOT), export supply and their price
fluctuations etc. aggravate the growth of public sector deficit; this view is called the
structural view. The third view, on the other hand, argues that inflation reduces real
tax revenue and thus causes high fiscal deficit. This is called the inflationary approach
to fiscal deficit. Sometimes the situation may be described as “olvera –Tanzi effect”
of fiscal deficit, i.e. deficit caused by a decline in real tax revenues during period of
high inflation as a result levels reduce real tax revenues of the government
significantly as the government collects and accounts its tax receipts in latter and
expenditures causing high budget deficit in the country (Carlos Rodriguez: 1994).
2.1.4 The effect of budget deficit on macroeconomic variables
The macroeconomic effects of government spending policy are important because the
effects of a particular spending program can be undone by these effects. We now turn
to the discussion of macro effect of deficit financing on some variables, namely
current account, private saving and investment. The open economy national income
identity provides a useful guide for the discussion.
Y=C+I+G+(X-M)……………………………3
Where Y=national income C=private consumption spending
X=export earning I=private investment spending
M=import bile G=total government expenditure
Subtracting tax (T) from both sides of the identity we obtained the disposable income
(Yd)
Y-T=C+I+G-T+(X-M)………………………………..4
Yd =C+I+ (G-T) + (X-M)…………………………..5
Since the disposable income (Yd) can either be saved or consumed (C+S) =Yd
C+S=C+I+ (G-T) +(X-M)…………………………6
Rearranging equation (3)
(T-G)= (I-S) +(X-M)…………………………….7
Identity (4) shows that government deficit has a counter balance in investment crowd
out or current account deficit or some combination of both. This implies that there is a
crowd out to budget deficit. For a given level of saving, the deficit implies either a
reduction in investment or an increase in current account deficit or both. The precise
effects, however, are not easy to define. The equation is merely an identity and says
nothing about causation. A direct linking would be wrong because it would imply
knowledge of ex-ante movements that could only be provided by additional
information about the way the saving and investment decision behavior of private
sector agents is determined. In this framework the effect of the deficit on current
account, investment and investment decision of individuals. There are two views in
this regard.
One of the views is the Riparian-equivalence view. In this view, it has been argued
that government deficit have no macroeconomic impact. The budget deficit in this
view affects either investment or external current account. In principle an increase in
government expenditure or a reduction of taxes results in an identical increase in
private saving and consequently has no lasting impact in the economy. Rational and
forward looking individuals are bound to anticipate future increased taxes, therefore,
raise their current rate of saving. However, Riparian-Equivalence presupposes inter-
alias the absence of constraints on borrowing and the neutrality of the tax system,
situation which is unlikely to be found (Mankiw, 2007; Dornbush, 1992).
The other view is the traditional view which believes that government deficits have a
larger impact on the economy, especially on private investment and external current
account. For a given overall spending, an increase in government spending entails a
drop in private investment. Likewise, at given private saving rate and in which
domestic borrowing opportunities are scarce, there is a closer relationship between the
budget deficit and the external current account (H.L.Blatia,1995).
However, this relation is not always direct as it depends in part on the level and
composition of government expenditure, the stance of monetary policy and related
changes in interest rates and real exchange rates (William
easterly.files.wordpress.com/2010/08).
2.1.5 Methods of deficit financing
In general, the problem of deficit financing is one of the controversial issues in
economics which does not have conclusive theoretical analysis. The method of
financing is also an important factor determining its effects. There are four sources of
deficit financing namely; drawing from reserves, domestic borrowing, external
borrowing and money printing (Clayton, 1995). These are discussed below.

2.1.5.1 Drawing from reserves


One of the methods of financing government budget deficit is to run down the foreign
exchange reserves. This method of financing can be used to pay for import bills. It has
two merit of reducing inflationary pressures in the domestic economy and reduces the
risk of external debt crises. However, it tends to appreciate the domestic currency by
increasing the supply of foreign currency. In turn, this will deteriorate the current
account by making import attractive and discouraging export. Moreover, there are
limits imposed by international organizations like the IMF to its use in financing
budget deficits.
2.1.5.2 Domestic borrowing
Government can obtain domestic borrowing from the banking system or the private
sector. Funds can be obtained by selling securities (government bonds and treasury
bills). It is non-inflationary and reduces the risk of external debt crisis. Nevertheless,
domestic borrowing from the banking system (excluding the central bank) requires a
relative well developed financial intermediation system. Moreover, the government’s
recourse to domestic financing reduces the supply of loan able funds. In countries
where the interest rates are relative flexible, the upward pressure on real interest rate
tends to decline private investment penalizing growth. Furthermore, if the exchange
rate and the interest rates are subject to government control, resorting to domestic
financing has a more crowd out effect on private sector and translates in to credit
rationing.
2.1.5.3 External borrowing
External borrowing often appears to be attractive because of the crowd out effect on
private investment and its being non-inflationary. It allows government to carry out
expansionary domestic policies without drawing upon domestic saving and there by
put
Ting pressure on domestic interest rate to rise. Moreover, it can induce greater fiscal
and monetary discipline, since it eliminates any incentive that the government might
have to generate inflation to reduce the real debt burden.
Although, government external borrowing does not directly affect domestic interest
rates and the supply of loan able fund, it may also crowd out private investment
through its impact on prices or the nominal exchange rate in flexible or managed
exchange rate regimes. When the budget deficit stems from expenditure on locally
produced goods, external borrowing brings about an appreciation of the real exchange
rate under a fixed or managed exchange rate regime. This has crowd out effect on
certain local producers (Mankiw, 2007).
2.1.5.4 Money printing
The budget deficit can be covered directly through money creation by the central bank
or more generally by increasing credit of the banking system. However, excessive use
of monetary financing results in excess over all demand, which in turn translate in to
inflation or under fixed exchange rate on the balance of payment.
However, the relationship between the monetary financing of budgets and inflation is
neither direct nor linear especially in the short run (Clayton, 1995). The unstable
nature of this link is generally attributes to several factors:
 Private saving may change as a result of change in inflation expectation
 The composition of budget financing may change over time
 The composition of demand for money is sometimes unstable and
 Expectation may be shaping future government policy

(Mania, 2007).

2.1.6 Determinants of budget deficit


There can be socio-economic development differences, political and structural factors
beyond the belief of policy practitioners about the role of government in economic
activities that can explain the reason why government run deficits. Budget deficits
may occur as a result of increasing government activities in the economy,
development theorizing problems related to the structure of the economy and political
pressures to spend more than what government collect as revenue.
The size of government in the economy is also an important determinant which may
be included in the structural determinants of budget deficits. Early work on the issue
at hand tried to explain why government’s involvement in the economy increases over
time. The earliest explanation of this kind is thee wager’s law. The law states that
there is an increasing trend of government expenditure over time. Wagner stated that
based on historical factors in Germany: there is a functional relationship between the
growth of public expenditure and the growth of the economy. Wagner did not clearly
refer the growth of the ratio of government expenditure to GDP or the absolute size.
Budget deficit may also accrue due to inflation. But, the direct link of the case can be
problematic. Reversed causation can be established for instance monetization of
deficit may prove to be inflationary. On the other hand, in economies where there is
inflation there is pressure for running budget deficits. This is because, taxes collection
exhibits collection lags that when there is inflation the real balance of tax revenue
would decrease at the collection. This situation is known as Oliver Tanzi effect.
An alternative approach to the explanation of budget deficit is the examination of
structural factors. These factors refer to those factors which the can not change them
in the short-term. As summarized by Tadesse (1994) structural factors to:
1. Level of economic development
2. Growth in government revenue
3. Instability of government revenue
4. Extent of government participation in the economy
5. Size of government

2.1.6.1 Level of economic development

For countries with low level of economic development (measured by PCI) the initial
levels of investment in infrastructure and expenditures on social services are very low.
Hence, expenditures as a percentage of the GDP, requires expanding the infrastructure
build new school, clinic….etc becomes significant. On the other hand, revenues on
those countries grow at a very slow rate. The structure and collection of taxes is such
that it adds to a growing deficit. The system of taxes collection in most low income
countries incur considerable time lags. Moreover, the money income elasticity of their
tax system is low. These makes the impact of inflation on the budget deficit,
especially when the deficit is very high, very server. Also, private saving ratios are
positively correlated to the level of economic development. Hence, it is assumed that
government of low income countries will be focused to supplement private saving
more to respond to public expectations. Here the saving is considered as the
mobilization and diversion of the money incomes through deficit financing that would
otherwise be spent for consumption.
2.1.6.2 Growth in government revenue
The need to deficit financing decreases when revenue grows rapidly while
expenditure grows at a lesser pace or decrease. However, in adequate growth of
revenue leads to deficit financing. Moreover, governments with slowly growing
revenues are more likely to resort to deficit financing to support expenditure than
government whose revenues are growing at significant rates.
2.1.6.3 Instability of government revenue
Fiscal discipline in a stiff challenge for government with large revenue fluctuation
government find more difficult to manage fiscal balances when revenue fluctuate
largely than when its growth is slow but stable. This is especially true when the
government is revenue follower.
2.1.6.4 Extent of government participation in the economy
The government’s ability to control expenditure is influenced by institutional,
ideological and structure factors. Among the factors which put upward pressure on
spending are lacks of coordination between financial and physical plans, development
theorizing high share of recurrent expenditure and revenue instabilities.
2.1.6.5 Size of government
The relative size of government sector tends to be highly correlated with increasing
government role in production, consumption and distribution of goods and services.
This includes upward pressure on spending and is a difficult process to reverse it in
the short-term.

2.2. Empirical literatures


It is known that fiscal deficit adversely affect the economic structure and this impact
depends on the ways of financing. Generally to analyze budget deficit and related
issues ten case studies were taken out of fifty nine countries (LDCS and DCS) and
econometric estimation was applied to each cases. The ten countries are Argentina,
Chile, Colombia, cotdivore, Ghana, Mexico, morocco, Pakistan, Thailand and
Zimbabwe. The reason for choosing them is because of their representatives of
developing worlds and their diverse fiscal and macro experience. This section thus
summarizes the major empirical finding of the studies. (Carlos A. Rodiguer: 1994).
The study concludes that the simple correlation between fiscal deficit and individual
indicators of macroeconomic imbalances (interest, inflation etc) are weak to non-
existent. This clearly due to different mechanism that each country rely to finance its
deficit. However, the study found strong negative correlations between deficit and
growth.
The study also examined the degree to which deficit are affected by external shocks
(foreign exchange, commodity price fluctuation and foreign interest rates or by
domestic macroeconomic variables (inflation, domestic interest rates, real exchange
rates) and found that only in Colombia, more than 50% of the variation in deficit was
explained by foreign shocks, but in Chile, Ghana and Zimbabwe foreign shock do not
seem to explains the deficit variation. the Oliver-Tanzi effect by inflation lower tax
revenue was seen only in Ghana, Colombia, Thailand and Zimbabwe. Revenue from
money creation associated with inflation was found in Thailand and Argentina. An
additional percentage in GDP from money creation induces 5% additional: inflation in
Thailand and 97% inflation in Argentina (ibid).
According to this study, high fiscal deficit are significantly associated with high
negative real interest rate in Ghana, Mexico and Zimbabwe. These countries relied on
financial repression as means of financing their deficit and all control their nominal
interest rate which latter resulted in negative real rates. The idea that high fiscal
deficit crowd out private investment was also reaffirmed by the case study. For
example in Colombia a debt financed expansion of 0.9% of GDP led private
investment to fall by 0.5 to 0.9% (ibid).
Regarding the effect of public investment on private investment, the study found
negative effect of public investment on private investment in Chile, Colombia, Ghana,
Mexico, and zero in Argentina. The study again tells us the existence of ambiguity of
the relationship between the two, and the relation depends on whether the former
substitutes or compliment the latter. The “Recardian equivalence theory”, which
propose that deficit and taxes are equal in their effect on private consumption, and the
private sector save more in anticipation of high future tax obligation during high fiscal
deficit, was found to be true in the case study in Argentina, Morocco, and Zimbabwe.
The study has confirmed that fiscal deficit appreciate real exchange rates, and the
existence of close association between fiscal and internal balances (ibid).
The study also pointed out the key role that economic growth plays in determining the
macro economic consequences of high public deficit. Fast growth economies in the
1980s like India, Pakistan, and Thailand were able to remove demand and other
means of financing while those with low (negative) growth such as Argentina, Brazil,
and Mexico faced large fiscal deficit were forced to under take adjustments.
(Nadeem UI. Haqu; 1991) has found Pakistan as the only exceptional country which
was able to use monetary financing of the deficit with our inducing high inflation.
However, high fiscal deficit did contribute to weak external sector of the country. But
[Westerly: 1989] has precisely quantified the theoretical relationship between fiscal
deficit and inflation in Colombia. He found that 22% of annual inflation, the revenue
from money creation is about 2% of GDP. And increasing the revenue figure to
2.65% level further required inflation to go over 100% a year. More over, he has
found that the high fiscal deficit in the 1970s in the country was caused by boom in
coffee price. During this time the government relied on foreign exchange reverses and
faced appreciated real exchange rate which damaged the non-traditional export sector.
During the 1986 high fiscal deficit, in country learning from its past mistake resorted
to issuing of public bonds to finance the deficit.
(C. chamely and H. chahe; 1991) has pointed out that Cot Divorces’ macro economic
shocks were largely caused by booms in cocoa and coffee prices in the mid 1870s
which caused high fiscal deficit. During these periods the government used external
debt to finance it, and faced appreciated real exchange rate. While in Argentina
monetary financing of the deficit immediately led to inflation. The monetary
financing in the country has caused a triple inflation levels, up to 4,927%. (Carlos
Alfredo Rodriguez; 1991).
The Chilean data confirmed all the relationship hypothesized earlier. Monetary
financing creates inflation, dept financing raises interest rates, and external financing
induces real appreciation and trade deficit. (Gorge marshal and K. sehimidt Hebeel).
Prior to 1983 Morocco highly relied on foreign borrowing to finance its deficit. It
covers 60% of the deficit. But in 1983 the government stopped this method because of
its bad consequence and relied on accumulation of arrears as a source of involuntary
financing from the private sector
(http://apcentral.collegeboard.com/apc/public/repository/ap06).

In the past the Thailand government usually ran a budget deficit. But in recent years
the deficit has become a surplus because tax capacity relative to GDP has increased
significantly and expenditure over revenues were also reduced. Evidence from the
country show also that the government adopted non-inflationary means of deficit
financing more relaying on domestic borrowing through bonds. Evidence also shows
that there is no clear relationship between inflation and seignorage in the country.
(Virabongase Ramangkura; 1991).
CHAPTER THREE
DESCRIPTIVE ANALYSIS
3.1 GOVERNMENT BUDGET DEFICIT, GOVERNMENT REVENUE,
GOVERNMENT EXPENDITURE IN ETHIOPIA
3.2 Government Budget deficits
This section would reveal the descriptive analysis followed by various variables
incorporated in the model with that of the budget deficit. The aim of the descriptive
analysis is to serve as a base for the econometric analysis.
On average, during the 1974/75-2009/10 period, 61.7 percent of the total government
spending has been on current expenditure the rest being on capital expenditure. This
large size of current expenditure may have impeded growth by reducing the resources
available for capital expenditure. Defense expenditure, poverty targeted expenditure
(which includes education, health and agriculture) and expenditure on interest
payment constitute the most important components of current expenditure with 27.9,
30.7 and 9.1 percent respectively. Over 55 percent of the interest obligation has been
on domestic borrowing. Over the 20 years period, current expenditure has grown
steadily at a rate of 12 percent per annum.
Though its share in government expenditure for the 1974/75-2009/10 period was, on
average, less than that of current expenditure, capital expenditure has grown at a
higher rate (21 percent per annum). The share of capital expenditure has overtaken
that of current expenditure in 2006/07 reaching 52 percent. Growth in capital
expenditure is often considered as a welcome development. Yet capital projects do
not go into operation within the period in which expenditures are made and usually
have long gestation period to bear fruit. Thus, at least for some time to come the
money injected into the economy for financing such projects might lead to
inflationary pressure. The magnitude of such inflationary pressure, however, depends
on the source of finance used and the food supply elasticity in the country.
The government has financed its capital expenditure from three sources. About 66
percent of capital expenditures were financed from central treasury while the rest
from external assistance (15 percent) and external loans (19 percent) (MoFED,
2007/08). The lion’s share of the means of capital expenditure financing coming from
central treasury indicates that domestic money creation might have played significant
role in the current inflationary process. Between 2002/03 and 2006/07, the National
Bank of Ethiopia’s (NBE) direct advance to the government has almost quadrupled
while there has been no major change in central government deposit with NBE or the
foreign assets of the NBE (NBE, 2006/07),despite statutory limits to this.

Figure 3.1 Budget deficits as % of GDP


15
budget deficit as %age of GDP
5 0 10

1970 1980 1990 2000


year

(Source: National Bank of Ethiopia data archive)

In general in the last 20 years, government deficit as percentage of GDP has averaged
about 10 percent. To finance this budget deficit, the government resorted to external
and domestic borrowing as well as privatization of public enterprises. For the
1997/98‐2006/07 period external borrowing and domestic borrowing each accounted
for around half of the total deficit with the balance slightly swinging towards
domestic borrowing, while a very small share (4 percent) is taken up by receipts from
privatization proceeds. In the post 2002/03 period, the means of financing the budget
deficit has, shifted from external to domestic bank and non-bank sources, especially
following the 2005 election where donors protested over the election by reducing the
amount of aid and latter changing modality of its delivery from budget support to
provision of basic services (PBS). This has led to the monetization of the deficits. The
use of domestic means of deficit financing as percentage of budget deficit has grown
from 34.4 percent in 2002/03 to 63.8 percent in 2009/10. The budget deficit and the
means of financing it thereof might have played a major role in the current
inflationary process. As can be seen from Figure 2.3, before 2002/03, budget deficits
(including grants) were not associated with inflation mainly due to the conservative
monetary policy stance of the government. After 2002/03 however, any given level of
budget deficit appear to have been associated with ever growing inflation. Domestic
borrowing reached over Birr 6 billion (the net stock of government debt being 42.3
billion birr which doubled since 2002/03) in 2006/07, with over 70 percent coming
from the banking system (see NBE, 2006/07; MoFED, 2007/08).
The budget deficit as a ratio of GDP reflects increasing of budget deficit. As shown in
fig. 3.1 budget deficit as a % of GDP were 3.5% in 1975. From the figure above
largest budget deficit was occurred in 1994/95 which is about 13.5% during post
reform. Average budget deficit as % of GDP was 7.95% during the Dergue regime. In
general we have seen that the overall increasing amount of government expenditure
over revenues has result in a wide budget deficit.
3.3 Government revenue
The government of Ethiopia classified the revenue schedule in to ordinary revenues,
external assistance and capital receipts. Ordinary revenues are further classified in to
direct tax, indirect tax, and foreign trade tax and non-tax revenue.

Fig.3.2 Government revenue as % of GDP


20
revenue as %age of GDP
10 5 15

1970 1980 1990 2000


year

Source: National Bank of Ethiopia (NBE) and own computation


3.3.1 Foreign trade tax
Foreign trade tax which include export duties and taxes, import duties and taxes and
transaction tax on exported as well as imported goods constituted a major part of
government revenue.
As shown in fig.3.1 above foreign trade tax have been the most important source of
government revenue. Until 1980/81 their shares from the total government revenue
were 44.29%, 42.61% and 40.16% during the year 1976/77 to 1978/79. Its
contribution to total government revenue decreased after 1984/85. In the year 1988/89
its contribution was only 13.45% from the total government revenue. After this period
the trend of foreign trade tax contribution to total government revenue increased.
3.3.2 Indirect tax
The other component of government revenue is indirect tax. These taxes constitute
excise and transaction taxes on locally manufactured good and turn over tax on stamp
duties. The relative importance of this tax has been declining although the absolute
amount has been increased.

3.3.3 Direct tax


Direct taxes have remained stable contribution to total government revenue. They
have shown relatively less fluctuation in their contribution to total government
revenue. Direct tax includes personal income tax, tax on business and profit,
agricultural income tax and land use fee. Of these taxes tax on business is the most
important one followed by personal income tax.
3.3.4 Non-tax revenue
A component of government revenue which has shown constant increase in terms of
contribution to the total government revenue is non-tax revenue categories. The
important components of non-tax category are charges and fees, sales of goods and
services, surplus and capital charges.
As can been seen from fig.3.2 above starting from 1970/71 the total government
revenue increase constantly until 1983/84. Then the trend for government sharply
declined until 1985/86. The reason for the sharp decline of the trends in the last two
years of the Dergue regime were political and economic crisis faced by the
governments which have eroded the tax basis and consequently decline the
government tax revenue. However, various reforms were introduced by the post
1990/91 government of Ethiopia that increased government revenue. As a result the
trend for government revenue show a sharp rise and increased constantly then after.
3.4 Government expenditure
In Ethiopia government expenditure is categorized in to current and capital
expenditure. Current expenditure refers to outlays for running government
administration and other social and economic services. On the other hand, capital
expenditure refers to investment outlays on developing project.

Fig.3.3 Government expenditure as % of GDP


30
expenditure as %age of GDP
15 20
10 25

1970 1980 1990 2000


year

Source: National Bank of Ethiopia (NBE) and own computation


3.4.1 Current expenditure
In Ethiopia, government expenditure is more concerned with those expenditure that
do not render returns in the long run, i.e. why the current expenditure takes the major
proportion of total expenditure. From the figure 3.2, the government expenditure
shows the growth of government expenditure through time. During the pre-reform,
the highest government expenditure occurs in 1982/83, which was 20.4% of GDP,
and in 1988/89 was about 23.4% of GDP. However, post reform period the highest
government expenditure occurs in 200/02 which was about 29% of GDP and after this
time it showed decreasing trends.
In general the share of current expenditure to total government expenditure during
pre-reform period on average was 73.8% while it was 64.3% in post reform period.
During pre and post reform period annual growth rate of current expenditure to over
all expenditure has dominated government expenditure for whole period of the study.
3.4.2 Capital expenditure
Capital expenditure is broadly defined as an outlay on development project that result
in the acquisition on fixed and there by enhance the capacity of the economy for the
production of the goods and services. Such outlays include spending on construction
and purchase of machine and equipment. It also includes payment for project study
and design, management provision and direct labor costs.
Significant increases have been witnessed in the capital expenditure during the post
reform period. It grew by 26.5% and 29.2% in nominal terms successively been
1992/93 and 1995/96. This was triggered by an all out effort for rehabilitation and
reconstruction of war devastated infrastructure and up lifting the status of neglected
regions. Such a significant annual increase lifted the share of capital expenditure in
total expenditure from 22.63% in1994/95 to 45.24 in 1995/96. Annual average growth
rate of capital expenditure was 13.8% during pre reform period while it increased to
22.93% on average during post reform period. On average, capital expenditure was
34.77% of the total government expenditure in the post reform period while it was
26.52% during pre-reform period.
Since 1970/71 the government expenditure as a whole shows an increasing path. Its
annual growth rates which was 7.79% and 5.43% during the year 1972/73 and
1973/1974 shoot to 41.12 on 1974/75. After 1974/75 the government expenditure
shows a sharp rise. The major reason for the sharp increase after 1974/75 was the
increase in the public sector economy. During the Dergue regime, the economy was
put under the state control with considerable socialist transformation. As a result, the
whole production and distribution processes were mainly carried out by the public
sector.
The government expenditure sharply declined from 1989/90 to 1991/92 annual
growth rate during those periods -7.73, -8.12 and -13.37. The reason for the sharp
decline was a cut in government expenditure as a result of a huge decrease in tax
revenue.
The sharp declined trend was reversed after 1992/93. Annual growth rate of
government expenditure were 24.11% and 35.91% during the year 1992/93.

CHAPTER FOUR
ECONOMETRICS ISSUES
4.1 model specification
Cambes and Saadi-Sedik (2006) used simple theoretical model of budget deficit
determination. To find the determinants budget deficit during the period of 1995 up to
2006. Cambes and Saadi-Sedik specify the following models.
BDi=f (CBIi, lnRGDPi, GRGDPi,URBi, AGRIi, ILLYi, OPENi)
Where i=denotes the time
Where Ei=error term and variables are defined as follows
BDi = denotes the government budget deficit
CBIi= central bank independence
LRGDPi= the log of real per capita GDP
GRGDPi= the real GDP annual growth rate
URBi=the degree of urbanization
AGRIi=the share of agriculture in the GDP
ILLYi= the ratio of liquid liabilities of the financial system to GDP
OPENi=the trade openness
The linear equation is therefore
BDi = BO+ BoCBIi + B1lRGDPi+ B2GRGDPi+ B3URBi+B4 AGRIi+B5 ILLYi+B6
OPENi+ Ei
Following the model specified by Cambes and Saadi-Sedik the budget deficit model
of this study can be specified by using the share of agriculture in the GDP, real GDP
annual growth rate, log of real per capita GDP, real interest rate, real exchange rate,
inflation rate, and broad money supply printed as explanatory variables and budget
deficit as dependent variable.
BDi= f (LRGDPt, GRGDPt, RERt, AGRIt, RIRt, LM2t, IRt)
Where t=denotes time
BDi = denotes the government budget deficit
LRGDPi= the log of real per capita GDP
GRGDPi= the real GDP annual growth rate
AGRIi=the share of agriculture in the GDP
RERI=real exchange rate
RIRt= real interest rate
LM2t= money printed
IRt=inflation rate
The estimated equation is therefore as follows:
BDi = BO+ B1lRGDPt+ B2GRGDPt+B3AGRIi+ B4RERt +B5RIRt+B6M2t+B7IRt+Et
Et= error term
The dependent variable is the budget deficit in the percent of GDP and is taken from
National Bank of Ethiopia (NBE). This measure of budget surplus which excludes net
interest payments of the treasury is probably the most appropriate measure of the
fiscal discipline because it assesses the orientation of the fiscal policy over a year.
Furthermore, from an econometric point of view, this measure allows us to avoid a
potential simultaneity bias between the dependent variable and the principal
explanatory variables.
LRGDP: Real GDP per capita is introduced in the function to indicate the level of
economic development. So higher level of income per capita reflecting a higher level
of development is held to indicate greater capacity to levy and collect taxes (Mankiw,
7th, 2007). This is true concerning the management and efficiency of public
expenditures. Moreover, according to Roubini (1991), the GDP per capita may
capture some sociopolitical effects if social conflicts are more important in poor
countries. The of GDP per capita is positive.

GRGDP- Real GDP annual growth rate- is included in the model as a proxy for
economic activity because government budget balance is sensitive to economic
fluctuation. Indeed , when the level of economic activities low or moderate the
amount of tax revenues collected by the government decreases while social
expenditure increase, that lead to a deterioration of budget balance. Conversely, a
higher economic growth generates an improvement of budget balance (automatic
stabilizers). However, some authors (Talvi and Vegh, 2000) have suggested that fiscal
policy can be procyclical in developing countries with weak governments, because
political pressures to increase public spending go hand in hand with the growing tax
revenue due to higher economic growth. The strong increase in the fiscal demands
during economic boom is called “voracity effect” (lane and tornell, 1999).
AGRI- share of agriculture in the GDP-According to Tanzi (1992) country’s
economic structure is an important factor that could influence the level of taxation.
For this reason, the share of agriculture in GDP is included. Nevertheless, the
expected sign of AGRI is uncertain because the theory distinguishes two opposite
effects of the share of agriculture in GDP on the tax share. Thus, on the supply side
the share of agriculture in GDP: is expected to have a negative effect on tax revenues
because political constraints could encourage the government to cut taxation in this
sector, often heavily taxed in many implicit ways through import quotas, tariffs,
collected prices of output or overt valued exchange rates. Moreover, agricultural
sector in LDC’s is mainly characterized by subsistence farming and the predominance
of small farmers and so, it appears difficult for a government to tax the main foods
that are used for subsistence (Stotsky and Wolde Miriam, 1997). Conversely on the
demand side, the share of agriculture in GDP is expected to have a positive effect on
budget surplus because many public sector activities being city oriented, the demand
for public goods and services and so the public expenditures are theoretically reduced
(Teere, 2003).
M2 :( Woo, 2003) included as a process for the financial market development level,
the so called “financial depth”. According to Woo (2003) “countries with higher
developed financial markets can more easily finance the budget deficit by issuing
bond, without having resort to inflationary finance.

IR: Keynesian believes that the great depression could have been averted if the
government had engaged in more government spending and income tax cuts (fiscal
policy). Inflation reduces real tax revenue and thus causes high fiscal deficit. This is
called the inflationary approach to fiscal deficit. Sometimes the situation may be
described as “olvera –Tanzi effect” of fiscal deficit, i.e. deficit caused by a decline in
real tax revenues during period of high inflation as a result levels reduce real tax
revenues of the government significantly as the government collects and accounts its
tax receipts in latter and expenditures causing high budget deficit in the country
(Carlos Rodriguez: 1994).
RIR: According to Keynesian economics lower real interest rate will lead both the
consumer and business capital spending both of which increases equilibrium national
income. This is because lower interest rates induce investment which in turn increases
output and hence increases revenue. Higher interest rate inhibits investment; lowers
output and hence reduces revenue.
RER: export supply of the country decreases due to the appreciation of exchange rate
then this aggravate the growth of public fiscal deficit.
4.2. Results of ordinary least square (OLS) estimation
BD=-23.99426--542.3924LRGDP+5.658793IR-839.3519LM2+506.6771RIR--
6.019493AGRI-1037.398RER-5.719545GRGDP
The value of the constant term -23.99426, which is also significant, shows that BD
will have a value of -23.99426 units if all the explanatory variables (included in the
model) are zero. It may also imply the impact of excluded variables on BD, other
variables kept constant.
As economic theoretically, holding other variables constant, as real GDP per capita
increases by 1 million birr budget deficit decreases by 542.3924 units. The
relationship between BD and real GDP per capita negative, which is not confirmed
with hypothesis of the study. This relationship is insignificant. Economic theory say
that higher level of income per capita leads to higher level of development which
indicates greater capacity to levy and collect taxes.
Inflation rate is positively related with budget deficit and which doesn’t conform to
the hypothesis of the study and its t-value shows that it not statistically significant.
When there is high inflation there is high budget deficit because taxes collection
exhibits lags and tax revenue would decrease at the time of collection.
Budget deficit and money printing have negative relationship which does confirm
with hypothesis of the study and its t-value shows that it’s statistically significant.
When the broad money supply in the country is high due to inflow of money from the
foreign the expenditure of government increase this leads to high budget deficit.
Real interest rates have positive relationship with budget deficit and its t-prob is
statistically significant and it’s confirmed with the hypothesis of the study. This result
contrary to economics theory which say that when real interest rate is low the
investment expands and the GDP expands which leads to low budget deficit.
Budget deficit have negative relationship with agriculture share of GDP and it’s
statistically insignificant which is confirmed with the hypothesis of the study.
Agriculture in Ethiopia is mainly characterized by subsistence farming and the
predominance of small farmers and it appears difficult for government to tax the main
foods that are used for subsistence. Due to this government revenue decreases and
budget deficit becomes high.
An increase in the real exchange rate by1 birr, keeping other things constant, reduces
budget deficit by 1037.398. This relationship is found to be significant. As theoretical
expectations Real exchange rate is negative relationship with budget deficit and it’s
statistically significant which is not confirmed with hypothesis of the study. When
real exchange rate is overvalued this leads to expansion of exportable and import
competing sectors and the GDP increase this leads to low budget deficit.
Real GDP annual growth rate have negative relationship with budget deficit and it not
statistically significant even though it not confirmed with the hypothesis of the study.
When the level of economic activities is low the amount of tax revenues collected by
government decreases while social expenditure increases (further information refer
appendix ).
4.2.1 Testing stationary
To estimate a more specific relationship between budget deficit and its determinants
we have to sure that the time series data is stationary. Most economic data are non-
stationary (random walk). There exists a trend element in which both the independent
variables and dependent variables grow up ward or decreases down ward
continuously together.
Running ordinary least squares (OLS) on this data give higher R2 which seems as the
explanatory variables well explain the regressed. However, the higher magnitude of
the multiple coefficient of determination (R2) arises from spurious (false) relation
ship between the dependent and independent variables (Thomas, 1993).
If the time series data are found to be non-stationary most of classical assumption for
econometric estimation will be violated clearly, if available data mean and variances
will change over time. In such cases econometric results may not be ideal for policy
making because the OLS estimation gives inconsistent estimates (Gujarati, 1995).
In the past the popular method to overcome the problem was to estimate the rates of
changes between variables instead of using their absolute levels. However, this
doesn’t capture the long-run information of the model. Besides, the disturbance term,
this is obtained after first differencing auto correlates.
The common tests used are Dickey Fuller (DF) and Augment Dickey Fuller (ADF)
tests. These tests are basically required to ascertain a number of times variables have
to be differenced to arrive at stationary. A time series data are said to be differenced
of ordered ‘p’ if it become stationary after differencing it ‘p’ times. Economic
variables stationary from the outset are I (0) series and a variable that requires to be
differenced once to be stationary is I (1) series (Gujarat, 1995).
Table 4.1 Stationarity test results
Variables ADF-test 1% Critical 5% Critical 10% Critical
value value value
BD -3.592 -3.702 -2.980 -2.622
RGDP -3.858 -3.696 - 2.978 -2.620
CPI -6.014 -3.702 -2.980 -2.62
M2 19.729 -3.689 -2.975 -2.619
NIR -6.220 -3.696 -2.978 -2.620
AGRI -6.437 -3.702 -2.980 -2.622
RER -2.675 -3.702 -2.980 -2.622
GRGDP -4.787 -3.696 -2.978 -2.620
Using the augment dickey fuller (ADF) test real GDP annual growth rates, agriculture
share of GDP and broad money are stationary without differencing the values or with
(0) lags. But budget deficit, real per capita GDP, nominal interest rate, real exchange
rate and inflation becomes stationary after differencing the values of the variables and
with (1) lags.

4.2.2 Testing the significance of the model


The statistical values that measure the goodness of fit and significance of the model
show that the model fits reasonably well. To be specific the R 2 value shows that about
51.101% of the variations in budget deficit are explained by the explanatory variables.

From the above results, it can be shown that 51.01 % of the variation in consumer
price index is explained by the independent variables. In simple the model explains
51.01% of the variation in the dependent variable, i.e. CPI. The adjusted R2 value,
which accounts for the number of variables, shows that the explanatory variables
account for 51% of the variation in consumer price index. The over all significance of
the model is also significant. This shows the variable that the variables incorporated in
model account for the changes in the dependent variable (refer appendix)

4.2.3 F- test
Testing of the significance of the whole explanatory variables simultaneously using F-
test.
F –test H0-all slope coefficients are zero
H1-all slope coefficients are different from zero
F = ESS/Df = ESS/K-1
RSS/Df RSS/n-k+1

The probability that critical F-value at 5% level of significance greater than F-


calculated is 0%, then we will reject our null hypothesis means that all parameters are
different from zero (refer appendix)

4.2.4 Autocorrelation test


In running ordinary least squares estimation the most important assumption is that the
consecutive error terms are not correlated or there is no autocorrelation. Running OLS
estimation by disregarding autocorrelation will result in inefficiency on the estimated
result and its standard errors are estimated in the wrong way.
Therefore detecting autocorrelation is a vital issue for my study since I use OLS
estimation.
Detecting autocorrelation
The Durbin Watson (DW) test
H0- no autocorrelation
H1-autocorrelation exist
Table 4.2 autocorrelation test result
estat Watson
Durbin-Watson d-statistic( 8, 33) = 2.018383

Range of Durbin Watson is between zero and four


If the calculated’ value is much smaller (close to zero) or much longer than two (close
to four) we will reject our null hypothesis which indicate that there is autocorrelation.
However, if the value of‘d’ is expected to be about two there is no serial
autocorrelation. From my regression I get‘d’ value of
This is much closer to two so I accept our null hypotheses which say there is
no autocorrelation. Or depending on the upper and lower‘d’ and finding where our‘d’
would be found and check. The Durbin -Watson value of 2.01883 which is
approximately 2 shows the absence of auto correlation in the model. The R2, the
adjusted R2, the F value and the Durbin-Watson value show that the model is strong.

4.2.5 Multicollienarity test


Multicollinearity refers to case which two or more explanatory variables in the
regression model are highly correlated and making it difficult to isolate their
individual effects on the dependent variable. Using VIF test since the mean of
variance inflation factor is below ten which is 6.04 there no Multicollienarity problem
in the model (refer table 4.3).
Table 4.3 Result of testing Multicollinearity
Variable VIF 1/VIF
RGDP 15.55 0.064300
IR 12.79 0.078188
M2 4.30 0.232796
RIR 3.97 0.252007
AGRI 2.06 0.486468
RER 2.00 0.499066
GRGDP 1.62 0.616027
Mean VIF 6.04

CHAPTER FIVE
5.1 Conclusion and Policy Implications
It was attempted to achieve objectives of the study with the data obtained from NBE
using two techniques of analysis. This chapter aims to link the objectives of the study
with the result obtained and draw some policy implications. It has been stated that
fiscal deficit are at the forefront of macroeconomic problems after the 1980s and
1990s in both developing and developed countries. The result of works on deficit has
been mixed that some empirical work have resulted positive relationship while other
have found clear cut negative relationship between budget deficit and its financing.

As to the causes of rising budget deficits, it has been argued that it was mainly caused
by the growth of expenditure over revenue. Apart from this, there were four reasons
discussed in the literature as to why government may open to high fiscal deficit level;
political reasons, that governments may deliberately favor high spending levels and
low tax rates to make there government legitimate; structural reasons; which makes
the economy inflexible in the short term; inflation, that reduces the real balance of tax
revenue as a result of the existence of collection lags and finally development
theorizing.

The descriptive analysis has reveled that expenditure has been persistently growing
due to the growth of the public sector economy. In turn, this has created budget deficit
since it was not followed by equally proportionate growth revenue. Since 1974, large
and rapid expansion of state activity in the economy has led to the growth of both
government revenue and spending with the dominance of the latter. As a result,
budget deficit have been growing over time.
Looking further in to the structure of current expenditure and capital expenditure the
relative share of the current expenditure was high.

For much reliable conclusion econometric analysis was employed to examine the
empirical relationship between budget deficit and it’s financing. To this effect, the
model specified in chapter four is estimated using the available data and the ordinary
least square method.
In the econometric analysis, the attempt has been made to test the hypothesis that
financing budget deficit through money creation leads to high inflation, domestic
borrowing to finance budget deficit leads to crowding out effect and excessive resort
to foreign borrowing for financing budget deficit leads to high external debt a service
burdening.
To reduce budget deficit the government should restructure its budgeting system and
mode of financing. To this end, the policy implication that emanates from the study is
the following.
 To increase tax revenue and decreases the deficit problems, fundamental
reform of tax structures should be made and the reform should focus on
broadening the tax bases (as opposed to mounting high tax rates), minimize
tax exemption and improve the tax administration system which would affect
the tax collecting systems.
 The government should also exercise control over its expenditures and its
financing techniques especially through limiting its domestic borrowings from
banks.
 Finally the fact that GDP growth in the country follows agricultural growth
trend implies that agriculture is the key to economic growth in the country.
Since the sector is subjected to several of the nature which are beyond policy
measures and control, due emphasis should be given to other sector of the
economy. Apart from policies directed towards improving productivity in
agriculture. In addition, the government should, not only formulate, but also
implement appropriate policies to further encouraging private sector
investment and saving which should gear the country to the pace of rapid
economic growth.

APPENDIX

Results of OLS estimation


Number of
Source SS df MS = 34
obs
F( 7, 26) = 3.89
Model 164.794852 7 23.5421217 Prob > F = 0.000
Residual 157.464978 26 6.0563453 R-squared = 0.5114
Adj R-
= 0.3798
squared
Total 322.25983 33 9.76544939 Root MSE = 2.461
BD Coef. Std. Err. t P>t [95% Conf. Interval]
LRGDP -542.888 405.0244 -1.34 0.192 -1375.428 289.6516
IR 5.18894 28.43587 0.18 0.857 -53.26182 63.6397
LM2 -850.5063 418.7873 -2.03 0.053 -1711.336 10.32327
RIR 509.7333 154.491 3.30 0.003 192.1725 827.294
AGRI -6.17865 8.713583 -0.71 0.485 -24.08968 11.73238
RER -1050.251 294.6288 -3.56 0.001 -1655.869 -444.6324
GRGDP -7.117993 8.903583 -0.80 0.431 -25.41957 11.18358
_cons 24.37767 6.620313 3.68 0.001 10.76942 37.98591
Stationarity test
1. Dfuller budget deficit, lags (0)
Dickey-Fuller test for unit root Number of obs = 33

---------- Interpolated Dickey-Fuller ---------


Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z (t) -3.570 -3.696 -2.978 -2.620

MacKinnon approximate p-value for Z (t) = 0.0064

2. Dfuller Lreal percapita GDP, noconstant lags(0)

Dickey-Fuller test for unit root Number of obs = 33

---------- Interpolated Dickey-Fuller ---------


Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z(t) -3.838 -2.647 -1.950 -1.603

3. Dfuller inflation rate, drift lags (0)

Dickey-Fuller test for unit root Number of obs = 33

----------- Z (t) has t-distribution -----------


Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z (t) -6.748 -2.453 -1.696 -1.309

P-value for Z (t) = 0.0000

4. Dfuller broad money, drift lags (0)

Dickey-Fuller test for unit root Number of obs = 33

----------- Z (t) has t-distribution -----------


Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z (t) 3.726 -2.453 -1.696 -1.309

P-value for Z (t) = 0.9996


5. Dfuller real interest rate, drift lags (0)

Dickey-Fuller test for unit root Number of obs = 33

----------- Z (t) has t-distribution -----------


Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z (t) -6.220 -2.453 -1.696 -1.309


P-value for Z (t) = 0.0000
6. Dfuller agriculture share of GDP, drift lags(0)
Dickey-Fuller test for unit root Number of obs = 33
----------- Z (t) has t-distribution -----------
Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z (t) -9.077 -2.453 -1.696 -1.309

P-value for Z(t) = 0.0000


7. Dfuller real exchange rate, noconstant lags(1)

Augmented Dickey-Fuller test for unit root Number of obs = 32

---------- Interpolated Dickey-Fuller ---------


Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z(t) -1.905 -2.649 -1.950 -1.603


8. Dfuller real GDP annual growth rate, noconstant lags(0)

Dickey-Fuller test for unit root Number of obs = 32

---------- Interpolated Dickey-Fuller ---------


Test 1% Critical 5% Critical 10% Critical
Statistic Value Value Value

Z(t) -7.186 -2.649 -1.950 -1.603

REFERENCE
Befekadu Degefe, Berhanu Nega, Getahun Tafesse, 200/01 “Second annual report on
the Ethiopia economy” volume two.
Carlos Rodriguez, westerly, k. Schmidt Hebbel, 1994 “fiscal performance with fixed
exchange rates, in carols Rodriguez, world bank, Washington DC.
Clayton, 1995 “economics principles and practices” video disc edition.
Easterly W and Fisher’s (1990)”The economics of the government budget deficit
constraint” the World Bank research observer, VOL-5.
Gujarati, N, Damodar (1995),”basic econometrics”, 4th edition, MC Graw Hill,
network
Mankiw, G. (2007)”macro economics”, 7th edt. Worthy publisher.
National bank of Ethiopia (NBE), various annual bulletins.
Rudiger Durnbusch, Stanley Fischer and Richard Startz, 2001 “macroeconomics” 8th
edition.

INTERNET SOURCES
Http //:www.William easterly .files.wordpress.com/2010/0
http://apcentral.collegeboard.com/apc/public/repository/
ap06_macroecon_syllabus1.pdf
http://www.cardiff.ac.uk/carbs/econ/workingpapers/papers/E2008_25.pdf
http://swbplus.bsz-bw.de/bsz118062182kap.pdf

You might also like