International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

FISCAL DEFICITS AND PRIVATE INVESTMENT:

ECONOMETRIC EVIDENCE FROM NIGERIA

BY
VINCENT N. EZEABASILI, Ph.D, CNA, HCIB
Department of Banking and Finance,
Anambra State University, Igbariam Campus, Anambra State, Nigeria
And
CLEM I. NDIKIFE NWAKOBY, PhD
Department of Banking and Finance
Nnamdi Azikiwe University, Awka, Anambra State, Nigeria

ABSTRACT
There has been considerable controversy about the possible crowding out effect of
government expenditure in general and particularly deficits on private sector
investments. This paper reexamines the controversial relationship within the
Nigerian context, using data over 1970-2006. A modeling technique that
incorporates co-integration and structural analysis was adopted. Evidence shows
that there is a positive long run relationship between private investment and real
growth of the national economy. This confirms the relevance of the accelerator
principle to Nigeria, with contemporaneous accelerator parameter of 1.84. On
aggregate, a 1% improvement in national income engenders 1.84% increase in
private investment in Nigeria. In addition, the result indicates that fiscal deficits has
had a depressive effect on private investment in the country. The estimation results
suggest that a 1% increase in fiscal deficit leads to 0.267% decline in private
investment. The results also indicate that Nigerias debt profile has had strong and
negative impact on private investment in Nigeria.
Keywards: Fiscal Deficit, Investment, Co-integration, Nigeria
i. INTRODUCTION
A commonly observed phenomenon in both developed and developing countries is the dominant role
of the public sector in initiating and financing economic growth (Ezeabasili, 2009). The resultant
growth in public spending is expected to be financed by public revenues from taxes and non-tax
sources but the revenues always lag behind the level of public spending leaving large deficits in the
process. The macroeconomic theory concerning deficit financing has undergone considerable
transformation since the Keynesian revolution of the early 1940s (Anyanwu, 1997).
1

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

Although deficits were common before the emergence of the Keynesian theories which advocated the
management of the economy by government through the use of fiscal policy. Thus the magnitude of
government fiscal surplus of deficit remain one of the most important statistics used to measure the
impact of government fiscal policy on the economy (Siegal, 1979; Tanzi and Blejer, 1984). In
advanced countries like United States of America, the Federal deficit provided the impetus for a reassessment of the effect of fiscal deficits on economic activities (Islam and Wetzed, 1991).
In less developed countries including Nigeria fiscal deficits have been blamed for much of the
economics crises that beset them since the 1980; over indebtedness and the debt crises; high inflation;
poor investment performance and sluggish growth (Onwioduokit, 1999). Given the volatile revenue
base of government and upwards tread in government expenditure pattern over the years, the
occurrence of fiscal deficits has become inevitable as it should be regarded as an essential element in
the developmental process (NCEMA, 2004). However, the possible adverse effect of large fiscal
deficits has been a subject of wide discussion in the popular newspapers and magazines. The
argument is that large fiscal deficits increase national debt and interest rates. Again, in the
macroeconomic literature, the debate focuses upon the so- called crowding out effect (Albatel,
2004).
Thus fiscal deficit is seen as government borrowing competition with the private sector, which will
result in fewer funds available for private sector investments as interest rates will rise. According to
Albatel (2004), the debate on the crowding out effect of private investment focuses on the impact
of the method of financing of deficits which affects the composition of private wealth. In Nigeria,
budget deficits, were generally financed by excessive borrowing from the banking sector and external
sources (NCEMA, 2004), the Central Bank of Nigeria (CBN) accounted for a large proportion of the
financing from the banking sector (CBN, 2004). Again empirical evidence in most developing
countries has shown that the method of financing have resulted in high monetary expansions, high
inflation, high public debt, exchange rate depreciation, deterioration in balance of payments, sluggish
or negative growth rates, high interest rates including crowding out of private investments,
corruptions, financial sector distress and unemployment (Ezeabasili, 2009; Onwoduobet, 1999).
Nigeria has had large and persistent fiscal deficits in most of thirty one (31) years out the thirty seven
(37) years under study, leaving only six (6) years in surplus. Several studies have dealt with the effect
of fiscal deficit, either by testing for the Keynesian proposition or the Ricardian equivalence
hypothesis. Most of these studies cover developed countries and other contexts different from
Nigeria. Even the few studies done in the Nigerian Context used shorter time horizon in their
estimation. Again the methodologies employed in these studies are not robust enough to interrogate
research data. These observed lapses have definitely left a trail in knowledge gap in the literature on
the effect of fiscal deficit on private investment. This underscores the need for the study
The present study attempts to present a more systematic examination of the relationship between the
variables by firstly using a data set spanning 1970-2006, which is longer than those used in previous
studies. Using recent developments in time series econometrics as provided by Engle and Granger
(1987), Andrew (1991), Phillips and Perron (1988), Dickey and Fuller (1979), Newey and West
2

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

(1994) Mackinnon (1996), Johansen (1988, 1991), Engsted and Bentzen (2001), this work is able to
estimate the relationship between the variables in the model adopted to the study of the relationship
between fiscal deficits and private investment.
The paper is organized as follows:
Following the introduction is section ii which overview the Nigerian Economy, section iii discuses
the theoretical background and literature, while section iv presents methodology, section v presents
data estimation and results of analysis, while section vi concludes the papers.
ii. OVERVIEW OF NIGERIA ECONOMY
The oil boom of the 1970s brought with it fundamental changes in the Nigeria economy. In the period
before independence and up to 1973, Nigeria relied heavily on agriculture as the main stay of the
economy. By 1973 with the emergence of oi8l boom, the c countrys main source of revenue changed
from agriculture to oil, and hence the economys heavy dependence on oil up till date. By 1980, the
oil sector provided about 80 per cent of government revenue and over 96 per cent of encouraged
import oriented production and consumption pattern with the little incentives for non-oil production
and exports. According to Anyanwu and Oaikhenam (1995), the competitiveness of the agricultural
sector in the world markets was eroded by overvalued naira exchange rate, inadequate pricing
policies, rural-urban migration and neglect arising from the oil syndrome. Thus its share of the 40 per
cent of government revenue in the early 1970s fell to 20 per cent in 1980.
In fact, low productivity in the agricultural sector became so acute that Nigeria became heavily
dependent on imported food and agro-allied industrial inputs. Moreover, the public sector became the
prime mover of the economy through huge investments from growing oil revenue in physical,
economic and social infrastructure such that in 1980 the government accounted for about 50 per cent
of GDP and over 70 per cent of modern sector employment (Anyanwu and Oaikhenam, 1995).
The period of mid 1981 witnessed the oil glut as the world oil market began to collapse. The resultant
fall in oil exports and prices were reflected in the government revenue. For example, crude oil prices
which rose rapidly from US $20.94 per barrel in 1979 to US $36.95 in 1980 and US $40 in 1981 fell
to US $29 in 1983 and to a low US $14.85 in 1986. It continued falling in the 1990s down to a low
US $14 in 1998. By 1999 crude oil price started rising again and as at 2004 it stood at about US $63
per barrel. Consequently, external reserves rose from US $175 in 1970 to about US 43 billion in
2007. In the face of rising imports; government deficits widened and efforts at containing the adverse
developments created some other serious problems such as economic depression, rising
prices/inflation from 21 per cent in 1981 to 73 per cent in 1995, unemployment and persistent balance
of payment deficits (Anyanwu, 1993). These created both internal and external imbalance. All these
shocks led to major deterioration in macroeconomic variables.
iii. Theoretical Background and Empirical Literature
Fiscal policies affect private investment through four major channels: They include, public
investment (or public capital), public deficits, corporate tax and investment incentives, and the user
cost of capital or real interest rate. Public capital could be a close substitute for private capital, and
will drive down the rate of return on private investment. Governments also invest in activities that do
3

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

not attract private sector investment like infrastructure for which it is difficult to charge user fees, but
that raise the return of other private projects. Thus the higher the complementarily of public and
private capital, the more likely that public investment directly affect private investment. Again, if
there is domestic financial repression of interest rates with the public sector given preferential access
to credit to finance its deficits, the implication is that the later will crowd out private investment
directly. This is because higher credit to the government may mean fewer funds available for private
sector. With the lack of access to credit by a large number of investors, investment has to be financed
by retained earnings. Net profit therefore, plays a vital role in growth of private investment. The
interest rates have been controlled in Nigeria, thus nominal interest rate ceiling accompanied by high
rate of inflation have led to poor private investment in Nigeria. But if there is no financial repression,
domestic financing of deficits tend to raise real interest rates and hence reduces the profitability of
investments when the cost of doing business rises due to high cost of funds (thereby lowering the
private investments) as the user cost of capital is raised.
Public investment might have an additional effect on private investment depending on whether it is a
substitute or complement to public investment. The argument here is that, if the government invests
in areas that the private sector would not invest in any way, or if the government undertakes
investment activities that would make private investment unprofitable, then higher public investment
would tend to lower private investment. On the other hand, if public investment consists of activities
that raise private investment and which the private sector does not find profitable to engage in
(example: road construction, rail transportation) then higher public investment may raise private
investment. The other factors that are expected to be significant in determining private investment
levels are corporate tax rates, investment incentives (subsidies), and the general investment climate,
caused by uncertainty regarding future government economic policy, political instability, etc.
For instance, the profit tax and investment incentive structure affects after-tax profits and the user
cost of capital. Although the general investment climate in Nigeria has not been particularly
encouraging in most of the past 37 years, uncertainty regarding the profitability of investment does
not seem to be a problem, agents are willing to invest more given the current economic environment
but are constrained by other factors such as access to credit and government policies. However, the
issue of private investment is presently of particular importance in Nigeria especially, when it is
believed that, to sustain high growth rates while simultaneously reducing government involvement in
the economy requires a substantial increase in private investment.
A number of studies have focused upon the possible effects of United States Federal borrowing on
interest rates. For instance Hoelscher (1983) and Makin (1983) empirically examined the possible
effects of federal borrowings in United States (US) on short term interest rates and three months
treasury bills respectively. Their funds are similar. While Hoelscher (1983) finds that federal
borrowing is a relatively unimportant determinant of short term interest rates, the study concludes that
federal borrowing does not have financial crowding out effects on investments. Again Makin (1983)
findings also affirms that federal borrowing has very little (if any) impact on three months treasury
bill rates and in the overall, the result reports that possible significance of crowding out can only be
judged as mixed to weak (Albatel, 2004).
4

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

Evans (1985) study on the impact of federal deficits upon three months treasury bill rate for four
separate time periods in the United States finds no evidence of positive relationship between deficits
and interest rate. However, Barth et al (1985) reports that after adjusting the federal deficits for the
effects of cyclical activity, the resulting structural deficit has a positive and highly significant impact
on 3 months treasury bills. The study by Akpokodge (1998) on the effect of fiscal deficit on private
investment in Nigeria reports that fiscal deficit is capable of contracting private investment as 1 per
cent increase in fiscal deficit leads to a 61 per cent decrease in private investment. This however
confirms the crowding out effect of private investment by fiscal deficit. Again, Paiko (2012)
examines the implication of deficit financing on private sector investment in Nigeria. The study
findings are that a negative relationship exists between deficit financing and investment in Nigeria;
and that deficit financing in Nigeria crowds out private investment.
The work by Albatel (2004) on the effect of government budget deficit on the crowding out of private
sector investment in Saudi Arabia shows that there is a crowding out of private investment by
government budget deficit. The study further suggests that financing government budget deficits by
borrowing from domestic markets reduces financial resources available to the private sector and
discourages private sector investment. The empirical work of Oussou and Bouabre (1993) in Cote
DIvoire reveals that, improvements in the use of public investments are more likely to bring about
additional tax revenues than an increase in private investments themselves. This result is in
consonance with that of Blejer and Khan (1984), as the findings of the latter revealed that public
investments affect private activities through their level rather than their changes. Ariyo and Raheem
(1991) cited in Tchokote (2004) demonstrates the evidence of crowing-in of private investment by
public investment in Nigeria.
In Kenya, Njeru and Randa (1998) found that an increase in money supply eases the constraints in
borrowing for the private sector and thus encourages the entrepreneurs to invest and that public
saving has a negative impact on private investment and this shows the degree of complementarily
between private and public investment. The results of their estimates also revealed that the short run
impact of fiscal deficit on private investment is broadly similar to the long run. The divergences in
the empirical studies were also confirmed with the work of Blejer and Khan (1984). The work reveal
a positive relationship between public capital and private investment in Pakistan, Thailand and
Zimbabwe. However, an inverse interaction was found between these variables in countries like
Chile, Ghana, Mexico and Columbia by the same (Bleyer and Khan, 1984).
Schmidt-Hebbel and Muller (1991) found from their study of Morocco that the causes of decline in
private investment in the eighties were great uncertainty about policy (proxied by foreign debt), a
rapid capital, a more stringent credit policy, and reduced public capital. They conclude that fiscal
stabilization, a consistent foreign debt policy, more investment in public infrastructure and a reform
of investment codes would increase private investment and growth. Again Aleshina et al, (1999)
found a sizeable negative effect of public spending and in particular of its public wage component in
business investment. Sturn (2001) studied 123 non-OCED countries and found that politicoinstitutional variables like ideology, political cohesion, political stability and political business cycles
5

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

do not seem to be important when explaining government capital formation in less-developed

economies. On the other hand, variables like public and private investment and foreign aid are
significantly related to public capital spending. Serven (1996) posits that in the long-run, capital for
public infrastructure projects crowds in private capital and other types of public capital have opposite
effect. But in the short-run, both kinds of public investment may crowd out private investment.
iv. METHODOLOGY
SOURCES OF DATA
The secondary data for the period 1970 to 2006 which were used as the macroeconomic variables in
this study were obtained from the Statistical Bulletin of the Central Bank of Nigeria. The choice of
this intervening period is informed by the following reasons:
i)
this period produced bouts of large fiscal deficits, interspersed, especially in the early 1970s,
with large surpluses in 1971, 1973 and 1974, and again in 1979, 1995 and 1996.
ii)
availability of data in Nigeria, and desire to capture the period of structural break-control
regime vis--vis deregulation and reforms.
Private investment model is therefore
PI = f (DCP, EXTD, FD, GDP, INT)
PI = f0 + f 1DCPR + f 2EXTD + f 3GDP + f 4 INT + f5 FD + Ut
.(10)
f0 is the intercept and,
f1, f2, f3, f4, f5 are coefficient of the regression equation.
A priori it is expected that f1 > 0, f2<0, f3>0, f4>0 ; f5 < 0;
The dependent variable is private investment PI; the independent variables are domestic credit to
private sector (DCP), External Debt (EXTD), Gross Domestic product (GDP), Interest Rate (INT),
and Ut is error term.
Estimation Technique Cointegration and Error Correction Model (ECM) Estimation
Technique
We investigated the time series characteristics of the data to test whether the variables are integrated.
The Augmented Dickey-Fuller (ADF), as specified in Dickey and Fuller (1979), and Phillips-Perron
(Phillips and Peron, 1988) was employed. For the ADF, the null hypothesis is that the variable being
considered has a unit root against an alternative that it does not. The model for the ADF is as
specified below:
P

yt T yt 1 dt yt 1 t

(11)

i 1

Where yt is the variable being considered, T is the time trend (which is only allowed if significant),
and t is a random error term. The Akaike Information Criterion is used in selecting p (the lag-length)
after testing for first and higher order serial correlation in the residuals. The lagged variables serve as
a correction mechanism for possible serial correlation. The Phillips-Peron (PP) test uses models
similar to the Dickey-Fuller tests but with Newey and West (1994) non-parametric correction for
correcting possible serial correlation rather than the lagged variables method employed in ADF. Also
Bartlett Kernel (Andrews, 1991) is used as an automated bandwidth estimator for lag truncation of
the Newey and West nonparametric correction. The test statistics of the PP have the same distribution
as those of Dickey-Fuller with critical levels as provided by MacKinnon (1996).
The fact that two series are unit roots can be an indication of a long run relationship between the two
series.
6

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

Error correction model

To test for the long run relationships between the variables, we apply the Engle-Granger (1987) two
step cointegration test which uses the residuals from the long run equation estimated with the nonstationary variables, and then test for the existence of unit root in the residual using the ADF
regression and compare the value to an appropriate asymptotic null distribution. If two time series yt
and xt are both integrated of order d (i.e. I(d)), then, in general, any linear combination of the two
series will also be I(d); that is, the residuals obtained on regressing yt on xt are I(d). If, however, there
exists a vector b, such that the disturbance term from the regression (e t = yt - bxt )is of a lower order
of integration I(d-b), where b>0, then Engle and Granger (1987) define yt and xt as cointegrated of
order (d,b).
The economic interpretation of cointegration is that if two or more series are linked to form an
equilibrium relationship spanning the long run, then even though the series themselves may be nonstationary, they will move closely together over time and their difference will be stationary. Their
long run relationship is the equilibrium to which the system converges over time, and the disturbance
term et can be interpreted as the disequilibrium error or the distance that the system is away from
equilibrium at time t. In order to estimate the long run relationship between yt and xt it is necessary to
estimate the static model:
yt = Xt + et
(12)
Although the equilibrium long run relationship can be estimated directly using (12), it is also
important to consider the short run dynamics of the variables under consideration, since the system
may not always be in equilibrium. A simple dynamic model of short run adjustment can be written as:
yt = 0 + 0Xt ++ 1Xt-1 + 1 yt-1 + ut
(13)
Reparameterising and rearranging (13) gives the error correction formulation (ECM):
yt = 0 Xt - (1- 1 ) [yt-1 - 0 - 1Xt-1 ] + ut
(14)
where 0 and 1 are coefficients estimated from equation.
The ECM incorporates both short run and long run effects. When equilibrium holds, [yt-1 - 0 - 1Xt1] = 0. But in the short run, when disequilibrium exists, this term is non-zero and measures the
distance that the system is away from equilibrium during time t. Thus (1- 1) provides an estimate of
the speed of adjustment of the variable yt. For instance, if [yt-1 - 0 1Xt-1] < 0, that is, yt-1 has moved
below its equilibrium level, since (1- 1 ) is negative, it will boost yt , thereby forcing it back to
its long run path. Engle and Granger show that two or more variables are cointegrated of order I(1,1)
if and only if an ECM exists.
The first stage in the Engle-Granger framework is to test whether the variables are cointegrated. This
is accomplished by testing the residuals of equation (9) for stationarity. That is, the null hypothesis of
et being I(1) is tested against the alternate of it being I(0). Although any unit root test can be used,
Engle and Granger advocated the use of Augmented Dickey Fuller tests on the residuals. The second
stage of the EG procedure comprises of estimating the short run ECM itself from the residuals of the
regression of the first stage. That is, having obtained t 1 = yt-1- Xt-1, we estimate equation (14) to
determine the dynamic structure of the system.
We expect that there is a long-run relationship between money supply, economic activity, domestic
interest rates and foreign interest rates. We therefore test for the existence of a cointegrating
7

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

relationship. This is done using the above ECM methodology. In the first step we estimate the
coefficients by OLS and test for the existence of a unit root in the residuals. The analysis is also
supplemented by testing for the number of cointegrating relationships using the Johansen procedure.
The deviations from the long run path are captured at the second stage. When the coefficients of the
lagged residual term from the first stage is negative, it suggests that the system comes back to the
long run path or adjusts. Therefore, there exists an error correction mechanism. The parsimonious
error correction mechanism (ECM) can be specified as:
PIt 1 DCPRt s EXTDt s GDPt s INTt s FDt s PIt 1 ECM t 1 t
Where the variables are defined in equation (1).

(15)

The Johansen Cointegration specification

The finding that many macro time series may contain a unit root has spurred the development of the
theory of non-stationary time series analysis. Engle and Granger (1987) pointed out that a linear
combination of two or more non-stationary series may be stationary. If such a stationary linear
combination exists, the non-stationary time series are said to be cointegrated. The stationary linear
combination is called the cointegrating equation and may be interpreted as a long-run equilibrium
relationship among the variables. The purpose of the cointegration test is to determine whether a
group of non-stationary series are cointegrated or not.
Consider a VAR of order p:
yt A1 yt 1 ... Ap yt p Bxt t

(16)

Where yt is a -vector of non-stationary I(1) variables, xt is a -vector of deterministic variables, and t

is a vector of innovations. We may rewrite this VAR as :
p 1

yt yt 1 i yt i Ap yt p Bxt t

(17)

i 1

Where:
p

p 1

i 1

j i 1

Ai I , i Ap Bxt t

(18)

Grangers representation theorem asserts that if the coefficient matrix has reduced rank r < k,
then there exist k < r, matrices and each with rank r such that / and / yt is I(0). r is the
number of cointegrating relations (the cointegrating rank) and each column of is the cointegrating
vector. The elements of are known as the adjustment parameters in the VEC model. Johansens
method is to estimate the matrix from an unrestricted VAR and to test whether we can reject the
restrictions implied by the reduced rank of .
Forecast Error Variance Decomposition
A shock to any variable in the Vector Error Correction (VEC) model not only directly affects the
variable but is also transmitted to all of the other endogenous variables through the dynamic (lag)
structure of the VEC. An impulse response function traces the effect of a one-time shock to one of the
innovations on current and future values of the endogenous variables. While impulse response
functions trace the effects of a shock to one endogenous variable on to the other variables in the VEC,
variance decomposition separates the variation in an endogenous variable into the component shocks
to the VEC. Thus, the variance decomposition provides information about the relative importance of
each random innovation in affecting the variables in the VEC. The general form of the vector error
correction model (VECM) for estimating the variance decomposition is therefore expressed as:
8

ISSN 2319 6912

(March 2013, issue 2 volume 3)

International journal of Innovative Research in Management

k 1

s 1

s 1

s 1

s 1

s 1

LnPIt 1,k vk ,t p 1,s LnDCPRt s 2,s LnEXTDt s 3,s LnGDPt s 4,s LN INTt s 4,s LN FDt s 5,s LnPIt 1 t
s 1

(19)
where:
p l(ishe
t optimal lag length of the VAR)
i, k= the adjustment coefficients
vk,t p= is the cointegrating vector
i = intercepts
The amount of forecast error variance of variable j accounted for by exogenous shocks to variable k is
given by jk,h :
h 1

jk ,h (e ji ek )2 / MSE[ y j ,t (h)]

(20)

i 0

Where MSE is the mean square error (MSE) of an estimator and is one of many ways to quantify the
difference between values implied by an estimator and the true values of the quantity being estimated.
MSE is a risk function, corresponding to the expected value of the squared error loss or quadratic
loss. MSE measures the average of the squares of the errors. The error is the amount by which the
value implied by the estimator differs from the quantity to be estimated.
The variance decomposition is displayed in which a table format displays a separate variance
decomposition for each endogenous variable. A column in the Table gives the forecast error of the
variable at the given forecast horizon. The source of this forecast error is the variation in the current
and future values of the innovations to each endogenous variable in the VAR. The remaining columns
give the percentage of the forecast variance due to each innovation, with each row adding up to 100.
As with the impulse responses, the variance decomposition based on the Cholesky factor can change
dramatically if you alter the ordering of the variables in the VAR. For example, the first period
decomposition for the first variable in the VAR ordering is completely due to its own innovation.
IV
Data Estimation and Results of Analysis
Table 1: Summary of Statistics of the Variables used in the Regression Analysis
Mean
Median Maximum Minimum Std.Dev. Obs
LEXTD
10.10308 10.45416 10.57844
8.922559 0.546046 37
LDCP
0.013754 0.03356 0.11814
-0.17259
0.075518 37
LGDP
11.12829 11.09621 11.3797
10.89509 0.12569 37
LFD
0.111193 0.066115 0.50274
-0.37609
0.182012 37
LINT
0.913497 1.004665 1.366273
0.425974 0.289786 37
LPINV
10.07345 10.10025 10.591
9.626741 0.245165 37
Source: Authors computation with data from CBN Statistical Bulletin using Econometric views
The characteristics of the data series used in the regression analysis are presented in Table 1. The
table reports the summary of statistics used in the analysis. It provides information about the means
and standard deviations of the main variables. The mean value of log of private investment is
10.07345 while the mean of the log of fiscal deficit and gross domestic product stood at 0.111 and
0.013 respectively.
Table 2: Table of Observed Result of the Augmented Dickey Fuller test (ADF)*

International journal of Innovative Research in Management

VARIABLES

LDCP
LEXTD
LFD
LGDP
LINT
LPINV

LEVEL
Model 1
-2.35211
-3.13116
-3.20642
-0.71521
-1.98285
-1.43743

Model 2
-2.14571
-1.05184
-4.67947
-1.82668
-1.95053
-1.22452

ISSN 2319 6912

(March 2013, issue 2 volume 3)

FIRST
DIFFERENCE
Model 1
Model 2
-6.60492
-6.68085
-1.81189
-6.21898
-5.84943
-5.86981
-5.97603
-5.92072
-2.84996
-3.08591
-4.01565
-3.98471

*The Null Hypothesis is the presence of unit root. Model 1 includes a constant while model 2
includes a constant and a linear time trend. Lags were selected based on Schwarz Information
Criterion. *, **, *** indicate significance at 1%, 5%, and 10% respectively.
The variables for our analysis were subjected to two types of unit roots test to determine whether they
are unit roots or stationary series. The tests employed were the Augmented Dickey Fuller test (ADF)
and the Phillips-Perron test (PP) test. For the ADF and PP tests, two models are considered viz, with
constant, with time trend. The null in both the ADF and PP test is the presence of unit root.
The ADF results in Table 2 show that the variables are integrated of order one in the two models of
unit root test considered. Only a few of the variables are significant at their levels and a reasonable
number of the variables were at the 5% level. One exception was however observable, the log of
external debt (LEXTD) which was found to be stationary in the model that includes a constant
without a linear time trend at levels. This variable is significant at 10% level. One interesting feature
noted in the results was that all the variables were stationary in model with constant as well as
constant and linear time trend at the first difference level. However, the log of the gross domestic
product (LGDP) is not significant in the ADF models that include a constant and time trend, and with
neither constant nor time trend but it is significant in the models that include only constant in first
difference.
Table 3: Table of Observed Result of the Phillips-Perron Test (PP)*
VARIABLES
LDCP
LEXTD
LFD
LGDP
LINT
LPINV

LEVEL
Model 1
Model 2
-2.34833
-2.10134
-2.92619
-0.86302
-3.17008
-4.54522
-0.78224
-2.17373
-1.42665
-1.52242
-1.31499
-1.68372

FIRST DIFFERENCE
Model 1
Model 2
-6.60492
-6.6834
-5.27047
-6.51706
-13.1983
-14.133
-5.98407
-5.92585
-6.96867
-7.09702
-3.98715
-3.94437

*The Null Hypothesis is the presence of unit root. Model 1 includes a constant, Model 2 includes a
constant and a linear time trend while Model 3 includes neither in the test regression as exogenous.

10

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

The Bandwith was chosen using Newey-West method with Barttlet Kernel spectral estimation*, **,
*** indicate significance at 1%, 5%, and 10% respectively.
The PP test statistics reported in Table 3 reinforces the result in the model that include only constant
in the ADF test and also supported those models that include a constant and a linear time trend. The
PP test supports the presence of unit roots in all the series. The few exceptions that were noticed in
the ADF model however remain. For example, the LEXTD was bound to be stationary in the model
that includes a constant without a linear time trend at levels. This variable is significant at 10% level.
It is evident from Table 2 and 3 that the variables become stationary series when appropriately
differenced. From the two types of integration tests carried out (above), it could be concluded that
that all the variables in our model contains unit roots. Therefore, we can safely proceed to use the cointegration method in analyzing our models as conventional regression model will generate spurious
results due to the integration level of the series. Following the findings that the data series are by
nature, mostly non-stationary stochastic processes, econometric developments regarding the concepts
of cointegration are particularly apposite in testing for equilibrium. Accordingly, the long run
properties of the variables in the behavioural equation was examined using the Engle-Granger twostep procedure.
Table 4: Table of Observed Result of the Unit Root Test of Residual of ECM of variables
Equation

Augmented Dickey
Phillips-Perron test
Fuller Test
Private Investment Equation
-5.6334
-5.6136
Note: (1) Lags were selected based on Schwarz Information Criterion in the ADF test (2) The
Bandwith was chosen using Newey-West method with Barttlet Kernel spectral estimation in the
Phillip-Perron test (3) *, **, *** indicate significance at 1%, 5%, and 10% respectively.
Following the findings that the data series are by nature, mostly non-stationary stochastic processes,
econometric developments regarding the concepts of co-integration are particularly opposite in
testing for equilibrium. Accordingly, the long run properties of the variables in the behavioural
equations were examined using the Engle-Granger two-step procedure. Presented in Table 4 above
are the results of the unit root tests of the residuals of the static long run models. The regression
residuals have zero mean, and as they are not expected to have the models that includes constant and
constant with time trend. The ADF test statistics and the Phillip-Perron statistics suggest that the
disequilibrium errors are mostly I(0), and as such, the variables in the static equations are
cointegrated.
Table 5: Table of Observed Result of the Johansen Multivariate Cointegration Test Results for the
Private Investment Equation
Sample(adjusted): 1971 2006
Included observations: 36 after adjusting endpoints
Trend assumption: No deterministic trend (restricted constant)
Series: LPINV LDCP LEXTD LFD LGDP LINT
Lags interval (in first differences): No lags
11

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

Unrestricted Cointegration Rank Test

Hypothesized
Trace
5 Percent
1 Percent
No. of CE(s) Eigenvalue Statistic
Critical Value
Critical Value
None **
0.743159
126.5948
102.14
111.01
At most 1 *
0.608234
77.66008
76.07
84.45
At most 2
0.401325
43.92477
53.12
60.16
At most 3
0.291167
25.45547
34.91
41.07
At most 4
0.218477
13.06662
19.96
24.6
*(**) denotes rejection of the hypothesis at the 5%(1%) level
Trace test indicates 2 cointegrating equation(s) at the 5% level
Hypothesized
Max-Eigen
5 Percent
1 Percent
No. of CE(s) Eigenvalue
Statistic
Critical Value
Critical Value
None **
0.743159
48.93475
40.3
46.82
At most 1
0.608234
33.7353
34.4
39.79
At most 2
0.401325
18.4693
28.14
33.24
At most 4
0.218477
8.874397
15.67
20.2
At most 5
0.109926
4.192222
9.24
12.97
*(**) denotes rejection of the hypothesis at the 5%(1%) level
Max-eigenvalue test indicates 1 cointegrating equation(s) at both 5% and 1% levels
Source: Authors Computation from cointegration test using Econometric views.
In view of the problems with the Engle-Granger framework for testing cointegration. The results were
validated using the Johansen (1991, 1995) approach. This framework provides the number of
cointegrating equations and estimates of all cointegrating vectors in the multivariate case. The
Johansen cointegration test results are contained in Tables 5 above. The trace test and the max-eigen
test were conducted to establish the number of cointegrating relations in each of the equations. The
trace test results are presented in the first part of the table while the max-eigen results were presented
in the second part of the table. Test results indicate the existence of one cointegrating equation in the
equations at the 1% and 5% significance level. In addition, the normalized cointegrating coefficients
show that the variables in the equations are relatively important. The consistency in the test results
confirms the existence of long run relationship among the exogenous and dependent variables in the
model.
As the data series are non-stationary and the vector of variables in the equations appear to be
cointegrated, execution of the second phase of the Engle-Granger technique led to the estimation of
error-correction forms of the stochastic equations. The equations represent the short-run behaviour
and the adjustment to the long run models. The residuals from the cointegrating regressions lagged
one period were used as error correction mechanism in the dynamic equations. The Ordinary Least
Squares (OLS) estimation method was used as it is an essential component of most other estimation
techniques. In addition, the OLS remains one of the most commonly used methods in econometric
12

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

investigations involving large models. Estimates of the preferred specifications obtained using
general-to-specific method are presented in Table 6 and discussed below. The results were evaluated
using conventional diagnostic tests.
The general discussion of the error correction model is useful here (see Tab. 6). All the diagnostic test
statistics are quite satisfactory. The magnitude of the coefficients confirms the absence e of redundant
regressors. Judged by the significance of the t-statistics, the coefficients are well determined. The
disequilibrium error term, ECMt-1, is statistically significant and negative (as expected) in the
equations. The significance of the error terms confirms the existence of long run relationship between
the variables in the error correction models. Of particular interest is the coefficient on the lagged
ECM in the private investment equation. The ECM induces about 56% adjustment per period in these
equations. The lowest adjustment of 56% is recorded by the private investment equation. In addition,
all the equations are statistically significant and the overall statistical fit is good. The marginal
significance level of the F-statistics is almost zero. Hence, the null hypothesis of the F-statistics is
rejected for the choice of significance level. Therefore, the conclusion is that, the regression
coefficients are significantly different from zero. The high value of the Durbin-Watson (DW)
indicates absence of autocorrelation. Finally, the relatively low value of the standard error of the
regression is a clear evidence of the goodness of fit of the equations.
Table 6: Parsimonious Model of Private Investment Equation
Dependent Variable: D(LPINV)
Method: Least Squares
Coefficient
C
-0.006366
D(LDCP)
0.128595
D(LEXTD)
-0.191044
D(LEXTD(-1)
0.280673
D(LEXTD(-2)
0.291853
D(LFD)
-0.267373
D(LFD(-2)
0.062500
D(LFD(-3)
-0.091186
D(LGDP(-1)
1.840189
D(LINT)
-0.162232
D(LINT(-1)
-0.217053
D(LINT(-2)
0.237551
D(LINT(-3)
0.484874
D(LPINV(-1)
1.775864
ECM6(-1)
-0.561412
R-squared
Adjusted R-squared
S.E. of regression

0.909093
0.838388
0.038304

Std. Error
0.010452
0.229130
0.087669
0.100808
0.096330
0.059253
0.050934
0.052231
0.525153
0.098096
0.099771
0.084684
0.102154
0.197448
0.304888
Mean dependent var
S.D. dependent var
Akaike info criterion

t-Statistic
-0.609090
0.561230
-2.179157
2.784243
3.029739
-4.512422
1.227099
-1.745808
-3.504102
-1.653807
-2.175520
2.805141
4.746505
8.994094
-8.401161

Prob.
0.5501
0.5816
0.0428
0.0122
0.0072
0.0003
0.2356
0.0979
0.0025
0.1155
0.0432
0.0117
0.0002
0.0000
0.0000
0.017673
0.095282
-3.383564
13

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

Sum squared resid

0.026410
Schwarz criterion
Log likelihood
70.82881
Durbin-Watson stat.
F-statistic
1.27E-06
Source: Regression results from analysis using Econometric views.

-2.703334
1.513449

The results of the private investment equation in Table 6 are insightful. Evidence emerges that there
is a positive long run relationship between private investment and real growth of the national
economy. This confirms the relevance of the accelerator principle to Nigeria, with contemporaneous
accelerator parameter of 1.84. On aggregate, a 1% improvement in national income engenders 1.84%
increase in private investment in Nigeria. In addition, the result indicates that fiscal deficits have had
a depressive effect on private investment in the country. The estimation results suggest that a 1%
increase in fiscal deficit leads to 0.267% decline in private investment. The results also indicate that
Nigerias debt profile has had strong and negative impact on private investment in Nigeria. This
confirms the relevance of the debt overhang hypothesis. Furthermore, although positive and
insignificant estimates show that the elasticity of private investment to domestic credit is 0.128. This
implies that a 1% increase real credit to the private sector results in a 0.128% increase in private
investment. This underscores the need to significantly increase credit to government in favour of the
private sector. The error correction estimates of 56.1% of the preceding periods disequilibrium are
eliminated in the current period, with intermediate adjustments captured by the differenced terms.
The value of the adjusted R2 shows that the model accounts for at least 83.8% changes in private
investment.
Variance Decomposition Error and Impulse Response
The effect of one-time shock to innovation in current and future relationships between fiscal deficit
private investment, domestic credit to private sector, external debt, gross domestic product and
interest rate using variance error decomposition and impulse response function within a 10- period is
shown below in table 7.

Period
1
2
3
4
5
6
7
8
9
10

Table 7: Variance Decomposition of Private Investment Equation

S.E.
LPINV
LDCP
LEXTD
LFD
LGDP
LINT
0.071885
100
0
0
0
0
0
0.096737 87.63212 0.130316 1.295697 7.049731 1.37001 2.522126
0.113116 68.25204 4.1252
5.9468 12.57415 6.451442 2.650365
0.130604 51.6517 8.22914 11.34807 10.16686 13.79229 4.81195
0.152969 38.00681 9.368967 17.64079 7.433431 20.96855 6.581444
0.173106 30.07511 10.05492 20.61608 5.953979 26.23439 7.065522
0.19387 25.16959 11.24117 21.32082 5.172926 30.64599 6.449503
0.214507 22.51721 12.43997 20.8616 5.174352 33.30723 5.69964
0.232323 20.97662 13.22963 19.90024 5.831005 34.91306 5.149448
0.246411 19.89104 13.77106 18.82373 6.734838 36.04086 4.738475

14

ISSN 2319 6912

(March 2013, issue 2 volume 3)

International journal of Innovative Research in Management

The gross domestic product (GDP) was found to explain about 36.04% of the future variations in the
private investment function. Fiscal deficit was only found to explain about 6.73 of the future
variations in the path of the private investment function. GDP and interest rate generates positive
impulses. Fiscal deficit initially generates positive impulses but start generating negative impulses at
the end of the fifth period.

IMPULSE RESPONSE GRAPH

Response to Cholesky One S.D. Innovations 2 S.E.

Response of LPINV to LPINV

Response of LPINV to LDCP

Respons e of LPINV to LEXTD

Response of LPINV to LFD

Response of LPINV to LGDP

Response of LPINV to LINT

.12

.12

.12

.12

.12

.12

.08

.08

.08

.08

.08

.08

.04

.04

.04

.04

.04

.04

.00

.00

.00

.00

.00

.00

-.04

-.04

-.04

-.04

-.04

-.04

-.08

-.08

-.08

-.08

-.08

-.08

-.12

-.12
1

10

-.12
1

Response of LDCP to LPINV

10

-.12
1

Response of LDCP to LDCP

10

-.12
1

Respons e of LDCP to LEXTD

10

-.12
1

Response of LDCP to LFD

10

Response of LDCP to LGDP

.05

.05

.05

.05

.05

.05

.04

.04

.04

.04

.04

.04

.03

.03

.03

.03

.03

.03

.02

.02

.02

.02

.02

.02

.01

.01

.01

.01

.01

.00

.00

.00

.00

.00

.00

-.01

-.01

-.01

-.01

-.01

-.01

-.02

-.02

-.02

-.02

-.02

-.02

-.03

-.03
1

10

-.03
1

Response of LEXTD to LPINV

10

-.03
1

Response of LEXTD to LDCP

10

Response of LEXTD to LEXTD

10

10

Response of LEXTD to LGDP

.12

.12

.12

.12

.12

.08

.08

.08

.08

.08

.08

.04

.04

.04

.04

.04

.04

.00

.00

.00

.00

.00

.00

-.04

-.04

-.04

-.04

-.04

-.04

-.08
1

10

-.08
1

Response of LFD to LPINV

10

-.08
1

Response of LFD to LDCP

10

-.08
1

Response of LFD to LEXTD

10

Response of LFD to LFD

10

Response of LFD to LGDP

.15

.15

.15

.15

.15

.10

.10

.10

.10

.10

.05

.05

.05

.05

.05

.05

.00

.00

.00

.00

.00

.00

-.05

-.05

-.05

-.05

-.05

-.05

-.10

-.10

-.15
1

10

-.10

-.15
1

Response of LGDP to LPINV

10

Response of LGDP to LDCP

-.10

-.15
1

10

Response of LGDP to LEXTD

10

Response of LGDP to LFD

10

Response of LGDP to LGDP

.04

.04

.04

.04

.04

.03

.03

.03

.03

.03

.02

.02

.02

.02

.02

.02

.01

.01

.01

.01

.01

.00

.00

.00

.00

.00

.00

-.01

-.01

-.01

-.01

-.01

-.01

-.02

-.02

-.02

-.02

-.02

-.02

-.03

-.03

-.04
1

10

-.03

-.04
1

Respons e of LINT to LPINV

10

Response of LINT to LDCP

10

Respons e of LINT to LEXTD

10

Response of LINT to LFD

10

Response of LINT to LGDP

.08

.08

.08

.08

.08

.04

.04

.04

.04

.04

.00

.00

.00

.00

.00

.00

-.04

-.04

-.04

-.04

-.04

-.04

10

10

10

10

10

10

10

10

Response of LINT to LINT

.04

-.04
1

.08

-.03

-.04
1

.01

-.03

-.04
1

10

Response of LGDP to LINT

.03

-.04

-.15
1

.04

-.03

-.10

-.15
1

Response of LFD to LINT

.10

-.15

-.08
1

.15

-.10

Response of LEXTD to LINT

.12

-.08

-.03
1

Response of LEXTD to LFD

.01

-.03
1

Response of LDCP to LINT

10

10

Figure 4.6: Accumulated impulse response functions for the private investment equation. The dashed
lines are 95% bootstrap confidence bounds. GDP and interest rate generates positive impulses. Fiscal
deficit initially generates positive impulses but start generating negative impulses at the end of the
fifth period.
15

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

Vi.
Summary and Conclusion
This study re-examines the effect of fiscal deficits on private investment in Nigeria between 19702006. After establishing the unit root status of the variables in the structural equation and the
existence of co-integration, the ordinary least squares (OLS) two stage approach as suggested by
Engle and Granger (1987) was utilized in deriving the estimates in the model adopted. The structural
analysis was done using impulse response analysis and variance decomposition to trace the one-time
shock to one of the innovations in the current and future values of the endogenous variables in the
model.

Evidence emerges that there is a positive long run relationship between private investment and real
growth of the national economy. This confirms the relevance of the accelerator principle to Nigeria,
with contemporaneous accelerator parameter of 1.84. On aggregate, a 1% improvement in national
income engenders 1.84% increase in private investment in Nigeria. In addition, the result indicates
that fiscal deficits have had a depressive effect on private investment in the country. The estimation
results suggest that a 1% increase in fiscal deficit leads to 0.267% decline in private investment. The
results also indicate that Nigerias debt profile has had strong and negative impact on private
investment in Nigeria. This confirms the relevance of the debt overhang hypothesis
Furthermore, although positive and insignificant estimates show that the elasticity of private
investment to domestic credit is 0.128. This implies that a 1% increase real credit to the private sector
results in a 0.128% increase in private investment. This underscores the need to significantly increase
credit to government in favour of the private sector. The error correction estimates of 56.1% of the
preceding periods disequilibrium are eliminated in the current period, with intermediate adjustments
captured by the difference terms. The value of the adjusted R 2 shows that the model accounts for at
least 83.8% changes in private investment.
When government budget deficits are financed through the banking system, it reduces the financial
resources in the hands of the banks to finance private sector investments. This will cause interest rates
to rise, thus making it difficult for private sector investors to borrow at the high interest rates which
will prevail and this crowds out private investments. However, economic growth of the nation can be
enhanced by directing public expenditure to public investments thus, creating enabling environment
for the private sector operators, and enhance economic growth. This will happen if government
adheres to fiscal discipline while spending on projects with positive net present value. This will
enhance capital formation so that the economy will return to long term growth path.
Again, more emphasis of government expenditure should be of on infrastructures that help in capital
formation instead of recurrent expenditures. The case in Nigeria is that over 75 per cent of the budget
goes to recurrent expenditures and thus a reduction in the physical infrastructure. Also the
government programmes should be financed through the bond markets as funds/savings can be
mobilized and channeled to specific projects.
REFERENCES
16

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

Albatel, AH (2004) Government Budget Deficits and the Crowding out of Private Sector Investment
in Sandi Arabia Journal king Saud University, Administrative Science Vol. 1 pp1-28.
Aleshina, A, Ardagna S. Perohi, R, and Schiantarell; F (1999). Fiscal Policy, Profits and Investment.
NBER Working Paper Series. Retrieved 26 February 2008 from http/ www.nberorg/papers/w7207.
Akpokodje (1998). Macroeconomic Environment, Investment Stimulation and Economic Growth and
Development. The Nigerian Experience cited in Paiko, I.I. (2012). Deficit Financing
and its Implication on Private Sector Investment: The Nigerian Experience, Arabian
Journal of Business and Management Review (OMAN Chapter), Vol. 1, No. 9, pp 4562.
Ariyo, A., and Raheen M.I. (1991). Effects of Fiscal Deficits on some Macroeconomic Aggregates
in Nigeria, cited in Tchokote (2004). Macroeconomics of Fiscal Deficts in Cameroun,
Doctoral Dissertation, University of Ibadan.
Aryanwu J.C (1997) Nigerian Public Finance.1st ed., Onitsha, Joanne Publishers.
Andrews, D.W.K (1991) Heteroskedasticity and Autocorrelation Consistent Covariance Matrix
Estimation, Econometrica Vol. 59, pp817-858
Barth, J.; Iden, G. and Russek, F. (1984). Do Federal Deficits Really Matter? Contemporary Policy
Issues, Vol. 3, pp 79-95.
Blejer, M.I and Khan, M.S. (1984). Government Policy and Private Investment in Developing
Countries. IMF Staff Paper vol 31, No 2, pp 379-403.
CBN (2004,2006) Statistical Bulletin (Lagos: Central Bank of Nigeria)
Dickey, D.A. and W.A. Fuller (1979). Distribution of the Estimators for Autoregressive Time Series
with a Unit Root, Journal of the American statistical association, Vol. 74, pp. 427-431.
Engsted, T. and Bentzen, J. (2001). Dynamic Modelling of Energy Demand: A Guided Tour through
the Jungle of Unit Roots and Cointegration. Energy, Vol.26, pp.45-55.
Ezeabasili, V.N. (2009) The Effect of Fiscal Deficits on Selected Macroeconomic Variables in
Nigeria. Doctoral Dissertation Department of Banking and Finance, Nnamdi Azikiwe
University, Awka, Nigeria.
Hoelscher, G. (1983). Federal Borrowing and Short Term Interest Rates. Southern Economic Journal,
vol. 50 pp. 319-333.

Islam, R. and Wetzel D. (1991). Macroeconomics of Public Sector Deficit, the case of Ghana.
Working Papers No. 672 (Washington D.C: The World Bank, Policy Research
Department).
Johansen, S. (1991). Estimation and Hypothesis Testing of Cointegration Vectors in Guassian Vector
Autoregressive Models, Econometrical, Vol. 59, pp.1551-1580.

17

International journal of Innovative Research in Management

ISSN 2319 6912

(March 2013, issue 2 volume 3)

Johansen. S (1988). A Statistical Analysis of Co-integration Vectors. Journal of Economic

Dynanmics & Control Vol. 12, pp231-254.
MacKinnon, J. G. (1996). Numerical distribution Functions for Unit Root and Co-integration Tests.
Journal of Applied Econometrics, Vol. pp. 601-618.
Makin, J. (1983). Real Interest, Money Supplies, Anticipated Inflation, and Fiscal Deficits. Review of
Economics and Statistics, vol. 65, pp 375-384
National Centre for Economic Management and Administration (2004) The macroeconomic Effects
of Deficit financing in Nigeria (NCEMA) workshop, Ibadan pp5-21
Newey, W. and K. West (1994). Automatic Lag Selection in Covariance Matrix Estimation. Review
of Economic Studies, Vol. 61, pp 631.
Njeru and Randa (1998). Macroeconomic Implication of fiscal Deficits: the case of Kenya, cited in
Tchokote, J (2004) The Macroeconomics of Fiscal Deficits in Cameroun. Doctoral
Dissertation, University of Ibadanm Ibadan.
Oussou k. and Bouabre B. (1993). The consequences and Limitation of Recent Fiscal Policy in
CotedIvoire. Research paper 51, AERC Nairobi Kenya, Cited in Tchokote, J (2004)
The Macroeconomics of Fiscal Deficits in Cameroun. Doctoral Disdsertation,
University of Ibadan, Ibadan.
Onwuoduokit, E.A (1999) Fiscal Deficits and Inflation Dynamics in Nigeria. An Empirical
Investigation of causal Relationship. CBN Economic and Financial Review Vol 37 No.
2 pp 1-16
Paiko, I.I. (2012). Deficit Financing and its Implication on Private Sector Investment: The Nigerian
Experience, Arabian Journal of Business and Management Review (OMAN Chapter),
Vol. 1, No. 9, pp 45-62.
Phillips, P.C.B. and Perron, P. (1988). Testing for a Unit Root in Time Series Regression. Biometrika,
Vol. 75, pp.335-346.
Sehmidt-Habbel, K and Muller T (1991) Private investment Under Macroeconomic Adjustment in
Morocco. Country Economic Department, working paper series 787, the World Bank,
Stum, J.E. and De Haan, J. (2001) Inflation in developed Countries: Does Central Bank
Independences Matter. New Evidence Based on New data set. Department of Economics
University of Groningen, the Nether Lands
Serven, L (1996) Does Public Capital Crowd our Private Capital? Evidence from India. Policy
Research Working Paper 1613. The World Bank Policy Research Department,
Macroeconomics and Growth Division.
Tchokote, J (2004) The Macroeconomics of Fiscal Deficits in Cameroun. Doctoral Dissertation,
University of Ibadan, Ibadan.

18