Download impact of oil shocks (revenue) on government revenue

IMPACT OF OIL SHOCKS (REVENUE) ON GOVERNMENT REVENUE AND
EXPENDITURE: THE NIGERIAN EVIDENCE.
By
Odeyemi Gbenga .A
Department of Economics
University of Lagos, Nigeria.
Email Address: [email protected]
Abstract
Nigeria’s dependability oil does not but respond to fluctuations in oil revenue which has
effect on government activities (revenues and expenditures) and major macroeconomic
variables such as consumer price index, money supply, capacity utilisation and economic
growth to mention a few. This paper investigated the impact of oil shocks (revenue) on
government revenue and expenditure in Nigeria. Empirical analysis was conducted by
applying vector autoregressive model (VAR) and vector error correction (VEC) with impulse
response and variance decomposition tools to the annual data on the Nigeria economy for
the period 1970-2012. The model was found to be stable and most of its estimates are as
expected. Moreover the findings indicate that government revenues, government
expenditures and money supply are important determinants of domestic price level in
Nigerian economy. The result does support the revenue-spending hypothesis for Nigeria. In
this context Nigeria should enhance the effectiveness of fiscal policy by making budget
expenditure less driven by revenue availability. This policy can help to avoid the costs and
instability that variations in public spending generate mostly due to the fluctuations in oil
revenues.
Keyword: Nigeria, oil revenue, government revenue and expenditure, VAR, VEC.
IMPACT OF OIL SHOCKS (REVENUE) ON GOVERNMENT REVENUE AND
EXPENDITURE: THE NIGERIAN EVIDENCE
Introduction
The Nigerian economy with her huge reliance on oil cannot be overemphasized as this
sector has been a major source of her income, the unique role revenues from oil plays in the
structure of government budgets and expenditures is a special characteristic of the
developing oil export economies like Nigeria. Oil revenues are the main source of financing
government expenditures and imports of products. The oil and gas sector accounts for
about 35 per cent of gross domestic product, and petroleum exports revenue accounts for
about 70 per cent of total exports revenue.
The budget is especially affected by sudden negative or positive shocks in oil prices.
Economic performance has been affected by oil revenue volatility and “stop-go” policies,
resulting in boom and bust cycles. Transitory oil price increases have caused the spending to
increase, often maintained even after oil revenues had decreased again. Despite higher oil
prices and revenues in recent years, the Nigeria government budget deficits are still a
challenging issue, in part due to the huge amount of state subsidies on energy and
comestible goods.
Moreover, the most recent challenge for Nigerian economy consists of oil spillage and theft
which mostly has effect both on oil export and oil revenue, however, the revenue effect
seems to be our concern here. It is likely that these challenges will reduce the amount of oil
revenues and consequently engenders budget deficit. This shows a dilemma to the policy
makers who are trying to keep up the momentum into the economy by injecting more
government expenditure into domestic economy while at the same time the oil revenues
are uncertain and expected to decrease (compare Van Marrewijk and Van Bergeijk 1993).
If policymakers understand the relationship between government expenditure and
government revenue, continuous government deficits can be prevented. In the past
decades, the relationship between government expenditure and government revenue has
attracted significant interest. Indeed this is because; the relationship between government
revenue and expenditure has an impact on the budget deficit (Eita & Mbazima, 2008).
However, a point to note is the submitted opinions as reasons why the nature of the
relationship between government expenditure and government revenue is important. First,
if the revenue-spend hypothesis- which indicates that the causality runs from government
revenue to government expenditure- holds, budget deficits can be avoided by implementing
policies that stimulate government revenue. Second, if the spend-revenue hypothesiswhich states that the causality runs from government expenditure to government revenue
holds it suggests that the government spends first and pay for this spending later by raising
taxes. This will result in the fear of paying more taxes in the future and encourage the
outflow of capital. Third, if the fiscal synchronization hypothesis- which states that the
causality runs from both directions- does not hold, it suggests that government revenue
decisions are made independent from government expenditure decisions. This can cause
high budget deficits should government expenditure rise faster than government revenue.
Oil revenues in most oil-exporting countries are paid directly to the government as the
guardian of the natural resources. Hence, the government becomes the conduit for the oil
revenues into the economy. If the revenue is unstable and/or transitory, then the other
macroeconomic variables will be unstable and the natural resource blessing could become a
curse. Fiscal policy is therefore a key element, for most countries, in causing or preventing
the resource curse (Devlin and Lewin, 2004).
Levin and Loungani (1996) argue that a country’s response to oil prices is determined by the
choices of rules adopted by the country and those followed by its trading partners.
Berument and Ceylan (2005) state that the impact of oil price changes depends on the
structure of the economy and whether the country is a net oil exporter or importer; the net
exporters of oil should benefit from the windfall profits and fiscal revenues created by oil
price hikes, while the net importers of oil will experience this situation as additional burdens
on their economies. However, Abeysinghe (2001) maintains that even net oil exporters
cannot escape the negative influence of high oil prices. He states that while the direct
impact of high oil prices on net oil exporters is positive, a negative effect is transmitted
indirectly through a trade matrix. Hence, net exporters cannot escape the contraction effect
which is passed on through their trading partners and in the long run, the positive effects of
high oil prices is mitigated.
From the political view and regarding the current challenges of the Nigerian economy for
passing the negative effects of the oil theft on its oil exports and foreign direct investment,
bearing in mind how so much the Nigerian economy rely on oil. And price also to determine
oil revenues at a point in time, influence the government expenditures and thereby the
wage bills of government employees, interest payment, employer contribution including
social security and pensions, subsidies and all other payments which are related to the
government functions.
The causal relationship between government revenue and expenditure has remained an
empirically debatable issue in the field of public finance. Many studies have considered links
between government expenditures and government revenues but most have been
conducted for countries where oil is not a major concern. This paper investigates the
dynamic interrelationship between the government revenues and government expenditures
in Nigeria as a developing oil export based economy. This study wants to take a different
approach to the fiscal policy issue as it deals with the government revenues and
government expenditures for a petroleum economy where income, government revenue
and exports are linked with oil revenue. It seems that it will be more interesting if we
consider the role of oil shocks on the government expenditures and government revenues
link when we are going to investigate their relationship in an oil exporting country like
Nigeria. Moreover in this study, the question of how oil price (revenue) shocks are affecting
the other major macroeconomic variables like money supply and domestic price levels
through influencing the government revenues and government expenditures is examined
answered.
Most of the empirical studies carried out have focused on the oil importing economies,
particularly the developed economies. This study also is considered in the study.
Consequently, the purpose of this study is twofold. First, it investigates the government
expenditure-government revenue nexus, examining which of the three hypotheses, spendand revenue, revenue-and spend or fiscal synchronization, is applicable to Nigeria. as
against previous works where bivariate framework is the usual approach; rather, given the
characteristics of the Nigerian economy, consideration is given to how government
expenditures and government revenues respond to an oil price (revenue) shock.
Furthermore, while most of the existing researches in this area use only total government
expenditures, this study will employ disaggregate government expenditures (current
expenditures and capital expenditures) in addition to total expenditures. Second, I want to
investigate how and to what extend the government expenditures and government
revenues nexus can transfer the effects of oil shocks to the other Nigerian key
macroeconomic variables such as money supply and CPI.
Section two of this paper reviews previous works on the relationship between government
expenditure and government revenue alongside the effects of oil shocks in the country
specific economy. Section three introduces the methodology used in this paper, section four
discusses results and the findings made while section five consist of summary and
conclusions.
2.0 Literature Review
Oil price (revenue) shocks receive considerable attention for their presumed
macroeconomic consequences. As the main focus of research directed towards net oil
importers and developed economies, there is a dearth of such studies for developing, oilexporting countries. Generally, there appears to be little consensus in the current literature
on the macroeconomic impact of oil fluctuations.
Darby (1982) and Hamilton (1983) focused on the US economy. While Darby was not able to
identify a significant relationship between oil prices and macroeconomic variables, Hamilton
found that oil price shocks were an important factor in almost all US recessions from 1949
to 1973. He concludes that changes in oil prices Granger-caused changes in unemployment
and GNP in the US economy.
Taking an analytical view of the impact of oil on the Nigerian economy, Obadan (1983),
sought to evaluate the impact of the development in the oil sector on the Nigerian economy
through government finances, he found that effect of oil on government revenue is positive.
That is, there is a positive relationship between oil price and government expenditure,
claiming that this relationship is significant and have fiscal implications and linkages. And
these linkages arise from the use of increasing oil revenue by the government to develop
other sectors of the economy such as Agriculture, Education, and Infrastructures etc which
are components of various government capital and recurrent expenditures.
Ubogu (1984) sought to establish the impact of oil on the Nigerian economy. Examining the
growth and development of the oil industry, government participation and the stages of the
development of the industry, government revenue, foreign exchange earnings, employment
generation and industry linkages effects, he noted that oil has been responsible for the
radical increase in revenue and further buttressed the stronger dependence on oil revenue
as envisaged in our development plans due to the unanticipated decline in oil earnings. He
was however, strongly in support of diversification and the need for judicious use of the
current limited revenue.
Hess (2000) observes that oil price shocks led to lower real gross domestic product (GDP)
prior to the 1980s. Since that time, changes in oil prices had no effect on US economic
activity. He concludes that oil price spikes are generally short-lived and may not even have a
direct effect on US economic activity. Furthermore, He notes that oil prices became more
volatile in the 1980s and 1990s relative to real GDP. However, Blanchard and Gali (2007)
claim the main reasons behind the weak response of economies to oil shocks in recent years
are smaller energy intensity, a more flexible labour market, and improvements in monetary
policies.
Eltony and Al-Awadi (2001), use quarterly data for the period 1984-199 to examine the
impact of oil price fluctuations on key macroeconomic variables for the Kuwaiti economy.
The results of the vector autoregression (VAR) and vector error correction (VEC) models
indicate that oil price shocks and hence oil revenues have a notable impact on government
expenditure, both development and current. However, government development
expenditure has been influenced relatively more. The results also point out the significance
of the CPI in explaining a notable part of the variations of both types of government
expenditure. On the other hand, the variations in value of imports are mostly due to the oil
revenue fluctuations and also the fluctuation in government development expenditures.
The results from the VECM approach indicate that a significant part of LM2 variance is
explained by the variance in oil revenue.
Jimenez-Rodriguez and Sanchez (2004) assess empirically the effects of oil price shocks on
the real economic activities in a sample of seven OECD countries, Norway and the Euro area
as a whole. Their results show that oil price increases have a larger impact on GDP growth
than oil price declines.
They emphasized the difference between oil importing and oil exporting countries. They
conclude that oil price increases have a negative impact on economic activity for oilimporting countries, while the relationship for oil exporting countries is mixed. UK’s
economy, according to them exhibits a surprising behaviour: while it is expected that an oil
price shock has positive effects on the GDP growth for a net oil exporting country, an oil
price increase of 100% actually leads to a loss of British GDP growth rate of more than 1%
after the first year in all specifications.
The transmission mechanisms of oil shocks to the economy mostly in the advanced, oil
importing countries are the supply effect, the demand effect and the terms of trade effect
(Brown et al., 2004; Schneider, 2004; Lardic and Mignon, 2006; Sill, 2007; Jbir and ZouariGhorbel, 2009). On the supply side, increased oil prices result in a reduction of an input for
production and this leads to higher production costs, thus leading to a slowdown of output
and productivity. On the demand side, higher oil prices increase the general level of prices
and with a reduction in real income available for consumption, demand falls (Farzanegan
and Markwardt, 2009). On the terms of trade side, oil importing countries face worsening
terms of trade conditions as demand falls in these countries and this results in wealth
transfer from oil-importing to oil exporting countries (Iwayemi and Fowowe, 2010).
Olomola and Adejumo (2006), examined the effect of oil price shock on output, inflation,
the real exchange rate and the money supply in Nigeria using quarterly data from 1970 to
2003. Using a VAR model they indicate that oil price shocks do not have substantial effects
on output and inflation rate in Nigeria. However, oil price shocks do significantly influence
the real exchange rates. The implication is that a high real oil price may give rise to wealth
effect that appreciates the real exchange rate. This may squeeze the tradable sector, giving
rise to the “Dutch Disease ”.They conclude that oil price shock is an important determinant
of real exchange rates and in the long run money supply, while money supply rather than oil
price shocks that affects output growth in Nigeria.
Mehrara and Oskoui (2007), study the sources of macroeconomic fluctuations in four oilexporting countries using annual data: (Iran, 1970–2002; Indonesia, 1970–2002; Kuwait,
1972–2002; Saudi Arabia, 1971–2002).using a structural VAR approach and by imposing
long-run restrictions on a VAR model, four structural shocks are identified: nominal demand,
real demand, supply, and oil price shocks. Their results show that oil price shocks are the
main source of output fluctuations in Saudi Arabia and Iran, but not in Kuwait and
Indonesia. They claim this is because of the relatively successful experience of Kuwait in the
use of stabilization and savings fund and the right structural reforms particularly diversifying
away from resource-based production in Indonesia.
Farzanegan and Markwardt (2009), analyse the dynamic relationship between oil price
shocks and major macroeconomic variables in Iran by applying a VAR approach. Their full
sample comprises quarterly observations for the 1975:II-2006:IV period. Their main results
examine the post-war period (1989:I–2006:IV) and for the robustness check, they compare
these results with the pre-1989 period (1975:II–1988:IV). They analyse the effects of oil
price shocks in three different channels: the supply side, the demand side, and the terms of
trade. Their results on the supply side of economy reveal that positive oil price shocks
stimulate Nigerian industrial production and real imports. On the other hand, negative
shocks on oil prices undermine the process of real industrial production and play a
significant role in lowering the real level of imports. On the demand side, both positive and
negative oil price shocks have inflationary effects and drive up the general level of prices.
Their results show just a marginal impact of oil price fluctuations on real government
expenditures. They conclude that increasing oil prices improves the terms of trade and
appreciates the real effective exchange rate. The imprecision in previous studies is however
evident, hence the need for this study.
3.0 Methodology
The vector autoregression model (VAR), vector error correction model (VECM) is employed
for the purposes of this study more so; some useful tools on these techniques such as
impulse response functions and variance decomposition are used.
The linear dynamic vector autoregression (VAR) method has presented by Sims (1980).
Sometimes, economic theory may not be adequate to determine the specific relationship
between variables. According to Pindyck and Rubinfeld (1991), there are times when it is
more logical to allow the data to specify the dynamics in a relationship. VAR makes little or
no theoretical demands on the structure of the relationships in a model. VAR helps
researchers to understand interrelationships among economic variables (Enders, 1996).
Maddala (1992) notes that the VAR model is a critical starting point in the analysis of
interrelationships among different time series. Darnell and Evans (1990) observe that the
VAR model provides a straightforward method of producing forecasts that do not discern on
how the variables in the model affect one another. The mathematical representation of a
VAR is
𝑦𝑡 = 𝐴1 𝑦𝑡−1 + ⋯ + 𝐴𝑝 𝑦𝑡−𝑝 + 𝐵𝑥𝑡 + 𝜀𝑡
(1)
Where 𝑦𝑡 is a k vector of endogenous variables, 𝑥𝑡 is a d vector of exogenous variables,
𝐴1 ,..., 𝐴𝑝 and B are matrices of coefficients to be estimated, and 𝜀𝑡 is a vector of innovations
that may be contemporaneously correlated but are uncorrelated with their own lagged
values and uncorrelated with all of the right-hand side variables. Since only lagged values of
the endogenous variables appear on the right-hand side of the equations, simultaneity is
not an issue and OLS yields consistent estimates. Moreover, even though the innovations 𝜀𝑡
may be contemporaneously correlated, OLS is efficient and equivalent to GLS since all
equations have identical regressors. In the standard VAR, disturbances are generally
characterized by contemporaneous correlations which it causes the response of the system
to an innovation in one of the variables be the response of all those variables that are
contemporaneously correlated with it. However, this contemporaneous correlation is
purged by Cholesky orthogonal procedure. Unrestricted VAR models are sometimes
superior to the structural VAR models since the latter models are “very often misspecified”
(Tijerina- Guajardo and Pagán, 2003). The VAR model is a dynamic simultaneous equation
system which is free of a prior restriction on the structure of the model. The VECM is
basically a VAR system that builds on Johansen's test for cointegration and is usually
referred to in the literatures as the restricted VAR. Indeed a vector error correction model
(VEC) is a restricted VAR designed for use with non-stationary series that are known to be
cointegrated. The VEC has cointegration relations built into the specification so that it
restricts the longrun behavior of the endogenous variables to converge to their
cointegrating relationships while allowing for short-run adjustment dynamics. The main
applied tools in the VAR and SVAR models estimation are the impulse response functions
(IRFs) and the variance decomposition analyses (VDC). An examination of the entire system
is studied by analyzing the variance decomposition of the system. Variance decomposition
assigns the variance of forecast errors in a given variable to self-shocks, as well as those of
the other variables in the VAR (Bron and Yücel, 1999). The Choleski decomposition method
is adopted, in order to construct the variance decompositions.
For the most part, the Cholesky procedure implicitly assumes recursivity in the VAR model
as it is estimated. Although theoretical considerations may help in determining the order of
the variables in the VAR model and ex-post sensitivity analysis may further help provide
insights regarding appropriate ordering, it remains largely at the discretion of the modeller
(Eltony and Al- Alwadi, 2001). To determine the appropriate number of lag length of the
VAR model, following Judge at al. (1988) and McMillin(1988), but for the purpose of this
paper, Akaike Information Criterion (AIC) is chosen.
3.1 Data Discussion
Six variables are considered for this study and they are discussed below; one external
shock captured by oil revenue (oilr), three policy variables; government recurrent and
capital expenditure, (cap) and (rec) and money supply (m2) and two macroeconomic
variable government total revenue (ttr) and consumer price index (cpi).the data’s were
sourced from world development indicator (WDI) and the statistical bulletin of the central
bank of Nigeria. However, in other to have all these variables on the same platform, they
are all converted and used at their logarithm form.
The first variable in the Cholesky ordering is usually the variable with the largest expectation
or the most evidence for exogeneity among the variables of the VAR system. This ordering,
indicates that oil revenues have an influence on total revenues and then later on all other
variables in the model. While the oil revenues are deeply depended on oil prices, then its
behaviour is the least determined by other variables included in the model. Oil prices and
consequently oil revenues are largely determined by the world market conditions rather
than within the a country’s economy, Also, this ordering reflects that government
expenditures are largely determined by government revenues, which is a plausible
assumption with regard to characteristics of Nigeria as a developing oil export based
economy.
Following theoretical considerations, one may say that oil revenue should come before
government revenue as it is the price of oil per time that determines oil revenue which is
totally exogenous in the model. One may also say that oil revenue would determine
government revenue especially in a nation like Nigeria. That current expenditure comes
before capital expenditure is understandable as capital expenditure is rather more rigid than
government expenditure. Capital expenditure can be revised as often as possible but
recurrent expenditure cannot be so done to.to buttress this Farzanegan(2011), claims that
capital expenditures are sensitive to fluctuation of oil revenues.
In this same light.one may also assay that money supply is determined by government
expenditures and money supply and government expenditures and revenues are
determinant factors for the domestic price level (CPI).
4.0 Results
It is customary to understand the nature of the data that is to be used especially on case of
time series data, their stationarity (Unit Root Test) must be tested for and a number of
methods are available to decipher the nature of the data capturing individual variable.
4.1 Unit Root Tests
Table 1
Variable
ADF
level
loilr
-2.020380
lttr
-1.452317
lcap
-1.669909
lrec
-0.765120
Lm2
-0.493306
lcpi
-0.616157
1st
difference
7.087045***
7.008805***
7.209285***
7.483030***
4.090720***
3.241058***
Remark
PHILLIPS
PERRON
Level
I(1)
-2.131927
I(1)
-1.500028
I(1)
-1.698495
I(1)
-0.745609
I(1)
-0.114773
I(1)
-0.474761
1st
difference
7.087045***
7.011139***
7.162340***
7.609985***
4.108570***
3.073775***
Remark
I(1)
I(1)
I(1)
I(1)
I(1)
I(1)
Composed by the author.
The null hypothesis is rejected at 1%.
From the above table, it is safe to say that all the variable are stationary at 1 st difference
using ADF and Phillips Perron unit root test.
The next step is to check for the existence of a long run relationship between the variables
using a cointegration test but before this is done, it important to check for the stability of
the model, the Figure shows the AR graph which reports the inverse roots of the
characteristic AR polynomial (see Lütkepohl, 1991). This figure shows that in the VAR model
all roots have modulus less than one and lie inside the unit circle and the VAR model is
stable
Figure 1
Inverse Roots of AR Characteristic Polynomial
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
Having confirmed that the model is good and stable, the test for a long run relationship is
next and this result is presented below.
4.2 impulse response
The impulse response functions trace out the response of current and future values of the
variables to a one standard deviation increase in the current value of oil revenues errors.
Following Sims and Zha (1999), 68% confidence intervals for the IRFs are estimated in this
study. To build these confidence bands, 1000 Monte Carlos simulations are employed.
The impulse response of each variable is shown below. It is evident from figure 1 that oil
revenues have positive and statistically significant effect on itself, government total
revenue, capital and recurrent expenditure especially in the long run but becomes
insignificant in the long run, however money supply response to oil revenue is all the way
statistically significant while totally insignificant for consumer price index all the way .it is
also seen that total government revenue follows oil revenue which implies that government
revenue is so sensitive to oil revenue and responds strongly to fluctuations in oil revenue
and same is it with both capital and recurrent expenditure. This is more pronounced
following the response of money supply which relies majorly on government expenditure to
innovations in oil revenues.one can tell from here that oil revenue in Nigeria practically
affect and dictates economic activities.
Figure 3 makes it clear that capital and recurrent expenditure, government revenue, money
supply, oil revenue and has a positive significant effect on capital expenditure in the short
run but fades out gradually while consumer price index is not significant.
Figure 4 shows that an innovation in consumer price index will have a significant and
positive effect on all other variables though their level of significance varies as government
revenue gets more significant in the long run. This shows that causality runs from consumer
price index to all other variable.
Figure 5 indicates that an innovation in money supply enjoys a positive and significant effect
on all other variable in the model.
Figure 5 shows a positive and significant relationship of all variables to government
recurrent expenditure excerpt capital expenditure showing a negative relationship .
Response to Cholesky One S.D. Innovations ± 2 S.E.
Res pons e of LCAP to LOILR
Res pons e of LCPI to LOILR
.08
.04
.04
.00
.00
-.04
-.04
-.08
-.12
-.08
1
2
3
4
5
6
7
8
9
1
10
2
3
Res pons e of LM2 to LOILR
4
5
6
7
8
9
10
9
10
9
10
Res pons e of LOILR to LOILR
.08
.15
.10
.04
.05
.00
.00
-.05
-.04
-.10
1
2
3
4
5
6
7
8
9
1
10
2
3
Res pons e of LREC to LOILR
4
5
6
7
8
Res pons e of LTTR to LOILR
.08
.15
.10
.04
.05
.00
.00
-.04
-.05
-.08
-.10
1
2
3
4
5
6
7
8
9
1
10
2
3
4
5
6
7
8
Figure 3
Response to Cholesky One S.D. Innovations ± 2 S.E.
Res pons e of LCAP to LCAP
Res pons e of LCPI to LCAP
.08
.12
.04
.08
.04
.00
.00
-.04
-.04
-.08
-.08
-.12
-.12
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LM2 to LCAP
3
4
5
6
7
8
9
10
9
10
9
10
Res pons e of LOILR to LCAP
.08
.10
.05
.04
.00
.00
-.05
-.04
-.10
-.08
-.15
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LREC to LCAP
3
4
5
6
7
8
Res pons e of LTTR to LCAP
.08
.10
.04
.05
.00
.00
-.04
-.05
-.08
-.10
-.12
-.15
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
Figure 4
Response to Cholesky One S.D. Innovations ± 2 S.E.
Res pons e of LCAP to LCPI
Res pons e of LCPI to LCPI
.25
.20
.20
.15
.15
.10
.10
.05
.05
.00
.00
-.05
-.05
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LM2 to LCPI
3
4
5
6
7
8
9
10
9
10
9
10
9
10
9
10
9
10
Res pons e of LOILR to LCPI
.16
.3
.12
.2
.08
.1
.04
.0
.00
-.04
-.1
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LREC to LCPI
.25
.20
.20
.15
.15
.10
.10
.05
.05
.00
.00
-.05
-.05
2
3
4
5
Figure 5
6
7
8
4
5
6
7
8
Res pons e of LTTR to LCPI
.25
1
3
9
1
10
2
3
4
5
6
7
8
Response to Cholesky One S.D. Innovations ± 2 S.E.
Res pons e of LCAP to LM2
Res pons e of LCPI to LM2
.16
.12
.12
.08
.08
.04
.04
.00
.00
-.04
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LM2 to LM2
3
4
5
6
7
8
Res pons e of LOILR to LM2
.12
.16
.10
.12
.08
.08
.06
.04
.04
.00
.02
.00
-.04
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LREC to LM2
.16
.12
.12
.08
.08
.04
.04
.00
.00
-.04
-.04
2
3
4
5
6
7
8
4
5
6
7
8
Res pons e of LTTR to LM2
.16
1
3
9
10
1
2
3
4
5
6
7
8
Figure 6
Response to Cholesky One S.D. Innovations ± 2 S.E.
Res pons e of LCAP to LREC
Res pons e of LCPI to LREC
.12
.12
.08
.08
.04
.00
.04
-.04
.00
-.08
-.12
-.04
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LM2 to LREC
3
4
5
6
7
8
9
10
9
10
9
10
9
10
9
10
9
10
Res pons e of LOILR to LREC
.12
.16
.12
.08
.08
.04
.04
.00
.00
-.04
-.04
-.08
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LREC to LREC
3
4
5
6
7
8
Res pons e of LTTR to LREC
.16
.12
.12
.08
.08
.04
.04
.00
.00
-.04
-.04
1
2
3
4
5
6
7
8
9
1
10
2
3
4
5
6
7
8
Figure 7
Response to Cholesky One S.D. Innovations ± 2 S.E.
Res pons e of LCAP to LTTR
Res pons e of LCPI to LTTR
.04
.02
.02
.00
.00
-.02
-.02
-.04
-.04
-.06
-.08
-.06
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LM2 to LTTR
3
4
5
6
7
8
Res pons e of LOILR to LTTR
.02
.04
.01
.00
.00
-.01
-.02
-.04
-.03
-.04
-.08
-.05
1
2
3
4
5
6
7
8
9
1
10
2
Res pons e of LREC to LTTR
3
4
5
6
7
8
Res pons e of LTTR to LTTR
.04
.04
.02
.02
.00
.00
-.02
-.02
-.04
-.04
-.06
-.06
-.08
-.08
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
From the last figue we can tell from the negative impact that a change in government
expenditure has on the all the variables that governmant revenue in nigeria depends
heavilly on other factors and that it doesnot just exist by itself. This can be justified from
figure 2 where it showed cleary that government revenue relies on revenue from oil to
functions and also that government expenditure depends on the revenue of government.
4.3 Cointegration test
Another bone of contention lies with whether an unrestricted VAR should be used where
the variables in the VAR are cointegrated. There is a body of literature that supports the use
of a vector error correction model (VECM), or cointegrating VAR, in this situation. It is shown
through Monte Carlo simulations that short run forecasting performance of unrestricted
VAR models are better than VECM performance (Farzanegan and Markwardt, 2009). Vector
error correction (VEC) model is a restricted version of the VAR model. If there is
cointegration, imposing this restriction will yield more efficient estimates (Naka and Tufte,
1997). However as the robustness of my results for the second group of the variables in this
study, I will try to use the VEC model and compare its results with the results of the
unrestricted VAR model.
Since the VEC specification only applies to cointegrated series, I should first run the
Johansen cointegration test to determine the number of cointegrating relations. A vector of
variables integrated of order one can be cointegrated if there exists linear combination of
the variables, which are stationary. Following the approach of Johansen and Juselius (1990)
two likelihood ratio test statistics, the maximal eigenvalue and the trace statistic, were
utilized to determine the number of cointegrating vectors. The results of the maximal
eigenvalues and trace test statistics are presented below:
Table 2
Unrestricted Cointegration Rank Test (Trace)
Hypothesized
No. of CE(s) Eigenvalue
Trace
Statistic
0.05
Critical Value Prob.**
None *
At most 1 *
At most 2 *
At most 3 *
At most 4 *
At most 5
286.2561
185.0319
102.9450
54.15531
23.73905
2.075519
95.75366
69.81889
47.85613
29.79707
15.49471
3.841466
0.930317
0.884696
0.723056
0.550863
0.434527
0.053154
0.0000
0.0000
0.0000
0.0000
0.0023
0.1497
Trace test indicates 5 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Having confirmed also the existence of cointegration among the variables, one can confirm
a long run relationship between the variables and this is a justification for the use of VEC. A
quick step to the variance decomposition is taken to see the contribution of each variable to
the other over a period of time. The result of variance decomposition is hereby presented
below. A 10 period season is chosen for the purpose of this study.
Table 3: Vector error correction (VEC) estimates variance decomposition
V.D of
LCAP
Period
S.E.
LCAP
LCPI
LM2
LOILR
LREC
LTTR
1
2
3
4
5
10
0.112044
0.158294
0.180128
0.200592
0.224335
0.344193
100.0000
54.24458
44.35585
38.04448
30.51051
16.69033
0.000000
3.993531
14.43238
24.65789
32.87267
50.00804
0.000000
23.42570
21.95556
17.98431
14.74652
6.552010
0.000000
0.779785
3.439479
3.908355
3.800324
5.274591
0.000000
16.13198
12.58473
10.51795
13.00152
17.03639
0.000000
1.424425
3.231997
4.887019
5.068453
4.438646
V.D of
LCPI
Period
S.E.
LCAP
LCPI
LM2
LOILR
LREC
LTTR
1
2
3
4
5
10
0.041274
0.085563
0.127930
0.161330
0.184285
0.275001
3.960284
3.394260
1.712681
1.216629
0.960531
2.726794
96.03972
89.05617
84.14183
81.85751
79.51635
73.82969
0.000000
3.483653
3.272254
2.725262
2.614088
1.534439
0.000000
1.854700
3.015863
4.432724
5.392110
7.266075
0.000000
1.281015
5.535962
6.811835
7.779570
10.24797
0.000000
0.930205
2.321405
2.956036
3.737351
4.395035
V.D of
LM2:
Period
S.E.
LCAP
LCPI
LM2
LOILR
LREC
LTTR
1
2
3
4
5
10
0.046497
0.073391
0.087662
0.096234
0.102947
0.181693
1.726797
2.507863
9.111395
10.16608
8.885487
8.897436
0.391246
1.016917
0.828821
1.541808
1.572860
28.03527
97.88196
94.61425
86.81088
83.35627
77.94247
26.50861
0.000000
0.117518
0.866321
0.906163
1.248929
0.617300
0.000000
0.157163
0.126529
0.571651
5.891680
32.06088
0.000000
1.586291
2.256058
3.458023
4.458570
3.880504
V.D of
LOILR:
Period
S.E.
LCAP
LCPI
LM2
LOILR
LREC
LTTR
1
2
3
4
5
10
0.139634
0.153331
0.176708
0.197601
0.228356
0.390218
1.024235
2.477618
5.197655
4.211429
4.640454
6.975991
6.702378
11.58883
10.31058
11.92876
21.72856
53.32167
27.85187
23.76389
19.87210
15.98527
11.97731
4.999320
64.42151
57.18297
43.13640
36.00597
26.98787
12.14579
0.000000
1.838208
17.95516
28.33818
31.27104
19.54261
0.000000
3.148483
3.528101
3.530395
3.394772
3.014627
V.D of
LREC:
Period
S.E.
LCAP
LCPI
LM2
LOILR
LREC
LTTR
1
2
3
4
5
10
0.094229
0.116722
0.130531
0.138970
0.156431
0.280046
5.309475
3.883150
3.221913
3.081305
5.993853
7.275603
1.013779
15.97828
15.98254
17.69644
23.76177
54.46793
3.469562
12.72166
11.97607
10.67099
8.943650
2.984752
11.04223
7.998492
7.068453
6.610684
5.222476
4.571537
79.16495
59.08169
60.17632
58.08160
52.11535
27.53701
0.000000
0.336729
1.574703
3.858981
3.962905
3.163163
V.D of
LTTR:
Period
S.E.
LCAP
LCPI
LM2
LOILR
LREC
LTTR
1
2
3
4
5
10
0.117787
0.128796
0.154570
0.176446
0.205790
0.369582
1.007489
2.725005
4.011261
3.188901
3.531172
6.159327
6.496345
10.20155
9.999425
15.67197
25.66864
56.63874
23.23209
20.61003
15.70461
12.10802
8.902168
3.543761
66.07634
59.80190
41.85620
33.08365
24.38640
10.74546
0.017737
1.586045
23.17683
31.33942
33.17272
19.47923
3.169995
5.075476
5.251682
4.608042
4.338893
3.433489
The variance decomposition results corresponding to the estimated VECM are presented in
table 3
The estimates of variations for government expenditures and government revenues in the
VEC model in compare to VAR indicate that oil revenues have a more leading role in
explaining the fiscal policy stance as it is seen from the above that government total
revenue so much relies on oil revenue. It contributes about 66% to it and it is the level of
government revenue either actual or projected that determines the level of government
expenditure which in turn determines money supply which also determines consumer price
index as in each case, each represent the main contributor after individual variable. All these
submission are made from the above table.
5.0 Summary and Conclusions
Taking into account the heavy dependency of government expenditures on oil revenues in
Nigerian economy as a developing oil export based economy. This paper considers the
effect of oil shocks on the dynamic relationship between government expenditures and
government revenues. Using unrestricted VAR and VEC model, A group of the variables
(including oil revenues, government total revenues, government current and capital
expenditures, money supply and CPI) with respect to their special characteristics employing
annual data for the period 1970-2012.
The results of the VAR model from the impulse repose graph shows that the variations in
government total revenues and government current expenditures are sensitive to the
shocks to oil revenues. There also exists a unidirectional causality from government revenue
to government expenditure especially on the capital expenditure side.
It is also worthy of note that with VEC, oil revenue shocks can affect the other
macroeconomic variables more directly and strongly while in the VAR model this influencing
has changed to be more via total revenues channel. An unexpected result is that the
variations in government current expenditures caused by the shocks to oil revenues and
total revenues are stronger than the variations in government capital expenditures caused
by the mentioned resources. This indicates that oil revenue fluctuations may affect the
standard of life of the Nigerian populace which are mostly the government employees. As
another result of this study, both VAR and VEC models confirm the results of the previous
studies indicating that money supply, government revenues and government expenditures
can be important determinants in explaining the Nigerian domestic prices.
These results suggest that fiscal policy appears to be effective in Nigeria as the oil shocks
impact government expenditure and then government expenditure accounts for a relatively
considerable part of the CPI and money supply variations.
Generally speaking, the results support the revenue-spending hypothesis for Nigeria. In this
context Nigeria should enhance the effectiveness of fiscal policy by making budget
expenditure less driven by revenue availability. This policy can help to avoid the costs and
instability that variations in public spending generate mostly due to the fluctuations in oil
revenues. The expenditures can be planed and isolated from volatile short-term revenue
availability by using a medium term expenditure framework. Moreover the Nigerian
government should reduce the dependence of its expenditures to oil revenues by financing
these costs through non-oil sources such as taxes.
Reference:
Abeysinghe, T. (2001) “Estimation of direct and indirect impact of oil price on growth”,
Economics letters, 73, pp. 147-53.
Blanchard, O.J., Gali, J. (2010) The macroeconomic effects of oil price shocks: why are the
2000s so different from 1970s? In: Gali, J., Gertler, M. (Eds.) International Dimensions of
Monetary Policy. University of Chicago Press, Chicago, IL, pp.373 - 428
Darby, M.R. (1982) ‘The price of oil and world inflation and recessions’, American Economic
Review 72, 738–751.
Eita, Joel Hinaunye and Mbazima, Daisy (2008) ‘The Causal Relationship between
Government Revenue and Expenditure in Namibia’, MPRA Paper, No. 9154.
Eltony, M.N., Al-Awadi, M. (2001) ‘Oil price fluctuations and their impact on the
macroeconomic variables of Kuwait: a case study using a VAR model’, International Journal
of Energy Research, 25, 939-959.
Farzanegan, M.R. and Markwardt, G. (2009) ‘The effects of oil price shocks on the Nigerian
Economy’, Energy Economics 31 (2009), 134-151
Farzanegan, M.R. (2011) ‘Oil revenue shocks and government spending behaviour in Iran’,
Energy Economics, 33(2011), 1055-1069.
Hamilton, J.D. (1996) ‘This is what happend to the oil price-macroeconomy relationship’,
Journal of Monetary Economics 38, 215–220.
Iwayemi, A., Fowowe, B. (2011) ‘Impacts of oil price shocks on selected macroeconomic
variables in Nigeria’, Energy policy 39, 603-612.
.
Jbir, R., Zouari-Ghorbel, S. (2009) ‘Recent oil price shock and Tunisian economy’, Energy
Policy 37, 1041-1051.
Jimenez-Rodriguez, R., Sanchez, R.H. (2005) ‘Oil price shocks and real GDP growth: empirical
evidence for some OECD countries’, Applied Economics 37, 201-228.
Lütkepohl, H. (1991) Introduction to Multiple Time Series Analysis. Springer-Verlag, New
York.
Marrewijk, C. van and Bergeijk, P.A.G. van, (1993) ‘Endogenous Trade Uncertainty: Why
Countries May Specialize against Comparative Advantage’, Regional Science & Urban
Economics, 23 (3), pp. 681-94
Obadan,. (1983), ‘Prospects for Diversification in Nigeria’s Export Trade’ in Annual
Conference of the Nigerian Economic Society, Heinemann Press,Ibadan page 33-53,
unpublished e mass: press.
Olomola, P.A., Adejumo, A.V. (2006) ‘Oil price shock and macroeconomic activities in
Nigeria’, International Research Journal of Finance and Economics 3, 28-34.
Sims, C.A. (1980) ‘Macroeconomics and reality’, Econometrica 48, 1-48.
Ubogu A.E (2006) ’’ An overview oil pollution and heavy metal concentration in Warri area,
Nigeria ‘‘, Management of environment quality; An International Journal, vol. 17 1ssue 2,
pp. 209-215