See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/377729795

Dynamic Effects of Oil Price Shocks on Sectoral Outputs in Nigeria: Application of Linear and
Nonlinear ARDL
Preprint · January 2024
DOI: 10.13140/RG.2.2.25740.90243

CITATIONS READS

0 122

2 authors, including:

Isiaka Akande Raifu

Centre for the Study of the Economies of Africa
106 PUBLICATIONS 549 CITATIONS

SEE PROFILE

All content following this page was uploaded by Isiaka Akande Raifu on 27 January 2024.

The user has requested enhancement of the downloaded file.

 Dynamic Effects of Oil Price Shocks on Sectoral Outputs in Nigeria: Application of
Linear and Nonlinear ARDL

Isiaka Akande Raifu

Department of Economics
Faculty of Economics and Management Sciences
University of Ibadan
Ibadan, Oyo State, Nigeria [email protected]
+2348022975084

and

Deborah Bolanle Oloyede

Department of Economics
Faculty of Economics and Management Sciences
University of Ibadan
Ibadan, Oyo State, Nigeria
[email protected]
+2349038936381

All correspondences should be directed to Isiaka Akande Raifu of the Department of Economics, Faculty
of Economics and Management Sciences, University of Ibadan, Oyo State, Nigeria via the e-mail address:
[email protected]

Abstract
This study was carried out to examine the dynamic effects of oil price shocks on sectoral outputs in
Nigeria over the period of 1982-2016 using linear and nonlinear autoregressive distributed lag methods
(ARDL). The results from the linear ARDL estimation showed that oil price shocks affect agricultural
sector negatively in the short-run and the long-run. However, oil price shocks affect industrial sector
positively in the short-run while the services sector is affected positively in both runs. The findings from
the nonlinear ARDL showed that negative oil price shocks posively influence all the sectors of the
economy while the positive oil price shocks are harmful them in both runs. These findings, however,
depend on the measures of oil price shocks. Co-integration also exists between oil price shocks and
sectoral outputs with the speed of convergence varying across the sectors. Based on these findings, the
study recommended the following: 1) A special fund be established to cater for various sectors affected by
unfavourable oil price shocks 2) Favourable policies are formulated and implemented and 3) The
economy is diversified and alternative sources of energy be prioritised.

Keywords: Oil price shocks, agricultural output, industrial output, services output, ARDL,
NARDL

JEL Classification: C22, E30, L71, P22, Q10

1.0. Introduction
It was in the 1970s when the world first experienced a significant jump in the price of crude oil on the
international market. This sudden rise in the price of crude oil brought concerns among economists and
policymakers on the implications of crude oil price hikes for the economies, particularly of developed
countries that were the largest importers of crude oil at that time. There was a series of empirical studies
that followed to examine the effects of such an increase in oil prices on some macroeconomic
fundamentals, particularly the gross domestic product. The submission from the initial series of studies
was that the oil price hike causes a decline in economic growth (Rasche and Tatom, 1977; Hamilton,
1983; Burbridge and Harrison, 1984). The rationale for this finding is as follows: crude oil is considered
as an input in the production of goods in oil-importing countries. Consequently, a rise in crude oil prices
increases the cost of production of firms. This rise in production cost is in turn passed down to consumers
in the form of higher prices. If salaries or wages remain unchanged, an increase in the prices of goods
will weaken the consumer’s purchasing power which further results in a decline in effective demand. The
decline in effective demand, coupled with the increase in the cost of production, will lead directly or
indirectly to a decline in production.1 When industrial production falls, the aggregate output will fall and
unemployment will rise. The reverse is the case of a decline in the price of crude oil (Barro, 1984; Gisser
and Godwin, 1986; Lee, Ni, and Ratti, 1995).

However, as crude oil price continues to experience ups and downs, attention is shifted to the developing
economies, particularly to the oil-exporting countries. Different authors have found different effects of oil
price changes on developing economies and the effects have been found to depend on the extent to which
a country relies on the revenue from crude oil, as well as whether it is a net oil importer or exporter. In
most of these countries, government revenues are closely linked to oil production and price. When the
price of crude oil increases, it is expected to bring fortune to the country which will avail the government
the opportunity to have enough money for investment in critical infrastructural facilities such as pipe-
borne water, roads, hospitals and electricity generation. This, however, is not the end of the story,
particularly for some oil-producing countries. Some authors have noted that the increase in oil price has
not resulted in any meaningful investment that could have spurred growth and development in those
countries, owing to what they referred to as “The Dutch Disease Problem”. Instead of engaging in
meaningful investment, most governments in oil-producing countries usually engage in unproductive

investment, misallocation or misappropriation of oil funds and other forms of wasteful spending (Akpan,
2009; Okonjo-Iweala and Osafo-Kwaako, 2007; Olomola, 2006; El-anshasy, Bradley, and Joutz, 2005).

1 To minimise the burden of the high cost of production, the firms may cut production or reduce the number of workforce if
the salary of the workers is to remain unchanged.
 In carrying out empirical studies on the oil price shocks-growth nexus, the initial series of studies
assumed that the relationship between oil price shocks and the economy is linear (Rasche and Tatom,
1977;
Hamilton, 1983; Burbridge and Harrison, 1984). However, in response to Hamilton’s work of 1983, Mork
(1989) noted that macroeconomic parameters or variables might respond differently to an increase in the
price of crude oil compared with a price decrease. This brought about the term, ‘asymmetric response of
the economy to oil price shocks’. Several empirical studies that followed Mork’s work have examined the
asymmetric effect of oil price shocks on different macroeconomic variables, albeit, with mixed findings
(Pindyck, 1991; Lee, Ni, and Ratti, 1995; Hamilton, 1996; Borenstein, Cameron, and Gilbert, 1997;
Brown and Yucel, 2002; Barsky and Kilian, 2004; Kliensen, 2006; Lescaroux and Mignon, 2008; Du,
Yanan, and Wei, 2010). Along this line, several empirical studies have also been conducted in
developing oilproducing economies (Walker and Ekpemauzor, 1990; Ayadi, 2005; Iwayemi and
Fowowe, 2011a; Iwayemi and Fowowe, 2011b; Chuku, 2012; Rezazadehkarsalari, Haghiri, and
Behrooznia, 2013; Moshiri, 2015; Abdulkareem and Kilishi, 2016; Sadeghi, 2017).

It has been submitted that to formulate and implement good and desirable policies to tackle the negative
effects of oil prices on the economy, policymakers and experts must understand the effects of oil price
shocks on the various sectors of the economy. This is because it is believed that before the shocks
emanating from oil price fluctuations affect the aggregate economy, it must have first affected the sectors
that constitute it. The explanation given is that crude oil is part of the inputs used in producing the outputs
of those sectors. Hence, when the price of oil fluctuates (rises or falls), its consequences fall first on
sectoral outputs. To this end, some studies have been carried out both in oil-consuming and oil-producing
countries to examine the effects of oil price shocks on the various sectors of the economy (Hanson,
Robinson and Schluter, 1993; Lee and Ni, 2002; Dhuyvetter and Kasterns, 2005; Jimenez-Rodriguez,
2007; Tokgoz, 2009; Mehrara and Sarem, 2009; Bolaji and Bolaji, 2010; Guidi, 2010; Torul and Alper,
2010; Fukunaga, Hirakata, and Sudo, 2010; Herrera, Lagalo, and Wada, 2011; Eski, Izgi, and Senturk,
2011; Shaari, Pei, and Rahim, 2013; Binuomote and Odeniyi, 2013; Baumeister and Kilian, 2013; Ikram
and Waqas, 2014; Nigatu, Hjort, Hansen, and Somwanu, 2014; Ee, Gugkang, and Husin, 2015; Aimer,
2016; Copiello, 2016; Seth, Giridhar, and Krishnaswami, 2016; Riaz, Sial, and Nasreen, 2016; Mahboub
and Ahmed, 2017; Liew and Balasubramanian, 2017; Okoye, Mbakwe, and Igbo, 2018; Raifu, 2021;
Raifu and Oshota, 2023; Raifu and Aminu, 2021; Raifu and Afolabi, 2023; Raifu, 2023a; Raifu, 2023b;
Raifu, 2023c).

From the studies above, empirical evidence is varied across countries and sometimes controversial. The
rationale for the mixed findings stemmed from the methodological approaches employed, particularly in
terms of estimation techniques adopted by different researchers. While the majority of the initial studies,
as aforementioned, assumed a linear relationship between oil price shocks and sundry macroeconomic
variables, others believe that the relationship cannot be linear over time. Since oil price usually fluctuates,
that is, it rises and falls depending on the situation in the international community; macroeconomic
variables are believed to be responding differently to the rise and fall of oil prices. In response to this
assertion, several econometric techniques have been used to test the asymmetric relationship between oil
price shocks and the economy.

The employment of asymmetric econometric methods to model the oil-price shocks-economy nexus
involves two approaches. The first approach is through the application of linear econometric techniques
such as OLS, VAR and SVAR. Different methods are used to decompose oil prices into positive and
negative components and linear econometric techniques are used to examine the response of
macroeconomic variables to the positive and negative oil price shocks (see Mork, 1989; Lee, Ni, and
Ratti, 1995; Hamilton, 1996; Brown and Yucel, 2002). The second approach entails the use of
econometric techniques specifically developed to capture asymmetric or nonlinear relationships between
macroeconomic variables. These econometric techniques include Threshold VAR, Multivariate Threshold
VAR, Smooth Transition VAR, Logistic Smooth Transition VAR, Markov Switching Method (MSM) and
Nonlinear Autoregressive Distributed Lag Method (NARDL). These econometric techniques have been
used extensively in the literature to model oil price shock-macroeconomic variables nexus. For instance,
Sardosky, (1999) used Threshold VAR to examine the impact of oil price shocks on market activity in the
US. Huang, Hwang, and Peng, (2005) investigated the asymmetric impact of oil price shocks on economic
activities employing a Multivariate Threshold VAR econometric technique. The same method was applied
by Catik and Onder, (2013) for the Turkish economy. Logistic Smooth Transition VAR was applied by
Rahman and Serletis, (2010) to examine the asymmetric effect of oil price shocks on monetary policy
while Cologne and Manera, (2009) employed the Markov-Switching econometric technique to analyse the
asymmetric impact of oil price shocks on the output growth of G-7 countries. Recently, Liew and
Balasubramanian, (2017) employed nonlinear ARDL to investigate the effects of oil price shocks on the
sectoral outputs for Malaysia.

This study joins the family of symmetric and asymmetric econometric techniques to examine the
dynamic effects of oil price shocks on the sectoral outputs (Liew and Balasubramanian, 2017). To the
best of our knowledge, this is the first study that applies a combination of ARDL and NARDL to
investigate the dynamic effects of oil price shocks on the economy, particularly at the sectoral level in
Nigeria. Most of the studies that focused on the analysis of oil price shocks in different sectors of the
economy in Nigeria used a linear or static technique like the Computable General Equilibrium Model
(Bolaji and Bolaji, 2010; Binuomote and Odeniyi, 2013; Okoye, Mbakwe, and Igbo, 2018). The
contributions of this study to the existing literature are as follows: First, this study examines the effects of
oil price shocks on the three sectors of Nigeria, namely agricultural, industrial and services sectors
comparing linear (symmetric) and nonlinear (asymmetric) effects by using ARDL and NARDL. Second,
the adjustment of sectoral outputs in response to oil price shocks is computed using a dynamic multiplier
method based on the NARDL technique. Third, for robust analysis, consistency and validation of our
baseline results in which we used nominal oil price, this study also considers the case of the effects of real
oil price shocks on the aforementioned sectoral outputs.

Following the introductory section, the rest of the study is structured as follows. Section two reviews the
extant literature on the effects of oil price shocks on different sectors of the economy. Section three
presents information about data sources and descriptions as well as methodological approaches employed
in this study. Section four presents the empirical findings and discussion while section five concludes and
offers policy recommendations.

2.0. LITERATURE REVIEW

There are vast empirical studies on the effects of oil price shocks on the economy of both advanced and
developing economies, particularly since the very work of James Hamilton in 1983. As the studies cut
across different countries and continents, the findings of the studies are diverse. For an extensive review
of the literature on the impact of oil price shocks on the macroeconomic variables, see the work of Jones
and Leiby (1996); Kilian (2008); Oladosu, Leiby, Bowman, Uria-Martinez, and Johnson (2018). In this
study, we concentrate on the review of existing studies that focus on the analysis of the effects of oil price
shocks on the sectors or subsectors of the economy, either at macro or micro levels.

Beginning from the agricultural sector, the first strand of studies in this area used the Computable
General Equilibrium technique (CGE hereafter) to examine the effects of oil price shocks on the
agricultural sector (Hanson, Robinson and Schluter, 1993; Tefera, Worku and Ayalew, 2012; Nkang,
2018). Hanson, et al., (1993) investigated the sectoral effects of world oil price shocks on the agricultural
sector in the United States using CGE technique. They found that the world oil price shocks did not only
affect the agricultural sector through direct and indirect energy costs but also exchange rate and foreign
borrowing adjustment to higher oil import costs and government support programmes for agriculture.
Tefera, Worku and Ayalew, (2012) examined the impact of an increase in the price of crude oil on the
international market and oil subsidy scheme on the Ethiopian economy. They submitted that oil price
shocks lead to the depreciation of the Ethiopian currency (Ethiopian Birr), which further increases
agricultural traded goods output. In a similar version, Nkang, (2018) looked at the effect of a 50 per cent
decline in the world oil price on agriculture and household welfare in Nigeria employing the CGE
method. The findings from his study showed agricultural GDP and composite supply surge leading to a
decline in the price of agricultural commodities. Consequently, agricultural imports fell and exports rose.
Furthermore, according to their study, oil price shocks also lead to a reduction in incomes/expenditures of
all household groups except those in the urban and rural north.

Still, on the effects of oil price shocks on agricultural sector output, other studies used time series data
and different econometric techniques. Tokgoz, (2009) examined the impact of energy markets on the EU
agricultural sector using the panel OLS method. The study found that higher crude oil price leads to an
increase in ethanol consumption, production and grain prices. Binuomote and Odeniyi (2013) analysed the
impact of crude oil prices on agricultural productivity in Nigeria using the Error Correction Method
(ECM). Their study showed that oil price serves as a determinant of agricultural productivity only in the
short–run. In Pakistan, Ikram and Waqas (2014) focused on the impact of crude oil price shocks on
agricultural productivity and documented that oil price has a negative effect on agricultural productivity
growth. Ee et al. (2015) examined the effects of oil price shocks on three sectors including the agricultural
sector in Malaysia and found a long-run relationship between oil prices and the agricultural sector.
However, they noted that Granger-causality occurred only in the short-run. Aimer (2016) also showed that
a negative longrun relationship exists between oil prices and the agricultural sector in the case of Libya.
While these studies examined the effects of oil price shocks on agricultural sector output, assuming a
linear relationship between the two variables, a study by Liew and Balasubramanian (2017) investigated
the effect of the asymmetric effect of oil price shocks on sectoral outputs in Malaysia, including the
agricultural sector.
However, their study’s findings showed that there is no asymmetric relationship between oil price shocks
and agricultural sector output.

There are considerable studies that focus on how the oil price shocks affect the industrial sector and its
subsectors, particularly the manufacturing sector. Beginning with the study by Jimenez-Rodriguez (2007)
who analysed the impact of the price shocks on the manufacturing sector outputs of six OECD countries
comprising France, Germany Italy, Spain, the UK and the US, their study showed that there exists
heterogeneity in the response of industrial outputs to oil price shocks in France, Germany, Italy and Spain
in a way similar to the UK and the US. Mehrara and Sarem, (2009) also showed that co-movement exists
between oil price and industrial output in three oil-exporting countries namely Iran, Indonesia and Saudi
Arabia. They also documented the existence of Granger-causality between oil price and industrial output
with causality running from oil price to industrial output. However, Lee and Ni, (2002) observed that oil
price shocks reduce the supply of energy-intensive industries supply and that of energy demand
automobile-related industries. Bayar and Kilic, (2014) concluded that an increase in oil and natural gas
prices negatively affects industrial production in the Eurozone. A similar finding was reported by
AlRisheq (2016), who studied the impact of oil prices on industrial production in developing countries.
Specifically, the study found that an increase in oil prices negatively affects industrial production in
developing countries. Seth, Giridhar, and Krishnaswami (2016) submitted that oil price shocks affect
industrial production through the costs of raw materials. Bolaji and Bolaji (2010) went further to state that
it is not only the cost and quantity of raw materials that are affected by the increase in the price of oil but
also the demand for the manufacturing products, the profits and the rate of turnover of manufacturing
companies. Mahboub and Ahmed (2017) stated that the effect of price shocks on the manufacturing
output only occurred in the long-run, after 10 quarters in Saudi Arabia. However, Fukunaga, Hirakata, and
Sudo (2010) stated that the way the changes in oil prices affect industrial output depends on the
underlying shocks driving oil price changes as well as the characteristics of the industry. Thus, they
concluded that unexpected disruptions of oil supply act mainly as negative supply shocks to oil-intensive
industries and also act mainly as negative demand shocks to less oil-intensive industries. With regard to
the case of a nonlinear relationship between oil price shocks and industrial or manufacturing output, the
overriding submission from several empirical studies is the existence of an asymmetric relationship
between oil price shocks and industrial or manufacturing output, albeit using different estimation
techniques (Mehrara and Sarem, 2009; Guidi, 2010; Herrera, Lagalo, and Wada, 2011; Riaz, Sial, and
Nasreen, 2016; Liew and
Balasubramanian, 2017). Mehrara and Sarem, (2009) used nonlinear Granger causality, Herrera, et al.
(2011) used VAR, Riaz, Sial, and Nasreen (2016) employed GARCH/ARDL while Liew and
Balasubramanian (2017) used NARDL. However, Torul and Alper (2010) who conducted a study on the
asymmetric effects of oil prices on the manufacturing sectors in Turkey found that none of the
manufacturing subsector except coke and refined petroleum is statistically significant either in the case of
the linear or nonlinear specification.

While there appear to be vast studies on the effects of oil price shocks on industrial and agricultural
outputs, such studies on services sector output are very scanty. Even the few available studies yield
inconclusive findings. While the study by Ee, et al. (2015) showed the existence of short-run causality
between changes in oil price and the service sector, the study by Liew and Balasubramanian (2017)
concluded that there is no evidence of the nonlinear relationship between oil price and services sectors
output. Tefera, Worku and Ayalew (2012) considered the effect of an increase in the international price of
crude oil and oil subsidy on the economy of Ethiopian using CGE. They found that an increase in oil
prices leads to a depreciation of the exchange rate, which further results in a decline in service sector
output.

3.0. Data Sources and Description, Trend Analysis and Methodology

3.1. Data Sources and Description

This study employs quarterly data spanning 1982 to 2016 to examine the dynamic effects of oil price
shocks on sectoral outputs in Nigeria. The variables used include Brent crude oil price measured in dollars
per barrel, real agricultural output, real industrial output and real services output. The real agricultural
output, real industrial output and real services output are measured in Billion Naira. The Brent oil price
and sectoral outputs were sourced from the World Bank Commodity Price Data Sheet and the Central
Bank of Nigeria Database respectively.2The Brent crude oil price is a monthly data series aggregated to
quarterly data by using an average of three months of data while the sectoral output is annual data
disaggregated into quarterly series using the quadratic matching method in EVIEWS 9. 3 The growth rates
of the real agricultural output, real industrial output and real services output are computed using Excel.

Following existing studies such as Binuomote and Odeniyi (2013) and Ikram and Waqas (2014),
variables such as labour employed in the agricultural sector, agricultural land and agricultural machinery,
real effective exchange rate and real interest rate are included as explanatory variables in the agricultural
output-oil price model. In the industrial output-oil price model, this study follows Guidi (2010) and Torul
and Alper (2010) by including variables such as money growth, real interest rate, real effective exchange
rate, inflation rate and capital stock as explanatory variables. Concerning services output-oil price model,
variables such as real effective exchange rate, inflation rate and interest rate are included following the
study by Guidi (2010). The agricultural land (measured as a percentage of total land area), agricultural
machinery (measured as tractors per 100 square kilometres), money growth (M2 growth rate),
CPI/inflation rate, real effective exchange rate, real interest rate and lending prime rate are all sourced
from the World Development Indicators. Agricultural employment as a percentage of total employment is
sourced from the National Bureau of Statistics and National Manpower Board. Capital stock (real or
nominal) is obtained from Penn World Table 8.1. These are annual data disaggregated into quarterly data.

Table 1 presents the summary statistics of important variables. According to the table, the average Brent
crude oil price stood at 42.07 USD/barrel during the period under consideration. The minimum and the
maximum price of crude oil stood at 11.09 USD/barrel and 122.48 USD/barrel respectively. In the case of
world crude oil price (average of Brent crude oil price, Dubai crude oil price and WTI crude oil price), the
mean value stood at 41.10 USD/barrel, ranging from 11.79 USD/Barrel to 120.97 USD/barrel. Among the
2 According to CBN statistical bulletin (2016), the agricultural sector is production, livestock, forestry and fishery subsectors;
industrial sector majorly comprises of crude petroleum and natural gas made up of crop, solid mineral and manufacturing
subsectors and services sector consists subsectors such as transport, information and communication, utilities,
accommodation and food services, finance and insurance, real estate, professional, scientific and technical services,
administrative and support services, public administration, education, human health and social services, arts, entertainment
and recreation.
3 Quadratic-match average performs a proprietary local quadratic interpolation of the low-frequency data to fill in the high
observations.
sectors, the services sector recorded the highest average output, which stood at N10, 004.55 billion. The
services sector is closely followed by the industrial sector with an average output of, 383.81 billion while
the lowest sector, in terms of average, is the agricultural sector with an average output of N7, 293.31
billion. The minimum and maximum amount of output produced by the services sector stood at N3,
699.33 billion and N25, 478.52 billion while that of the industrial sector stood at N5, 181.53 billion and
N13, 849.31 billion respectively. The agricultural sector output also ranged from N2, 259.79 billion to
N16, 871.23 billion. It can be observed from the table that the standard deviation which shows the degree
of dispersion of the variables from their means behaves normally as its values do not deviate greatly from
their means.

Table 1: Summary Statistics of Major Variables

Variable BRENTOILPRI COPA AGR IND SER
Mean 42.07 41.10 7,293.31 9,383.81 10,004.55
Minimum 11.09 11.79 2,259.79 5,181.52 3,699.33
Maximum 122.48 120.97 16,871.23 13,849.31 25,478.52
Std. Dev. 31.89 30.49 4,691.14 2,477.20 7,064.82
Observations 140 140 140 140 140
Source: Computed by the author
Note: BRENTOILPRI, COPA, AGR. INDU and SER denote Brent crude oil price, world oil price, agricultural sector output, industrial
sector output and services sector output respectively

3.2. Oil Prices and Sectoral Output

This section presents the stylised facts of the oil price and sectoral output growth rates in Nigeria. Several
studies have discussed the evolution of oil price fluctuations over time and its implications for the
economy in Nigeria (Ayadi, 2005; Olomola, 2006; Akpan, 2009; Iwayemi and Fowowe, 2011a; Iwayemi
and Fowowe, 2011b; Chuku, 2012). The focus of this section is to analyse the trend of the relationship
between oil price fluctuations and each sector's output.

Figure 1 shows the trend of the relationship between oil price and industrial output growth rate over the
last 36 years. From the figure, it could be inferred that oil price fluctuations and the industrial sector
output fluctuation move together. Put in another way, in most of the years under consideration, an
increase in oil price is often associated with a decline in industrial sector output growth and vice versa in
the case of a decline in the price of crude oil. Towards the tail end of the graph, particularly during 2014
when the oil price suddenly declined in the international market, industrial sector output recorded a
negative growth rate, which led to the economic recession in the country in the second quarter of 2016.
Thus, it could be succinctly put that oil price fluctuations or shocks are the precursors to industrial output
fluctuation in Nigeria. This discovery is not surprising because Nigeria is a net exporting country. Apart
from the fact that Nigeria exports crude oil, it also imports refined oil products such as diesel, kerosene
and petroleum, otherwise known as Premium Motor Spirit (PMS). Besides, Nigeria is incessantly facing
the challenge of electricity generation. This, coupled with the unstable nature of oil prices on the
international market often raise the costs of production of firms. Most of the industries in Nigeria rely on
the use of generators to power their machines in the course of production. The rising costs of production
negatively affect the production of the firms. Thus, even though an increase in oil prices appears to be
important to the government in Nigeria, it seems to be detrimental to industrial output.

In Figure 2, the trend of the relationship observed between oil price fluctuations and the services sector is
similar to the one observed between oil price and industrial production. This implies that as oil price
fluctuates, the services sector output also fluctuates. However, the amplitude of fluctuation between oil
prices and the services sector is not the same as observed in the trend of the relationship between oil
prices and industrial outputs. This noticeable difference could be a result of the level of energy utilisation
by the services sector compared with that by the industrial sector.

Figure 3 depicts the relationship between oil price fluctuation and the agricultural sector output growth
rate. From the figure, it appears that there is no discernable pattern between oil price and agricultural
output except during the periods of 1985-1986 and 1999-2003. The period of 1985-1986 was the period of
the Structural Adjustment Programme, a period of economic and financial liberalisation policy which had
a significant impact on the sectors of the economy. The period of 1999-2003 coincided with the period of
the inception of the democratic era of the 4 th republic, which ushered-in several economic reforms and
tremendous economic growth. However, the lack of a discernable pattern between oil prices and the
agricultural sector is not surprising because there are shortages of intensive energy consumption by
agricultural industries in Nigeria. Most of the highly energy-intensive agricultural industries are in a
comatose state, particularly the textile industry while the bulk of cash crop production and arable crop,
rearing of animals, fishery (fishing business) and forestry are still done by the people in the rural areas
using crude implements such as a hoe, cutlass, hook and lots more. Hence, the lack of a pattern of the
relationship between oil price fluctuation and output growth rate in the agricultural sector is unclear.
 Figure 1: The Trend of Oil Price and Sectoral Output Growth in Nigeria (1982-2016)

3.3. Methodology

This study aims to examine the relationship (linear and nonlinear) between oil price shocks in the three
sectors of the Nigerian economy, industrial, services and agricultural sectors. To implement this objective,
the study employs the econometric methods known as Linear and Nonlinear Autoregressive Distributed
Lag (L/N-ARDL). However, we first perform unit root tests using Augmented Dickey-Fuller and
PhilipsPerron to examine the characteristics of our variable of interest. There are two rationales for
carrying out the unit root tests. The first is to avoid running spurious regression. The second is to ensure
that none of the variables is integrated of order 2, I (2), since the ARDL and NARDL are only applicable
when the variables are integrated of order 0 and 1, that is, I(0) and I(1).

We begin the model specification by specifying the ARDL framework following Pesaran, Shin and Smith
(2001) and empirical studies by (Raifu, Aminu and Folawewo, 2020; Aminu and Raifu, 2019; Raifu and
Raheem, 2018, Raifu and Aminu, 2020; Raifu and Aminu, 2021; Raifu, 2021; Nnadozie, Emediegwu and
Raifu, 2022) as follows:
p q1 q2

SOt = + 01SOt−1 +2OPt−1 +i' Xt−1 +iSOt−1 +iOPt−1 +i'Xt−1 +t (1)
i=1 t=0 i=0

Where SOis the sectoral outputs (agricultural sector output, industrial sector output and services sector
output),OP denote oil price shocks, X s' are the other explanatory variables, apart from the oil price, that
can affect each of the sectoral output. For the industrial sector model, we include explanatory variables
such as money growth, real effective exchange rate, real interest rate and inflation. In the model of the
agricultural sector, explanatory variables such as agricultural land, agricultural machinery, real effective
exchange rate and real interest rate. Real effective exchange rate, real interest rate and inflation are
included in the service sector model. 0 is a constant denoting a drift component of the model, 1 and 's
are referred to as the long-run coefficient parameters. i , i and i'sare the short-run dynamic parameters.
t is the error term assumed to be independently and identically distributed with zero mean and constant
variance.

Following theoretical arguments, it is expected that an increase in the oil price should have an adverse
effect on sectoral output and vice versa for a decrease in the oil price (Hamilton, 1983; Mork, 1989;
Mory, 1993). It is also expected that an increase in the money supply would positively influence output
including sectoral outputs. The increase in money supply lowers the cost of borrowing and the interest
rate. By lowering the cost of borrowing, there is more money available for borrowing in the economy
which in turn does not only raise household consumption and investment but also raises sectoral outputs.
However, this is not without a caveat. An increase in money supply can fuel inflation which could be
disastrous to the economy. Besides, the ultimate effect of money supply depends on the medium of
transmission (see Ezeaku, et al. 2018). The theoretical channels through which the real exchange rate
influences the economy depend on whether it is appreciating or depreciating. Depreciation of currency
(that is naira becomes weak against the US dollar) leads to an increase in the prices of foreign goods
relative to domestic goods. This implies that the depreciation of the exchange rate diverts economic
agents' spending from foreign goods to home goods thereby stimulating domestic economic activities.
The reverse is the case when the currency appreciates. The appreciation of the exchange rate harms
domestic economic activities because it makes foreign goods cheaper than domestic goods. Thus, the real
exchange rate can positively or negatively influence domestic outputs even at sectoral levels (Kandil and
Mirazaie, 2005; Baggs, Beaulieu and Fung 2010; Aminu, Raifu and Oloyede, 2019; Mlambo, 2020). As
in the case of the exchange rate, the theoretical channels through which interest rate affects the economy
 remain inconclusive. Theories such as the neoclassical growth model, Tobin's monetary growth model
and the neo-Keynesian model posit that interest rate negatively affects output because interest rate serves
as a cost of borrowing. Thus, a highinterest rate raises a firm's cost of capital and dampens investment
thereby affecting output negatively. However, some strands of theories, especially the theory of
irreversible investment, suggest that a high rate of interest rate can only affect economic growth
adversely in the short run. A high-interest rate can encourage investment activity that would spur
economic growth (see Bertola and Caballero, 1994). Thus, it can be said that interest rates can have either
a positive or negative effect on sectoral output (see Khatkhate 1988; Drobyshevsky, et al. 2017). Finally,
we expect that agricultural land and machinery should have a positive impact on agricultural output
(Ojiya, et al., 2013; Raifu and Aminu, 2019).

The null hypothesis of ARDL in the long-run is specified as follows:

  0 = = =1 i
'
s0

The alternative hypothesis is specified as follows:

  0  1i's0

When the null hypothesis is tested against the alternative hypothesis and the outcome shows that
cointegration exists, the error correction model (ECM) which captures the dynamic adjustment towards
the long-run equilibrium from the short-run disequilibrium is specified as follows:
p q1 q2

SOt = + 0  iSOt−1 +iOPt−1 +'Xt−1 +ectt−1 +t (2)

i=1 t=0 i=0

Where ectt−1is the error correction term. Theoretically, the coefficient of the error correction term ( ) must
be negative, less than one and statistically significant for the adjustment from the short-run disequilibrium
towards the long-run equilibrium can take place.

The next is to specify NARDL developed by Shin, Yu, and Greenwood-Nimmo, (2014). Following Liew
and Balasubramanian (2017) , who modelled the effects of oil price shocks on the sectoral outputs in
Malaysia, NARDL is specified as follows:

p q1 q2

 = +SOt  01SOt−1 +2OPt−+1 +3OPVt−−1 +i' Xt−1 +iSOt−1 +(i+OPt−+1 +i−OPt−

−1)+i'Xt−1 +t
i=1 i=0 i=1

(3)
 From equation 1, p , q1andq3 represent lag orders, 0, 1, 2, 3 and i's denote constant and long-run

coefficients of sectoral outputs and positive and negative oil price shocks and other explanatory variables
q q

respectively. + and− represent the asymmetrically distributed lag parameters which depict the
i=0 i=0

short-run impact of oil price shocks on sector outputs. OPt−+1 andOPt−−1 are the partial sum of positive and

 
negative changes in oil prices, denoting oil price shocks. 1+ =− 2 and 2− =− 3
capture the effects of
1 1
the increase and decrease in oil price shocks on the sectoral outputs in the long-run. Both the null
q q

hypotheses of the short-run symmetry(+ =− ) and the long-run symmetry ( 1+ = 2−) are tested
i=0 i=0

using the standard Wald test that follows the 2 distribution (Shin, et al. 2014). If the Wald test is not
statistically significant, then it means that only a symmetric relationship exists between Oil prices and
sector outputs. However, if the Wald test is statistically significant, there is an asymmetric relationship
between the two variables.

The partial sum of positive and negative changes in oil price shocks can be computed as follows:
t t t t

OPt+ =OPit =max(OP ,0);t OPt1 =OPt− =min(OPt ,0) (4)

i=1 i=1 i=1 i=1

The error correlation model (ECM) for NARDL is specified as follows:

p q1 q2

SOt = + 0  iSOt−1 +(i+OPt−+1 +i−OPt−−1)+i'Xt−1 +ectt−1 +t (5)

i=1 i=0 i=0

Where ectt−1remains as defined in the case of linear ARDL and is the coefficient of error correction
term.

Finally, the asymmetric dynamic multiplier effects of positive and negative changes in oil price shocks on
the sectoral output growth rates are computed as follows:

mq SOt i ;mq =q SOt i ,q = 0,1,2,... (6) q+ =i=0 OPT++ − i=0 OPT+−
Where q→, mq+ →+ and mq− →− , + and − denote the asymmetric long-run coefficients.

4.0. Presentation of Findings and Discussion

4.1. Correlation Analysis Results

Table 2 presents the results of the correlation analysis among the variables. As shown in the table, two
types of crude oil prices are considered in both nominal and real terms. The first is Brent crude oil price
while the second is the average of the world crude oil price, which consists of Brent crude oil price, West
Intermediate Texas crude oil price and Dubai crude oil prices. The correlations of these oil prices are
considered with the outputs of the three sectors-industrial, agricultural and services sectors as well as
some other explanatory variables included in this study. The results show that Brent crude oil price, either
in nominal or real terms (BRN and BRR), is negatively correlated with industrial sector output (INDR).
The same results are obtained in the case of the correlation between industrial sector output and average
world crude oil price (AOPN and AONR). However, the observed negative correlations are not
statistically significant. Similarly, both Brent crude oil price and average world crude oil price, capital
stocks (CRCS), money growth (MG), real effective exchange rate (REER), and inflation (INF) all have
negative insignificant associations with industrial sector output while the real interest rate (RIR) is
insignificantly related to industrial sector output. In the case of the agricultural sector, the correlations
between the Brent crude oil price, the average world crude oil price (nominal and real terms), and the
agricultural output are negative, albeit only real-term oil prices have a significant relationship with
agricultural output at 10 per cent level of significance. Agricultural land and machinery (AGL and AGR)
are positively associated with agricultural sector output; however, only agricultural land is statistically
significant at the 5 per cent level of significance. It is also observed that real effective exchange rate and
real interest have an insignificant negative relationship with agricultural output. With regard to the
services sector, nominal Brent crude oil and average oil price depict positive and significant relationship
with the services sector output while the real Brent oil price and average world oil price are found to be
significantly negatively correlated with services sector output. The real effective exchange rate and
inflation rate have a negative and significant relationship with the services sector output while the real
interest rate has a positive and significant correlation with the services sector in both nominal and real
models.

4.2. Unit Root Tests Results

In Table 3, the results of unit root tests are presented. We performed the unit root tests to determine the
stationary properties of the variables we used in this study. We carried out both Augmented Dickey-Fuller
(ADF) and Philips-Perron (P-P) unit root tests at constant and trend and constant. Both unit root tests rest
on the assumption that the variables contain unit roots. In other words, the variables are not stationary at
level and they can only be made stationary after the first difference. According to the Table, the results
show, either at intercept or trend and intercept, the mixture of I(0) and I(1), that is, the results depict the
mixture of integration of order zero and order one. This finding is suitable for the application of both
linear and nonlinear ARDL methods. Most importantly, none of the variables is integrated of order two I
(2), which would have rendered the use of these methods (ARDL and NARDL) useless.

4.3. Causality Results

The results of the causality test are presented in Table 4. We carried out both Granger-causality and
TodaYamamoto causality tests to determine the direction of causality, particularly between our variables
of interest; crude oil prices and sectoral outputs. The null hypothesis of both causality tests is that there is
no causality between the variables against which the alternative hypothesis (i.e. there is causality between
the variables) is tested. Beginning from the agricultural sector output and both oil prices (Brent and World
Oil Prices) in nominal and real terms using Granger-causality and Toda-Yamamoto causality tests, the
results show that there is no causality between agricultural output and Brent crude oil price (or world
crude oil price). However, in the case of industrial sector output, there is a unidirectional causality
between the nominal Brent oil price (or nominal World oil price) and industrial sector output with the
direction of causality running from nominal oil prices to industrial sector output in both methods
(Granger-causality and Toda-Yamamoto causality tests). As in the case of the agricultural sector, there is
no direction of causality between either nominal or real oil prices (Brent or World Oil Prices) and services
output under both Granger-causality and Toda-Yamamoto causality tests. Intuitively, the results could
imply that only the industrial sector, which consumes more energy to produce its output is where the
crude oil price fluctuations could have causal effects.
Table 2: The Correlation Analysis Results
Variable Brent Crude Oil Price Average World Crude Oil Price
INDR BRN BRR GRCS MG REER INF RIR INDR AOPN AOPR GRCS MG REER INF RIR
INDR 1.000 INDR 1.000

BRN -0.005 1.000 AOPN -0.007 1.000

(0.948) (0.933)
BRR -0.004 -0.333 1.000 AOPR -0.006 -0.351 1.000
(0.950) (0.000) (0.944) (0.000)
GRCS -0.132 0.835 -0.259 1.000 GRCS -0.132 0.835 -0.260 1.000
(0.120) (0.000) (0.002) (0.120) (0.000) (0.002)
MG -0.017 -0.119 -0.093 -0.260 1.000 MG -0.017 -0.117 -0.091 -0.260 1.000
(0.838) (0.161) (0.275) (0.002) (0.838) (0.170) (0.283) (0.002)
REER -0.128 -0.062 0.419 0.022 -0.503 1.000 REER -0.128 -0.059 0.421 0.022 -0.503 1.000
(0.131) (0.467) (0.000) (0.795) (0.000) (0.131) (0.485) (0.000) (0.795) (0.000)
INF -0.021 -0.441 0.105 -0.390 0.266 -0.179 1.000 INF -0.021 -0.444 0.106 -0.390 0.266 -0.179 1.000
(0.805) (0.000) (0.218) (0.000) (0.002) (0.034) (0.805) (0.000) (0.211) (0.000) (0.002) (0.034)
RIR 0.056 0.187 -0.232 0.181 -0.041 0.097 -0.482 1.000 RIR 0.056 0.187 -0.233 0.181 -0.041 0.097 -0.482 1.000
(0.509) (0.027) (0.006) (0.033) (0.628) (0.253) (0.000) (0.509) (0.027) (0.006) (0.033) (0.628) (0.253) (0.000)

AGRR 1.000 AGRR 1.000

BRN -0.028 1.000 AOPN -0.026 1.000
(0.743) (0.757)
BRR -0.160 -0.333 1.000 AOPR -0.159 -0.351 1.000
(0.060) (0.000) (0.061) (0.000)
AGL 0.167 0.369 -0.921 1.000 AGL 0.167 0.375 -0.923 1.000
(0.049) (0.000) (0.000) (0.049) (0.000) (0.000)
AGM 0.042 0.842 -0.651 0.528 1.000 AGM 0.042 0.844 -0.657 0.528 1.000
(0.624) (0.000) (0.000) (0.000) (0.624) (0.000) (0.000) (0.000)
REER -0.084 -0.062 0.419 -0.583 -0.127 1.000 REER -0.084 -0.059 0.421 -0.583 -0.127 1.000
(0.325) (0.467) (0.000) (0.000) (0.134) (0.325) (0.485) (0.000) (0.000) (0.134)
RIR -0.028 0.187 -0.232 0.146 0.319 0.097 1.000 RIR -0.028 0.187 -0.233 0.146 0.319 0.097 1.000
(0.743) (0.027) (0.006) (0.086) (0.000) (0.253) (0.743) (0.027) (0.006) (0.086) (0.000) (0.253)

AGRR BRN BRR AGL AGM REER RIR AGRR AOPN AOPR AGL AGM REER RIR
SERR BRN BRR REER INF RIR SERR AOPN AOPR REER INF RIR

SERR 1.000 SERR 1.000

BRN 0.441 1.000 AOPN 0.449 1.000
(0.000) (0.000)
BRR -0.452 -0.333 1.000 AOPR -0.452 -0.351 1.000
(0.000) (0.000) (0.000) (0.000)
REER -0.387 -0.062 0.419 1.000 REER -0.386 -0.059 0.421 1.000
(0.000) (0.467) (0.000) (0.000) (0.485) (0.000)
INF -0.260 -0.441 0.105 -0.179 1.000 INF -0.260 -0.444 0.106 -0.179 1.000
(0.002) (0.000) (0.218) (0.034) (0.002) (0.000) (0.211) (0.034)
RIR 0.175 0.187 -0.232 0.097 -0.482 1.000 RIR 0.175 0.187 -0.233 0.097 -0.482 1.000
 (0.039) (0.027) (0.006) (0.253) (0.000) (0.039) (0.027) (0.006) (0.253) (0.000)

Source: Computed by Authors using EVIEWS 9 Note: Probability values are in parenthesis in the table

Table 3: Unit Root Tests Results

Variables Intercept Trend and Intercept Decision

ADF P-P ADF P-P

Level First Diff Level First Diff Level First Diff Level First Diff
AGRR -3.53 -5.00 -2.89 -14.03 -3.47 -5.02 -2.88 -14.39 I(0)
( 0.0087) (0.0000) ( 0.0494) (0.0000) (0.0474) (0.0003) (0.1733) (0.0000)
INDR -3.57 -4.66 -3.63 -6.71 -3.69 -6.27 -3.40 -7.41 I(0)
(0.0078) (0.0002) (0.0064) (0.0000) (0.0268) (0.0000) (0.0552) (0.0000)
SERR -1.52 -3.05 -1.80 -6.50 0.03 -3.56 -1.45 -6.92 I(1)
(0.5182) (0.0328) (0.3792) (0.0000) (0.9963) (0.0379) (0.8428) (0.0000)
BNOP -1.19 -9.37 -1.39 -9.59 -2.80 -9.35 -2.30 -9.56 I(1)
(0.6764) (0.0000) (0.5846) (0.0000) (0.1984) (0.0000) (0.4308) (0.0000)
COPA -1.18 -9.46 -1.35 -9.47 -2.37 -9.44 -2.34 -9.44 I(1)
(0.6831) (0.0000) (0.6046) (0.0000) (0.3958) (0.0000) (0.4115) (0.0000)
BROP -1.93 -9.23 -1.92 -9.23 -1.94 -9.30 -1.64 -9.22 I(0)
(0.3156) (0.0000) (0.3230) (0.0000) (0.6272) (0.0000) (0.7733) (0.0000)
COPAR -1.89 -9.18 -1.97 -9.18 -1.90 -9.25 -1.74 -9.30 I(0)
(0.3362) (0.0000) (0.3004) (0.0000) (0.6469) (0.0000) (0.7304) (0.0000)
AGRL -5.30 -4.85 -4.28 -7.16 -2.23 -6.83 -1.88 -8.22 I(0)
(0.0000) (0.0001) (0.0007) (0.0000) (0.4713) (0.0000) (0.6616) (0.0000)
AGRMEC 0.50 -3.13 1.24 -6.12 -4.94 -3.89 -2.37 -5.68 I(1)
(0.9861) (0.0266) (0.9983) (0.0000) (0.0005) (0.0148) (0.3937) (0.0000)
INF -2.57 -3.98 -2.79 -5.80 -2.99 -3.97 -2.99 -5.78 I(0)
(0.1019) (0.0021) (0.0621) (0.0000) (0.1396) (0.0120) (0.1379) (0.0000)
RCSGR -1.24 -3.88 -1.91 -5.14 -1.70 -3.46 -3.91 -5.16 I(0)
(0.6560) (0.0029) (0.3257) (0.0000) (0.7469) (0.0481) (0.0141) (0.0002)
RIR -2.69 -4.29 -2.93 -13.37 -3.62 -4.28 -2.98 -13.32 I(0)
(0.0790) (0.0007) (0.0442) (0.0000) (0.0318) (0.0047) (0.1425) (0.0000)
LREER -2.98 -4.15 -2.26 -6.63 -2.87 -4.21 -2.15 -6.61 I(1)
(0.0398) (0.0012) (0.1872) (0.0000) (0.1751) (0.0057) (0.5138) (0.0000)
LMONG -3.89 -8.08 -3.18 -7.66 -3.89 -8.05 -3.20 -7.62 I(1)
(0.0028) (0.0000) (0.0230) (0.0000) (0.0149) (0.0000) (0.0896) (0.0000)
Source: Computed by authors using EVIEWS 9 Note: Probability values which depict the level of significance are put in parenthesis AGRR = agricultural sector
output, INDR= industrial sector output, SERR= industrial sector output, BNOP= nominal Brent crude oil price, COPA= nominal average world crude oil price, BROP = real
Brent crude oil price, COPAR = real World crude oil price, AGRL, agricultural land, AGRMEC= agricultural machinery, INF = inflation rate, RCSGR= real growth rate of
capital stock, RIR= real interest rate, LREER, the log of the real effective exchange rate, LMONG = log of money growth or broad money or money supply (M2)

Table 4: Causality Tests: Granger-causality and Toda-Yamamoto Causality Tests Results

Null Hypothesis Granger Causality Test Toda-Yamamoto
BNOP does not cause AGRR 0.52 (0.7196) 2.09 (0.7195)
AGRR does not cause BNOP 0.38 (0.8208) 1.53 (0.8213)
COPA does not cause AGRR 0.55 (0.6961) 2.22 (0.6958)
AGRR does not cause COPA 0.42 (0.7944) 1.68 (0.7948)
BROP does not cause AGRR 0.74 (0.5677) 2.95 (0.5658)
AGRR does not cause BROP 0.41 (0.7990) 1.65 (0.7993)
COPAR does not cause AGRR 0.73 (0.5700) 2.94 (0.5682)
AGRR does not cause COPAR 0.49 (0.7404) 1.97(0.7404)
BNOP does not cause INDR 2.38 (0.0547) 9.53 (0.0490)
INDR does not cause BNOP 1.30 (0.2731) 5.21 (0.2669)
COPA does not cause INDR 2.58 (0.0401) 10.34 (0.0350)
INDR does not cause COPA 1.19 (0.3189) 4.76 (0.3133)
BROP does not cause INDR 1.84 (0.1255) 7.35 (0.1184)
INDR does not cause BROP 1.09 (0.3597) 4.40 (0.3547)
COPAR does not cause INDR 1.92 (0.1107) 7.69 (0.1037)
INDR does not cause COPAR 0.99 (0.4155) 3.96 (0.4113)
BNOP does not cause SERR 0.42 (0.7940) 1.68 (0.7944)
SERR does not cause BNOP 1.25 (0.2945) 4.99 (0.2886)
COPA does not cause SERR 0.44 (0.7761) 1.78 (0.7763)
SERR does not cause COPA 1.25 (0.2931) 5.00 (0.2872)
BROP does not cause SERR 0.53 (0.7155) 2.11 (0.7153)
SERR does not cause BROP 0.24 (0.9151) 0.96 (0.9156)
COPAR does not cause SERR 0.53 (0.7147) 2.12 (0.7145)
SERR does not cause COPAR 0.21 (0.9303) 0.86 (0.9308)
Source: Computed by authors using EVIEWS 9 Note: Probability values which depict the level of significance are put in parenthesis AGRR = agricultural sector output, INDR=
industrial sector output, SERR= industrial sector output, BNOP= nominal Brent crude oil price, COPA= nominal average world crude oil price, BROP = real Brent crude oil
price, COPAR = real World crude oil price, AGRL, agricultural land, AGRMEC= agricultural machinery, INF = inflation rate, RCSGR= real growth rate of capital stock, RIR=
real interest rate, LREER, log of the real effective exchange rate, LMONG = log of money growth or broad money or money supply (M2)
4.4. Co-integration Test Results

The results of bounds testing approach to cointegration tests are presented in Table 5. Three
maximum lag length periods are chosen using both the Akaike Information Criterion (AIC) and
Schwarz Criterion (SC) to get the optimal orders for each model. In both linear and nonlinear
ARDL, the decision to determine the existence of co-integration among the variables is based on
whether the value of the computed F-statistic obtained falls between or outside the lower and
upper bound provided in Pesaran et al. (2001) Table. If the value of the F-statistic computed falls
within the lower and upper bound criteria values, then no decision will be made about the
existence of cointegration. Also, if the computed F-statistic value falls below the lower bound
criteria value, there is no existence of co-integration. Co-integration therefore only exists when
the computed F-statistic value falls outside the upper bound criteria value in the Pesaran et al.
Table. Based on the information provided above, the results show that, as presented in Table 5,
either for the linear or nonlinear model and irrespective of the sectors (industrial, agricultural or
services sectors) and the crude oil prices (nominal or real Brent and World crude oil prices)
considered, there is the existence of co-integration among the variables.

4.5. ARDL and NARDL Results for Short-run and Long-run Effects of Nominal/Real Oil
Price Fluctuations on Sectoral Outputs

This section focuses on the presentation of the results of linear and nonlinear short-run and
longrun effects of nominal oil price fluctuations on sectoral outputs. The results of the findings
are presented in Table 6. Beginning with the linear ARDL results, the results show that in all the
sectoral models (industry, agriculture and services) considered, there exists convergence between
measures of the crude oil prices (Nominal Brent and World Oil prices) and the sectoral outputs.
This is because the coefficient of the error correction model, implemented via the ARDL error
correction framework, is negative and statistically significant. Specifically, the coefficients of
error correction terms, considering nominal Brent crude oil price, are -0.29, -0.25 and -0.18 for
the agricultural sector, industrial sector and services sector respectively. Similar coefficients of
error correction terms were also obtained when considering world oil prices except for the
services sector model where the coefficient of the error term is -0.19. The values of coefficients of
error correction terms are relatively low, showing that about 29%, 25% and 18% of
disequilibrium caused by past fluctuations in the economy, particularly in the agricultural sector,
industrial sector and services sector are corrected for and adjusted towards the long-run
equilibrium point in the current year.

In the case of nonlinear ARDL, there is also evidence of convergence towards the long-run
equilibrium as shown by the coefficients of error correction terms. However, the speeds of
adjustment in the case of nonlinear models are very fast compared with linear models. For
instance, considering the nominal Brent crude oil price, the coefficients of error terms for the
agricultural sector, industrial sector and services sector are -0.32, -0.27 and -0.18 respectively. As
can be seen, only the services sector has a coefficient of error similar to that of the linear model.
For the nominal world oil price, the coefficients of error terms are -0.32, -0.29 and -0.18 for the
agricultural sector, industrial sector and services sector respectively.

Beginning from the ARDL results as presented in Table 6, in the short-run, particularly in the
case of nominal Brent oil price, it is observed that oil price fluctuation has a negative insignificant
effect on agricultural output and, a positive significant effect on industrial and services sectors’
outputs. Specifically, holding other variables constant, an increase in the price of crude oil by 1
percent reduces agricultural output by 2.40 percent and increases the industrial and services
sectors by 4.28 and 0.39 percent respectively. The same findings are observed, particularly in
terms of signs, when the world oil price fluctuates, albeit with different magnitudes in the
agricultural and services sectors. This implies that an increase in the prices of crude oil is
detrimental to the agricultural sector output and beneficiary to the industrial and services sector's
outputs in the short run. In the long-run, however, oil price fluctuation does not only negatively
affect the agricultural sector output but also the industrial sector output. Only the services sector
continues to benefit from the increase in the price of crude oil. The finding, particularly the
negative relationship between oil price and agricultural output, aligns with the findings of
Binuomote and Odeniyi, (2013); Ikram and Waqas, (2014) and Aimer, (2016). Specifically, the
findings from the studies by these authors showed that oil price shocks have a negative effect on
agricultural output in Nigeria, Pakistan and Tunisia respectively. Similarly, our long-run results
for the industrial sector are akin to the findings by Fukunaga, Hirakata, and Sudo, (2010) and
Bayal and Kilie (2014) who found a negative relationship between oil price shocks and industrial
output in the US and Japan.

For robustness, we also considered the effects of real oil prices (Brent and World Oil Prices) on
the sectoral outputs (albeit, the table of the results is not presented here), the findings in the case
of the agricultural sector, both in the short-run and in the long-run, are similar to the one obtained
above. With regard to the industrial sector, oil price shocks have positive impacts on output in
both runs, albeit the long-run effect is not statistically significant. This is, however, different from
the long-run effect obtained when nominal oil prices were used, where oil prices negatively affect
industrial output in the long-run. Contrary to the positive effects of nominal oil price shocks on
services output, the negative effects of real oil price shocks are observed in both runs.
On the effect of other explanatory variables on the agricultural output, both in the short-run and
the long-run, it is found that agricultural land and machinery positively affect agricultural output,
albeit the effect is not statistically significant in both runs. The real effective exchange rate has a
negative effect on the agricultural sector's output in the short-run and the long-run. While the real
interest rate is negatively related to the agricultural sector in the short-run the relationship turns
positive in the long-run.

With regard to the industrial sector, capital stock and money growth (money supply) have short-
run positive effects on the industrial output, however, these positively significant effects turn out
to be negatively insignificant in the long-run. This can be intuitively interpreted as excessive
money supply is detrimental to industrial output growth and that acquisition of capital stock
beyond the capacity utilisation of firms is equally detrimental to industrial sector output in the
long-run. Similarly, real effective exchange rate and real interest rate have a positive and
significant impact on the industrial output in the short-run, which later turns to negative effects in
the long-run. This is interpreted thus: a short time appreciation of the exchange rate and a rise in
interest rate are indispensable to industrial sector output growth. However, a long-term
appreciation in the exchange rate and a rise in interest rate are disastrous to the industrial sector
output growth. An increase inflation rate is not favourable to the industrial sector output growth in
both the short-run and the long-run.

In the case of the services sector, the real effective exchange rate, real interest rate and inflation
rate have negative effects on the services sector output in the short-run. However, in the long-run,
only a real effective exchange rate continues to have a negative impact on the services sector
output. The real interest rate and inflation have positive effects on services sector output in the
long-run.

Turning to nonlinear ARDL models for all the sectors under consideration, we carried out the
long-run and short-run asymmetric tests using the stepwise regression method and Wald-test
coefficient restrictions. The results show that in the case of the agricultural output-oil price nexus
model, there is neither short-run nor long-run asymmetric whether the Brent or World oil price is
used. For the industrial output-oil price nexus model, it is found that there is the existence of both
short-run and long-run asymmetric in both Brent and World oil prices. This suggests that the
relationship between industrial output and oil price is characterised by asymmetry over time.
These results are in line with the findings by Mehrara and Sarem (2009), Guidi (2010), Herrera,
Lagalo, and Wada (2011), Riaz, Sial, and Nasreen (2016) and Liew and Balasubramanian (2017)
who documented asymmetric relationship between oil price shocks and industrial or
manufacturing output. The same results are obtained in the case of services output-oil price nexus
model.

Considering the agricultural sector, the short-run results indicate that a positive nominal Brent oil
price shock has a significantly negative effect on agricultural output while the negative oil price
shock has a positively insignificant effect on agricultural output. However, it is also observed that
negative oil price shocks lagged in one period have a negatively significant impact on agricultural
output. In the long run, positive oil price shocks (Brent or World Oil Prices) continue to have a
negative effect on agricultural output while negative oil price shocks also exhibit an insignificant
positive effect on agricultural output (see Table 6). Comparing these findings with the case of real
oil price shocks, it is observed the results remain unchanged. Thus, irrespective of the measures of
oil price shocks (nominal or real), oil price shocks are hazardous to agricultural sector output in
Nigeria.

As shown in Table 6 as well, a positive oil price shock (Brent) has an insignificant negative effect
on industrial output while a negative oil price shock (Brent) has a significant positive effect on
industrial output in both runs. The same results were recorded when the real oil price shock
(Brent) was considered. This connotes that a negative oil price shock, such as a sudden reduction
in the price of crude oil is a blessing to industrial sector output. However, when the world oil
price shock is considered, mixed findings were observed. In the case of nominal oil price shocks,
oil price shock now has a positive effect, albeit it is not statistically significant. The negative oil
price shock continues to exert a significant positive effect on industrial output. In the long run,
both positive and negative Brent oil price shocks retain their sign effects on the industrial output,
but they are, however, not statistically significant. In the case of world oil price shock, both
positive and negative oil price shocks have insignificant positive effects on industrial output. With
regard to real oil price shocks, positive oil price shocks have negative impacts on industrial output
while negative oil price shocks have positive effects on the industrial output in both runs.

On the asymmetric effect of nominal oil price shocks on the services sector output, it is found that
a positive oil price shock (Brent) has an insignificant negative effect on services output while a
negative oil price shock exhibits a significant positive effect on the output of services sector in the
short-run. The same findings are discovered when the world oil price is considered. In the long-
run, however, both positive and negative oil price shocks have significantly positive effects on the
services sector output. In the case of real oil shocks, mixed results were also obtained. While both
positive and negative oil price shocks result in a decline in the output of the services sector in the
short-run, the long-run results show that positive oil price shocks have an increasing effect on
services output and negative oil price shocks have a reducing effect on the output of the services
sector.

Other explanatory variables included in each of the models exhibit different degrees of effects on
sectoral outputs both in the long-run and in the short-run. For instance, agricultural land and
agricultural machinery have positive effects on agricultural output in both runs. However, the real
effective exchange rate and real interest rate negatively impact agricultural output. The impacts
are not statistically significant except in the case of the real interest rate, which is statistically
significant in the short run. It can be submitted that agricultural land and agricultural machinery
are indispensable to the success of the agricultural sector, particularly crop production in Nigeria.
It is important also to note that the exchange rate fluctuation and a high interest rate are
detrimental to agricultural production.

For the industrial sector, it is found that the real effective exchange rate, real interest rate, money
growth (money supply) and capital stock have significant positive effects on industrial output in
the short-run. However, the inflation rate has a significantly negative effect on industrial output in
the short-run. In the long-run, all these variables mentioned above have a negative effect on
industrial output, albeit only real effective exchange rate and money growth are statistically
significant. This implies that moderate movement in the exchange rate, interest rate, money
supply and real capital stock are germane to industrial sector output growth in the short-run while
their persistent increase in the exchange rate, interest rate and money will not be helpful to
industrial sector output. More importantly, as in the case of the linear model, excessive
accumulation of capital stock above the capacity utilisation is detrimental to the success of the
industrial sector in the long-run. Rising inflation, either in the short run or in the long run, is
harmful to industrial output.

With regard to the services sector, the real effective exchange rate has a negative impact on the
output of the services sector both in the short-run and in the long-run. The real interest rate and
inflation have positive effects on services output in both runs, however, only the real interest rate
is statistically significant at the 10 percent level.

The post estimation results reveal that none of the models suffers heteroskedasticity and
autocorrelation problems based on the results obtained from the ARCH LM heteroskedasticity test
and Breusch-Godfrey Serial Correlation LM Test. Also, the Ramsey RESET test for a fit model
shows that the models are well specified. However, the models have the problem of non-
normality.
 Table 5: Linear and Nonlinear ARDL Approaches to Co-integration Test Results

Agricultural Sector (Nominal Brent Oil Price) Industrial Sector (Nominal Brent Oil Price) Services Sector (Nominal Brent Oil Price)
Linear Nonlinear Linear Nonlinear Linear Nonlinear
Test Statistic Value K Value K Value K Value K Value K Value K
F-statistic 7.09 5 6.72 6 7.08 6 6.74 7 5.27 4 4.13 5
Critical Values

Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper
Significance I(0) I(1) I(0) I(1) I(0) I(1) I(0) I(1) I(0) I(1) I(0) I(1)
10% 2.26 3.35 2.12 3.23 2.12 3.23 2.03 3.13 2.45 3.52 2.26 3.35
5% 2.62 3.79 2.45 3.61 2.45 3.61 2.32 3.50 2.86 4.01 2.62 3.79
2.5% 2.96 4.18 2.75 3.99 2.75 3.99 2.60 3.84 3.25 4.49 2.96 4.18
1% 3.41 4.68 3.15 4.43 3.15 4.43 2.96 4.26 3.74 5.06 3.41 4.68
Agricultural Sector (WA Nominal Oil Price) Industrial Sector (WA Nominal Oil Price) Services Sector (WA Nominal Oil Price)
F-statistic 7.08 5 6.82 6 7.05 6 6.85 7 5.34 4 4.22 5
10% 2.26 3.35 2.12 3.23 2.12 3.23 2.03 3.13 2.45 3.52 2.26 3.35
5% 2.62 3.79 2.45 3.61 2.45 3.61 2.32 3.50 2.86 4.01 2.62 3.79
2.5% 2.96 4.18 2.75 3.99 2.75 3.99 2.60 3.84 3.25 4.49 2.96 4.18
1% 3.41 4.68 3.15 4.43 3.15 4.43 2.96 4.26 3.74 5.06 3.41 4.68
Real Brent Oil Price Real Brent Oil Price Real Brent Oil Price

F-statistic 7.12 5 6.23 6 7.77 6 7.31 7 4.53 4 3.72 5

10% 2.26 3.35 2.12 3.23 2.12 3.23 2.03 3.13 2.45 3.52 2.26 3.35
5% 2.62 3.79 2.45 3.61 2.45 3.61 2.32 3.50 2.86 4.01 2.62 3.79
2.5% 2.96 4.18 2.75 3.99 2.75 3.99 2.60 3.84 3.25 4.49 2.96 4.18
1% 3.41 4.68 3.15 4.43 3.15 4.43 2.96 4.26 3.74 5.06 3.41 4.68
Real World Oil Price Real World Oil Price Real World Oil Price

F-statistic 7.12 5 6.32 6 8.42 6 4.80 7 4.53 4 3.68 5

10% 2.26 3.35 2.12 3.23 2.12 3.23 2.03 3.13 2.45 3.52 2.26 3.35
5% 2.62 3.79 2.45 3.61 2.45 3.61 2.32 3.50 2.86 4.01 2.62 3.79
2.5% 2.96 4.18 2.75 3.99 2.75 3.99 2.60 3.84 3.25 4.49 2.96 4.18
1% 3.41 4.68 3.15 4.43 3.15 4.43 2.96 4.26 3.74 5.06 3.41 4.68
Source: Computed by authors using EVIEWS 9
 Table 6: ARDL and NARDL Results for Short-run and Long-run Effects of Nominal Oil Price Fluctuations on Sectoral Outputs
Variable Agricultural Sector Industrial Sector Services Sector
Brent Nominal Oil Price (0.4329) (0.0060)
Linear Nonlinear D(AGM) 0.23 5.60
Coefficient Coefficient (0.7348) (0.0130)
D(LREER) -0.04 -0.50
D(AGP(-1)) 0.44 0.44 (0.9543) (0.5052)
(0.0000) (0.0000) D(RIR) -0.09 -0.12
D(AGP(-2)) 0.21 0.25 (0.0409) (0.0040)
(0.0108) (0.0028) D(INF)
D(IND(-1))
D(LMONG)
D(IND(-2)) WA Nominal Oil Price Brent Nominal Oil Price WA Nominal Oil Price Brent Nominal Oil Price WA Nominal Oil Price
Linear Nonlinear Linear Nonlinear Linear Nonlinear Linear Nonlinear Linear Nonlinear
D(SERV(-1)) Coefficient Coefficient Coefficient Coefficient Coefficient Coefficient
Coefficient Coefficient Coefficient Coefficient Cointegrating
D(SERV(-2)) Form
D(BNOP) 0.44 0.44
-0.89
(0.0000) (0.0000)
(0.3843)
0.22 0.25
D(BNOP)+ -2.40
(0.0109) (0.0027)
(0.0617)
0.48 0.44 0.47 0.46
D(BNOP)- 0.86
(0.0000) (0.0000) (0.0000) (0.0000)
(0.8054)
0.22 0.25 0.22 0.25
D(BNOP(-1))- -10.18
(0.0071) (0.0071) (0.0066) (0.0016)
(0.0083)
0.45 0.44 0.45 0.44
D(COPA)
(0.0000) (0.0000) (0.0000) (0.0000)
D(COPA)+ 0.17 0.17 0.18 0.17
(0.0544) (0.0644) (0.0504) (0.0621)
D(COPA)- 4.28 0.39
(0.0065) (0.0723)
D(COPA(-1))- -0.18 -2.27
D(AGL) (0.8305) (0.1357)
4.27 39.48 7.06 0.60
 (0.0447) (0.7807) (0.1825)
(0.0749) 1.08 6.78 0.67
3.72 (0.7581) (0.0038) (0.0519)
-10.37
(0.0072)
(0. 4.36 39.43
14 (0.4232) (0.0039)
69) 0.23 5.73
-0.91 (0.7342) (0.0091)
4.28 -0.03 -0.43 5.27 4.68 5.21 4.84 -0.55 -0.74 -0.56 -0.74
0.43 (0.9662) (0.5628) (0.0074) (0.0145) (0.0080) (0.0116) (0.0163) (0.0040) (0.0150) (0.0036)
(0.3840) -0.09 -0.13 0.05 0.06 0.05 0.05 -0.01 -0.01 -0.01 -0.01
(0.0574) (0.0403) (0.0036) (0.0746) (0.0201) (0.0725) (0.0652) (0.3697) (0.2873) (0.3669) (0.2817)
-2.71 -0.06 -0.03 -0.03 -0.10 -0.04 -0.05 -0.04 -0.05
0.23 (0.1676) (0.1310) (0.1621) (0.0736) (0.0979) (0.0376) (0.0969) (0.0428)
-2.10 0.70 0.47 0.70 0.32
(0.3113) (0.5033) (0.3061) (0.6422)

D(LMONG(- -1.21 -1.29 -1.22 -1.03

1)) (0.0690) (0.0566) (0.0666) (0.1185)
D(RCSGR) 2.75 3.30 2.79 2.63
(0.0019) (0.0003) (0.0017) (0.0037)
D(RCSGR(- -1.36 -1.33 -1.41 -1.22
1)) (0.1253) (0.1339) (0.1126) (0.1646)
Ect(-1) -0.29 -0.32 -0.29 -0.32 -0.25 -0.27 -0.25 -0.29 -0.18 -0.18 -0.19 -0.18
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
Long-run Coefficients

Constant -51.10 -583.71 -52.37 -581.27 23.40 42.43 23.93 43.73 12.32 28.03 11.79 27.93
(0.5431) (0.0050) (0.5326) (0.0031) (0.1316) (0.0012) (0.1253) (0.0006) (0.0806) (0.0008) (0.0900) (0.0006)
BNOP -3.09 -0.20 2.13
(0.3845) (0.9472) (0.0421)
BNOP+ -7.60 -0.65 -0.65 2.94
(0.0583) (0.8317) (0.8317) (0.0483)
BNOP- 5.42 0.48 0.48 3.41
(0.1961) (0.8843) (0.8843) (0.0782)
COPA -3.16 -0.29 2.27
(0.3843) (0.9261) (0.0304)
COPA+ -8.50 0.79 3.22
(0.0416) (0.7790) (0.0312)
COPA- 5.03 2.11 3.77
(0.2152) (0.4961) (0.0541)
AGL 14.90 124.94 15.20 123.55
(0.4260) (0.0044) (0.4160) (0.0027)
AGM 0.81 17.73 0.81 17.95
(0.7350) (0.0116) (0.7345) (0.0079)
LREER -0.15 -1.58 -0.11 -1.34 -2.41 -4.78 -2.46 -5.34 -3.00 -4.19 -2.99 -4.16
(0.9543) (0.5047) (0.9662) (0.5624) (0.1916) (0.0127) (0.1834) (0.0045) (0.0054) (0.0030) (0.0048) (0.0025)
RIR 0.09 -0.01 0.09 -0.02 -0.09 -0.02 -0.10 -0.03 0.08 0.09 0.01 0.10
(0.3193) (0.8619) (0.3204) (0.8050) (0.1966) (0.7688) (0.1887) (0.6012) (0.0991) (0.0849) (0.7980) (0.0763)
INF -0.11 -0.11 -0.11 -0.10 0.01 0.003 0.08 0.008
(0.1715) (0.1366) (0.1659) (0.1538) (0.8199) (0.9567) (0.0974) (0.8757)
LMONG -2.09 -3.09 -2.08 -2.81
(0.1252) (0.0174) (0.1293) (0.0253)
RCSGR -0.93 -0.31 -0.91 -0.54
(0.2505) (0.6976) (0.2652) (0.4636)
R2
0.8197 0.8339 0.8197 0.8351 0.8996 0.9048 0.8992 0.9042 0.9130 0.9156 0.9132 0.9156
(81.97%) (83.39%) (81.97%) (83.51%) (89.96%) (90.48%) (89.92%) (90.42%) (91.30%) (91.56%) (91.32%) (91.56%)
Adj R2 0.8069 0.8178 0.8069 0.8191 0.8861 0.8902 0.8858 0.8895 0.9068 0.9082 0.9071 0.9083
(80.69%) (81.78%) (80.69) (81.91%) (88.61%) (89.02%) (88.58%) (88.95%) (90.68%) (90.82%) (90.71%) (90.63%)
F-test 64.1513 51.88 64.15 52.31 67.14 62.28 66.92 61.85 148.06 123.26 148.55 123.41
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
DW 1.98 1.98 1.98 1.97 2.04 2.03 2.03 2.06 1.97 1.97 1.97 1.99
omputed by authors using EVIEWS 9 Note: Probability values which depict the level of significance are put in parenthesis

Diagnostic Test

Wald-Test SR 1.13 1.14 4.71 4.46 3.13 3.53

(0.3448) (0.3373) (0.0038) (0.0052) (0.0794) (0.0624)
Wald-Test LR 0.83 0.66 3.62 3.38 9.07 9.09
(0.4378) (0.5142) (0.0297) (0.0372) (0.0002) (0.0002)
Jarque-Bera 7872.48 5029.88 7856.47 4851.91 109.39 91.60 107.86 121.91 799.44 886.17 787.63 889.30
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
B-G SCLM 0.13 0.10 0.14 0.12 0.36 0.32 0.27 0.68 0.10 0.003 0.09 0.009
(0.7153) (0.7498) (0.7085) (0.7296) (0.5486) (0.5752) (0.6060) (0.4098) (0.7489) (0.9572) (0.7686) (0.9247)
HT ARCH 0.08 0.08 0.08 0.08 0.11 0.03 0.11 0.14 0.22 0.08 0.22 0.08
(0.7810) (0.7845) (0.7803) (0.7768) (0.7396) (0.8580) (0.7411) (0.7079) (0.6372) (0.7778) (0.6363) (0.7752)
Ramsey Reset 0.72 1.93 0.73 1.97 0.23 0.46 0.25 0.07 6.60E-05 0.0001 8.01E-05 0.0007
(0.3985) (0.1677) (0.3930) (0.1633) (0.6341) (0.4989) (0.6167) (0.7985) (0.9935) (0.9912) (0.9929) (0.9793)
Dynamic Multiplier Effects of Oil Price Shocks on Sectoral Output in Nigeria

Figure 2 shows the dynamic multiplier effects of oil price shocks on sectoral outputs in Nigeria. In Figure 2,
there are three separate figures which represent the response of each sector’s output to oil price shocks.
Specifically, Figure 2.1 denotes the response of agricultural output to positive and negative oil price shocks
while Figure 2.2 represents the multiplier effect of oil positive and negative price shocks on the industrial
output. Figure 2.3 symbolises the response of services output to positive and negative oil price shocks.
According to Figure 2.1, when the oil price shocks occur, particularly around 2013 towards 2014, the negative
oil price shock initially has an increasing multiplier effect on the agricultural output which later dissipates into
a negative multiplier effect. The decline in output, however, never returns to the initial level of agricultural
output even though there is a sign of returning to its original level. The positive oil price shock, as depicted in
Figure 2.1, results in a decline in agricultural output, which never returns to the initial level of the output
before the shock occurred.
With regard to industrial sector output, Figure 2.2 shows that positive oil price shocks have no significant
multiplier effects on industrial output. However, the figure shows that the negative oil price shocks, which
occurred, particularly in the 3rd quarter of 2013 by 1 percent initially resulted in a decline in industrial sector
output in the 2nd quarter of 2014 by 11.7 percent. This implies that a 1 percent decline in oil price has about an
11.7 percent multiplier effect on industrial output. Later on, there was an appreciable surge in the industrial
output of about 3.4 percent in the 4th quarter of 2015 before the output returned to its initial level before the
shocks occurred in the 4th quarter of 2016.
In the case of the services sector, both positive and negative oil price shocks have initial reducing multiplier
effects on the output of the services sector in the 2 nd quarter of 2014. Thereafter, positive oil shocks lead to a
significant multiplier effect on the output, which increases output. However, it never converges to its initial
level. On the other hand, the negative oil price shocks continue to have a decreasing multiplier effect on the
services sector output, which also never returns to its original level (See Figure 2.3).
 Figure 2: Dynamic Multipliers Response of Sectoral Outputs to Nominal Oil Prices Shocks

BRENTOILPRI +1% COPA +1%

BRENTOILPRI -1% COPA -
Difference 1%
Difference

Figure 2.1: Dynamic Multiplier Response of the Agricultural Sector Output to Nominal Oil Price Shocks

Figure 2.2: Dynamic Multiplier Response of the Industrial Sector Output to Nominal Oil Price Shocks
BRENTOILPRI -1% COPA -1%

Difference Difference
 BRENTOILPRI +1% COPA -1%
BRENTOILPRI -1%

Difference
Difference

Figure 2.3: Dynamic Multiplier Response of the Services Sector Output to Nominal Oil Price Shocks
Conclusion and Policy Recommendation

This study examines the dynamic effects of oil price shocks on the sectoral outputs in Nigeria over the period of
1982 to 2016. To implement this objective, both linear and nonlinear econometric technique versions of ARDL were
used, with consideration of different oil prices both in nominal and real terms. The main findings from this study are
as follows:
First, it was found that irrespective of the measures of oil price shocks (Brent or World Oil prices), nominal oil price
shocks have a negative effect on agricultural output when the linear ARDL is used. In real terms, the results do not
change either as real oil price shocks still have negative effects on agricultural output. In the case of asymmetric
effects, it is discovered that positive nominal oil price shocks have negative effects on agricultural output both in the
short run and in the long run. The negative oil price shocks have positive effects on the agricultural output in both
runs. However, the positive effect in the long run is not statistically significant. The multiplier effect of oil price
shocks on agricultural output shows that when negative oil price shocks occur, it has an initial increasing multiplier
effect on the sector output. However, this initial increase dissipated out and the output never returned to its original
level before the shocks occurred. The multiplier effect also reveals that positive oil price shocks have a reducing
effect on agricultural output, which as in the case of negative oil price shocks, never returns to its original level
before the shocks took place. From these findings, it can be submitted that an oil price increase is deleterious to
agricultural sector output in Nigeria and vice versa in the case of an oil price decline.
Second, nominal oil price shocks have a positive effect on industrial output in the short run. The short-run positive
effect however turns negative in the long-run. In terms of real oil prices, the shocks have positive effects in both
runs, albeit the effect is statistically significant in the short run. The asymmetric findings show that nominal positive
oil shocks (Brent) exhibit insignificant negative effects on industrial sector output in the short run and the long run.
On the other hand, nominal negative oil price shocks (Brent) have positive effects on industrial output. When the
world oil price is considered, both positive and negative oil price shocks have an insignificant positive impact on
industrial output. In real terms, positive oil price shocks have negative impacts on industrial output while negative
oil price shocks have positive effects on industrial output in both runs. On the multiplier effects of oil price shocks
on industrial output, it is found that positive oil price shocks do not have appreciable multiplier effects on industrial
output. The negative oil price shocks initially had reducing effects on the industrial output, which after the initial
decline, recovered to have increasing multiplier effects before the output returned to its original level. Thus, negative
oil price shocks will lead to industrial sector growth.

Third, it was also found that oil nominal price shocks have positive effects on services sector output in both runs.
Conversely, real oil price shocks have a negative impact on services sector output. With regard to the asymmetric
effects of oil price shocks, it is discovered that positive nominal oil price shocks have an insignificant negative effect
on services sector output while negative nominal price shocks have a significant negative effect on services sector
output. Using the world oil price, both positive and negative oil price shocks have significant positive effects on the
output of the services sector in the long run while the results of short-run remained unchanged. When the real oil
price shock is considered, the findings show that both positive and negative shocks have increasing effects on the
services sector in the short run. However, in the long run, positive oil price shocks increase services sector output
while negative oil price shocks result in a decline in services output. For the multiplier effect, both positive and
negative oil price shocks initially have a reducing multiplier effect on the output in the 2 nd quarter of 2014.
Thereafter, positive shocks lead to a surge in output while negative shocks lead to a fall in output. However, output
never converges to its original level before the shocks.

The important policy implications emanating from this study are highlighted as follows:

(1) Since the agricultural sector appears to be negatively affected by the increase in oil prices, the government
should establish special funds from excess revenue that may be realised from the increasing oil prices to cater
for the farmers affected by oil price hikes.
(2) For the industrial and services sectors, favourable policies such as low-interest rates and stable exchange rate
management must be pursued by the government to encourage the firms in the sector. This must also be
accompanied by sound fiscal policy in terms of a favourable tax system.
(3) Alternative sources of energy; all forms of renewable energy such as solar, wind and biofuel should be
prioritised by the government to improve electricity generation and supply to lessen the burden of dependence
on oil production such as diesel and premium motor spirit (PMS).
(4) The government should give economic diversification the utmost attention to minimise the negative impact of
the increase in oil prices on all sectors.
References

Abdulkareem, A., and Kilishi, A. A. (2016). Analysis of oil price-macroeconomic shocks in Nigeria. CBN Journal
of Applied Statistics, Vol 7, Issue 4, 1-22.

Aimer, N. M. (2016). The effects of price shocks on the economic sectors of Libya. International Journal of
Business and Social Research, Vol. 06, Issue 12, 13-24.

Akpan, E. O. (2009). Oil price shocks and Nigeria's macroeconomy. paper presented at the Annual Conference of
CSAE Conference, Economic Development in Africa, 22nd-24th March, Oxford.

Al-Risheq, S. M. (2016). The impact of oil prices on industrial production in developing countries. University of
Ottawa, Ottawa, Ontario

Aminu, A., and Raifu, I. (2018). Dynamic Nexus between Government Revenues and Expenditures in Nigeria:
Evidence from Asymmetric Causality and Cointegration Methods. University Library of Munich, Germany.

Aminu, A., and Raifu, I. A. (2019). Dynamic nexus between oil revenues and economic growth in
Nigeria. Economics Bulletin, 39(2), 1556-1570

Aminu, A., Raifu, I. A., and Oloyede, B. D. (2019). Financial Development and Manufacturing Output Growth
Nexus in Nigeria: The Role of Institutional Quality. DBN Journal of Economics and Sustainable Growth, 1(2), 1-
43.

Ayadi, O. F. (2005). Oil price fluctuations and the Nigerian economy. OPEC Energy Review, Vol. 29, Issues 3,
199217.

Baggs, J., Beaulieu, E., and Fung, L. (2010). Are service firms affected by exchange rate movements?. Review of
Income and Wealth, 56, S156-S176.

Barro, R. J. (1984). Macroeconomics. New York, NY: John Wiley and Sons.

Barsky, R. B., and Kilian, L. (2004). Oil price and the macroeconomy since the 1970s. Quarterly Journal of
Economics, 98(1), 115-134.

Baumeister, C., and Kilian, L. (2013). Do oil price increases cause higher food prices? Bank of Canada Working
Paper No. 52, 1-73.

Bayar, Y., and Kilic, C. (2014). Effects of oil price and natural gas prices on industrial production in the Eurozone
member countries. International Journal of Energy Economics and Policy, Vol. 4, No. 2, 238-247.

Bertola, G., and Caballero, R. J. (1994). Irreversibility and aggregate investment. The Review of Economic Studies,
61(2), 223-246.

Binuomote, S. O., and Odeniyi, K. A. (2013). Effect of Crude Oil price on agricultural productivity in Nigeria.
International Journal of Applied Agricultural and Apicultural Research, 9(1and2), 131-139.

Bolaji, B. O., and Bolaji, G. A. (2010). Investigating the effects of increase in oil prices on manufacturing companies
in Nigeria. Pacific Journal of Science and Technology, 11(2), 387-390.

Bollerslev, T. (1986). Generalised autoregressive conditional heteroskedasticity. Journal of Econometrics, 31(3),

307327.
Borenstein, S. A., Cameron, C., and Gilbert, R. (1997). Do gasoline prices respond asymmetrically to crude oil price
changes. The Quarterly Journal of Economics, 112(1), 305-339.

Brown, S., and Yucel, M. (2002). Energy prices and aggregate economic activity: an interpretative survey. Quarterly
Review of Economics and Finance, Vol. 42(2), 193-208.

Burbridge, J., and Harrison, A. (1984). Testing for the effect of oil price rises using Vector Autoregressions.
International Economic Review, Vol. 25, No. 2, 429-486.

Catik, A. N., and Onder, A. O. (2013). An asymmetric analysis of the relationship between oil price shocks and
output: The case of Turkey. Economic Modelling, 26, 884-892.

Chuku, A. C. (2012). Linear and asymmetric impacts of oil price shocks in an oil-importing and exporting economy:
The case of Nigeria. OPEC Energy Review, Vol. 36, Issue 4, 413-443.

Cologne, A., and Manera, M. (2009). The asymmetric effects of oil price shocks on output growth: A
Markovswitching analysis for G-7 countries. Economic Modelling 26, 1-29.

Copiello, S. (2016). Economics implications of the energy issue: Evidence for a positive non-linear relation between
embodied energy and construction cost. Energy and Buildings, 123, 59-70.

Dhuyvetter, K., and Kasterns, T. (2005). Impact of rising diesel prices on machinery and whole-farm cost.
Manhattan, Kansas, Kansas State University, (Accessed).

Drobyshevsky, S., Bozhechkova, A., Trunin, P., and Sinelnikova-Muryleva, E. (2017). The Effect of Interest Rates
on Economic Growth. Gaidar Institute for Economic Policy.

Du, L., Yanan, H., and Wei, C. (2010). The relationship between oil price shocks and China's macro-economy: An
empirical analysis. Energy Policy, 38, 4142-4151.

Ee, C. Y., Gugkang, A. S., and Husin, N. H. (2015). The effect of oil price in Malaysia economy sectors. Labuan
Bulletin of International Business and Finance, 13, 1-9.

El-anshasy, A., Bradley, A. D., and Joutz, F. L. (2005). Evidence on the role of oil prices in Venezuela's economic
performance. Working Paper, George Washington, University.

Eski, I. H., Izgi, B. B., and Senturk, M. (2011). Reconsidering the relationship between oil prices and industrial
production: Testing for co-integration in some of the OECD countries. Eurasian Journal of Business and
Economics, 4(8), 1-12.

Ezeaku, H. C., Ibe, I. G., Ugwuanyi, U. B., Modebe, N. J., and Agbaeze, E. K. (2018). Monetary policy transmission
and industrial sector growth: Empirical evidence from Nigeria. Sage Open, 8(2), pp. 1-12

Fukunaga, I., Hirakata, N., and Sudo, N. (2010). The effects of oil price changes on the industry-level production and
prices in the US and Japan. NBER Working Paper 15791, 1-23.

Gisser, H., and Godwin, T. (1986). Crude oil and the macroeconomy: Tests of some popular notions. Journal of
Money, Credit and Banking, 18, 95-103.

Guidi, F. (2010). The economic effects of oil prices shocks on the UK mafacturing and services sectors. IUP Journal
of Applied Economics, Vol. 9, Issue 4, 5-34.

Hamilton, J. D. (1983). Oil and the macroeconomy since World War II. Journal of Political Economy, 91(3),
593248.
Hamilton, J. D. (1996). This is what happened to the oil price-macroeconomy relationship. Journal of Monetary
Economics, 38(2), 215-220.

Hanson, K.; Robinson, S.; Schluter, G. (1993). Sectoral effects of a world oil price shock: Economicwide linkages to
agriculture. Journal of Agricultural Resource Economics, 18(1), 96-116.

Herrera, A. M., Lagalo, L. G., and Wada, T. (2011). Oil price shocks and industrial production: Is the relationship
linear? Macroeconomic Dynamics, 15(3), 472-497.

Huang, B.-N., Hwang, M. J., and Peng, H. (2005). The asymmetric of the impact of price shocks on economic
activities: An application of the multivariate threshold model. Energy Economics, 27, 455-476.

Huntington, H. G. (1998). Crude oil price and U.S. economic performance: Where does the asymmetric reside? The
Energy Journal, Vol. 19, No. 4, 107-132.

Ikram, H., and Waqas, M. (2014). Crude oil price and agricultural productivity growth in Pakistan. World Applied
Sciences Journal, 32(4), 642-649.

Iwayemi, A., and Fowowe, B. (2011b). Impact of oil price shocks on selected macroeconomic variables in Nigeria.
Energy Policy, 39, 603-612.

Iwayemi, A., and Fowowe, B. (2011a). Oil and the macroeconomy: empirical evidence from oil-exporting African
countries. OPEC Energy Review, Vol. 35, Issue 3, 227-269.

Jimenez-Rodriguez. (2007). The industrial impact of oil shocks: Evidence from the industries of siX OECD
countries. Banco de Espana, Documentis de Trabajo, No. 0731, 1-51.

Jones, D. W., and Leiby, P. N. (1996). The macroeconomic impacts of price shocks: A review of literature and issue.
Tennessee: Energy Division, Oak Ridge National Laboratory.

Kandil, M., and Mirazaie, I. (2005). The Effects of Exchange Rate Fluctuations on Output and Prices: Evidence from
Developing Countries. The Journal of Developing Areas, 38(2), 189-219.

Khatkhate, D. R. (1988). Assessing the impact of interest rates in less developed countries. World Development,
16(5), 577-588.

Kilian, L. (2008). The economic effects of energy price shocks. Journal of Economic Perspective, 46(4), 871-909.

Kliensen, K. L. (2006). Rising natural gas prices and real economic activity. Federal Reserve Bank of St. Louis
Review, 88(6), 871-909.

Lee, K., and Ni, S. (2002). On the dynamic effects of oil price shocks: A study using industry level data. Journal of
Monetary Economics, 49, 823-852.

Lee, K., Ni, S., and Ratti, R. A. (1995). Oil shocks and the macroeconomy: The role of price variability. The Energy
Journal, Vol. 16, No. 4, 39-56.

Lescaroux, F., and Mignon, V. (2008). On the influence of oil prices on the economic activity and other
macroeconomic and financial variables. OPEC Energy Review, Vol. 32, Issue 4, 343-340.

Liew, V. K., and Balasubramanian, A. (2017). Oil price shocks and sectoral outputs: Empirical evidence from
Malaysia. Economics Bulletin, Vol. 37, Issue 1, 38-47.

Mahboub, A. A., and Ahmed, H. E. (2017). The effect of oil price shocks on the Saudi manufacting sector.
Economics World, 5(3), 230-238.
Mehrara, M., and Sarem, M. (2009). Effects of oil price shocks on industrial production: Evidence from some
oilexporting countries. OPEC Energy Review, Vol. 33, Issue 3-4, 170-183.

Mlambo, C. (2020). Exchange rate and manufacturing sector performance in SACU states. Cogent Business &
Management, 7(1), 1787735.

Mork, K. N. (1989). Oil and the macroeconomy when prices go up and down: An extension of Hamilton's Results.
Journal of Political Economy, Vol. 97, No. 3, 740-744.

Mory, J. F. (1993). Oil prices and economic activity: is the relationship symmetric?. The Energy Journal, 14(4),
151163.

Moshiri, S. (2015). Asymmetric effects of oil price shocks in oil-exporting countries: The role of Institutions. OPEC
Energy Review, Vol. 39, Issue 4, 222-246.

Nigatu, G., Hjort, K., Hansen, J., and Somwanu, A. (2014). The impacts of energy price on global agricultural
commodity supply. Selected paper presented for presentation at the Agriculture ad Applied Economics
Association's 2014 AAEA Annual Metting, Minneapolis, MN, July 27-29, 2014, 1-33.

Nkang, N. M. (2018). Oil price shocks, agriculture and household welfare in Nigeria: Results from an economy-wide
model. Central Bank of Nigeria Publication, 1-37.

Nnadozie, O., Emediegwu, L. E., and Raifu, I. A. (2022). Examining the Response of Inflation to the Boom-Bust
Cycle of Oil Price: Evidence from Nigeria. Available at SSRN 4152273.

Ojiya E.A., Okoh S.A., Mamman A.B., and Chukwuemeka N. J. (2017). An Empirical Analysis of the effect of
Agricultural Input on Agricultural Productivity in Nigeria. Int J Agric Sc Food Technol 3(4): 077-085

Okonjo-Iweala, N., and Osafo-Kwaako, P. (2007). Nigeria's Economic Reforms: Progress and Challenges.
Washington, DC: Brookings Global Economy and Development.

Okoye, P. U., Mbakwe, C. C., and Igbo, E. N. (2018). Modelling the construction sector and oil prices toward the
growth of the Nigerian Economy: An econometric approach. Economies, 6, 16, 1-19.

Oladosu, G. A., Leiby, P. N., Bowman, D. C., Uria-Martinez, R., and Johnson, M. M. (2018). Impact of oil price
shocks on the United States Economy: A meta-analysis of the oil price elasticity of GDP of net-importing
economies. Energy Policy, 115, 523-544.

Olomola, P. A. (2006). Oil price shock and aggregate economic activity in Nigeria. African Economic and Business
Review, Vol. 4, No. 2, 48-61.

Pesaran, M. H., Shin, Y., and Smith, R. J. (2001). Bounds testing approaches to the analysis of level relationships.
Journal of applied econometrics, 16(3), 289-326.

Pindyck, R. (1991). Irreversibility, uncertainty and investment. Journal of Economic Literature, Vol. 29, No. 3,
110148.

Rahman, S., and Serletis, A. (2010). The asymmetric effects of oil price shocks and monetary policy shocks: A
nonlinear VAR approach. Energy Economics 32, 1460-1466.

Raifu, I. A. (2021). The Role of Institutional Quality in Oil Price-Unemployment Nexus in African and Asian Oil
Exporting Countries.
Raifu, I. A. (2023c). Examining the Time-Varying Causality Between Oil Returns and Stock Returns in
Norway. Energy RESEARCH LETTERS, 4(1), 1-9.

Raifu, I. A. (2023a). The Effect of Oil Revenue on Economic Growth in Nigeria: Evidence from Quantile
Regression Method. DBN Journal of Economics and Sustainable Growth,

Raifu, I. A. (2023b). Are All Shocks Alike? Evidence from the Effects of Oil Shocks on Military Expenditure in
Nigeria. Energy Research Letter

Raifu, I. A., Aminu, A., and Folawewo, A. O. (2020). Investigating the relationship between changes in oil prices
and unemployment rate in Nigeria: linear and nonlinear autoregressive distributed lag approaches. Future
Business Journal, 6, 1-18..

Raifu, I. A., and Afolabi, J. (2023). Simulating the Inflationary Effects of Fuel Subsidy Removal in Nigeria:
Evidence from a Novel Approach. Available at SSRN 4636464.

Raifu, I. A., and Aminu, A. (2019). Financial development and agricultural performance in Nigeria: what role do
institutions play?. Agricultural Finance Review, 80(2), 231-254.

Raifu, I. A., and Aminu, A. (2020). A reconsideration of Wagner's hypothesis for Nigeria using linear and nonlinear
ARDL methods. International Journal of Sustainable Economy, 12(4), 321-343.

Raifu, I. A., and Aminu, A. (2021). IMPACT OF OIL REVENUE ON UNEMPLOYMENT IN NIGERIA. The
Journal of Energy and Development, 47(1/2), 101-134.

Raifu, I. A., and Oshota, S. O. (2023). Re-examining oil price-stock market returns nexus in Nigeria using a two
stage Markov regime switching approach. International Journal of Energy Sector Management, 17(3), 489-
509.

Raifu, I. A., and Raheem, A. N. (2018). Do Government Revenues Matter for Economic Growth? Evidence from
Nigeria. European Journal of Government and Economics (EJGE), 7(1), 60-84.

Rasche, R. H., and Tatom, J. A. (1977). The effects of the new energy regime on economic capacity, production and
prices. Federal Reserve Bank of St. Loius Review, 59(4), 2-12.

Rezazadehkarsalari, A., Haghiri, F., and Behrooznia, A. (2013). The effects of oil price shocks on real GDP in Iran.
African Journal of Business Management, Vol. 7(33), 3220-3232.

Riaz, M. F., Sial, M. H., and Nasreen, S. (2016). Impact of oil price shocks on manufacturing production of Pakistan.
Bulletin of Energy Economics, 4, 23-34.

Sadeghi, A. (2017). Oil price shocks and economic growth in oil-exporting countries: Does the size of government
matter? IMF Working Paper, WP/17/287, 1-27.

Sardosky, P. . (1999). Oil price shocks and stock market activity. Energy Economics, 2, 449-469.

Seth, D., Giridhar, B. S., and Krishnaswami, S. (2016). Impact of crude oil price changes on select Industries.
International Journal of Social Sciences and Management, Vol. 3, Issue 2, 87-92.

Shaari, M. S., Pei, T. L., and Rahim, H. A. (2013). Effects of oil price shocks on the economic sectors in Malaysia.
International Journal of Energy Economics and Policy, Vol. 3, No. 4, 360-366.
Shin, Y., Yu, B., and Greenwood-Nimmo, M. (2014). Modelling asymmetric co-integration and dynamic multipliers
in a nonlinear ARDL framework. In W. H. (Eds), The festchrift in honor of Peter Schmidt: Econometric
Methods and applications (pp. 281-314). New York: Springer.

Tefera, B., Worku, F., and Ayalew, Z. (2012). Implications of Oil Price Shocks and Subsidizing Oil Prices to the
Ethiopian Economy: A CGE Analysis (No. 008). Ethiopian Development Research Institute.

Tokgoz, S. (2009). The impact of energy markets on the EU agricultural sector. Center for Agricultural and Rural
Development, Working Paper 09-WP 485, 1-24.

Torul, O., and Alper, C. E. (2010). Asymmetric effects of oil prices on the manufactuTTeferaring sector in Turkey.
Review of Middle East Economics and Finance, Vol 6, Number 1, 1-18.

Walker, I. O., and Ekpemauzor, D. (1990). The impact and prospects of oil wealth in an open monocultural
economy: Nigeria, a case study. OPEC Energy Review, Vol. 14, Issue 1, 23-43.

View publication stats