INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 331
Asymmetric Effects of Natural Gas Consumption by GENCOs on
the Nigerian Economy: A Nonlinear ARDL Approach
Ohabuenyi Edwin Onah
1
, Ijeoma Emele Kalu
2
, Osokogwu Uche
3
1
Emerald Energy Institute, University of Port Harcourt, Nigeria
2
Department of Economics, University of Port Harcourt, Nigeria
3
Department of Gas Engineering, University of Port Harcourt, Nigeria
DOI : https://doi.org/10.51583/IJLTEMAS.2024.131232
Received: 29 November 2024; Accepted: 07 December 2024; Published: 23 January 2025
Abstract: This research article examined the correlation between natural gas usage by power generation firms and economic
growth in Nigeria. The study analyzed many factors, including liquefied petroleum gas usage, industrial gas consumption, and gas
consumption for electricity production, and their effects on gross domestic product (GDP). Data on a quarterly basis from 2010 to
2020, sourced from reputable entities such as the Central Bank of Nigeria, Gas Exporting Countries Forum, and the United
Nations, were employed for the analysis. The study investigation utilized the Nonlinear Autoregressive Distributed Lag
(NARDL) approach. The results indicated that gas usage substantially affected Nigeria's GDP in short term and long run. The
research indicated that a 1% growth in gas utilization by electric power plants and LPG usage positively correlated with a 1.04%
and 16.64% rise in Nigeria's GDP, respectively. A 1% reduction in industrial gas usage would result in a 19.95% decline in
Nigeria's GDP. The findings demonstrated that augmenting gas use in power generating influenced the GDP. Based on the study,
it is advised that the Nigerian government undertake steps to increase the consumption of LPG and natural gas for industrial
applications and electricity generation to foster economic growth.
Keywords: GDP, NARDL, Natural gas consumption, Economic growth.
I. Introduction
Africa contributes less than 4% of the total global energy-related CO₂ emissions, notwithstanding the scenario applied. Under
current policies. The worldwide mean temperature is anticipated to increase by 2°C by 2050 (IEA Energy Transition Report,
2023). However, this increase would likely be higher in North Africa, with median temperatures climbing by 2.7°C, which could
lead to an estimated 8% reduction in African GDP by mid-century. Over 5,000 billion cubic meters (bcm) of natural gas have
been discovered in Africa yet remain undeveloped. These resources would provide an extra 90 bcm of gas per year by 2030,
which could be crucial for industries which include fertilizer production, steel, cement, and water desalination (Energy
Information Administration, 2018). Over the next 30 years, using these gas reserves could generate approximately 10 gigatonnes
of cumulative CO₂ emissions. If added to Africa’s current emissions, this would only increase its share of global emissions to
3.5% (IEA, 2022). This highlights the importance of developing these gas resources to support economic growth while aligning
with climate goals for a low-carbon future.
In Nigeria, where consistent electricity generation is essential for industrial expansion, employment creation, and general
economic stability, the power industry has important role to play in the development of the country's economy. Among the
different energy sources utilized in the nation to generate power, natural gas is the most common fuel. The Nigerian Electricity
Regulatory Commission (NERC) reports that natural gas powers the majority of the thermal plants run by generation companies
(GenCos), accounting for over 75% of the nation's electricity generation, with hydropower plants providing about 25% of the
total power generated (NERC, 2023). Figure 1 depicts the percentage contributions of the two energy sources (gas and hydro) to
Nigeria's overall energy generation from 1980 to 2015. All the same, Nigeria is still beset by severe power outages, and regular
failures of the national system result in large-scale financial losses. Due to inadequate gas supplies and inefficient infrastructure,
the nation's installed power capacity, which is expected to be 12,660 megawatts (MW), is drastically underutilized, with effective
generation hovering at 4,544 MW (NERC, 2023).
Figure 1: Electricity generation in Nigeria, 1980 – 2015; Source: Oyeleke and Akinlo (2019)
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 332
This inadequacy in electricity generation, primarily powered by natural gas, has had far-reaching consequences on Nigeria’s GDP
(Gross Domestic Product). Several studies have highlighted a strong correlation between consumption of energy and the growth
of the economy, especially in developing countries (IEA, Africa Energy Outlook, 2022). For Nigeria, where energy remains a key
input for industries and households, the consumption of natural gas by GenCos plays an integral role in driving productivity
across sectors (Malami et al., 2024). It is estimated that Nigeria loses approximately $25 billion annually, or about 6% of its GDP,
due to unreliable electricity (World Bank, 2021). About 55% of Nigerian citizens has access to the electricity grid, which can
meet just 30% of the country’s total electricity demand. The population of Nigeria is estimated at over 200 million citizens,
although its per capita electricity consumption is incredibly low at 151 kWh due to a combination of poor electrical generation
and a high population. The averages for Africa and sub-Saharan Africa are 550 kWh and 370 kWh, respectively, and this number
is much lower (Owebor et al., 2021). The economy is impacted asymmetrically by GenCo's struggles with irregular gas supply,
with periods of gas shortages leading to significant declines in GDP growth (Ummalla and Samal, 2019). Nigeria has the biggest
energy access deficit in the world, with 43% of Nigerians translating to 85 million citizens lack access to public electricity. This
has significant implications for economic growth.
Figure 2: Nigeria’s electricity access statistics; Source: World Bank, 2021
Figure 2: Nigeria’s unreliable electricity access statistics; Source: World Bank, 2021
As of 2018, Nigeria accounted for 29% of the oil reserves and 21% of the gas reserves on the continent, making it a significant
player in the energy industry (Udoudo et al, 2023). The production-to-reserve ratio of Nigeria is below 1%, the industry is still
mostly undeveloped even though the nation is ranked ninth in terms of gas reserve globally (Nwaoha and Wood, 2014). Nigeria is
number one in terms of gas reserve holder in Africa and the number three gas producer on the continent, with confirmed natural
gas reserves of 209.2 trillion cubic feet (tcf) as of January 1st, 2024, of which 139.4 tcf are considered recoverable (Obada et al.,
2024).
In light of this, it is critical to comprehend the asymmetric consequences that GenCos's natural gas consumption has on Nigeria's
economy. There are no studies that applied the nonlinear dynamics of gas consumption by power plants and its disproportionate
effects on GDP, despite previous research focusing on energy consumption and industrial production. Using NARDL (Nonlinear
Autoregressive Distributed Lag) approach, this research will investigate the effects of increase and decrease in GenCos' natural
gas consumption on Nigeria's GDP (Ozcan and Ozturk, 2019). This study will provide further insight into the degree to which
stabilizing the gas supply and reforming energy policies may support long-term economic development by capturing these
asymmetric impacts (Solarin and Ozturk, 2016). The study is particularly relevant in light of the Federal Government of Nigeria's
ongoing efforts to boost domestic natural gas usage under the "Decade of Gas" strategy. The link between GENCO gas
consumption and economic development may provide policymakers with valuable data to improve Nigeria's energy mix and
increase the efficiency of the power sector (Nwabueze and Joel, 2022).
A large portion of the literature on economics examined the causal relationship between growth of the economy and energy use.
There is still contention with regard to the nature of this link. While some empirical research see the opposite, others suggest a
unidirectional causal relationship between growth of the economy and power use. Furthermore, a few research has indicated no
causative relationship at all, showing a neutral association between energy consumption and economic growth, while other
studies have discovered evidence of a bidirectional causal link between the two variables (Akinlo, 2009).
A review of the literature that examines current theories and empirical research on the connection between energy use and growth
of the economy will come after this introduction. The NARDL strategy and the data sources used in the study will be covered in
the methodology section. The analysis's conclusions and suggestions will be presented in the section under "Results and
Discussion."
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 333
II. Literature Review
The literature review offers the basis for comprehending the overview of the generating firms (GenCos) in the electrical sector, as
well as the theoretical and empirical frameworks that underlie the unequal effects of GenCo's natural gas consumption on the
Nigerian economy. Using both global and regional viewpoints, this section reviews the literatures that dealt with the relationships
among the growth of the economy, energy consumption (especially natural gas consumption), and NARDL models. The literature
will also discuss how using natural gas has policy consequences and how it relates to economic outcomes and electricity
generation.
Overview of Power Generation in Nigeria:
Nigeria electricity industry was unbundled in year 3013 and its ownership is predominantly private. Nevertheless, the transition is
yet to deliver the anticipated outcomes. Following the enactment of the EPSRA (Electric Power Sector Reform Act) in 2004, the
electricity industry was broken into six companies that generate electricity (GenCos), eleven companies to distribute (DisCos),
and one Transmission Company (TCN) (World Bank, 2021). Figure 3 shows the Nigerian power sector, and its ownership after
its unbundling.
Figure 3: Nigeria's power sector, unbundled and privately owned. Source: World Bank: 2021
According to the 2023 report by the Nigerian Electricity Supply Industry (NESI), 27 grid-connected power plants are currently
operational with capacity of all installed generators 12,660 MW, but the average available capacity stands at 4,544.30 MW,
resulting in an overall plant availability factor of 35.90% as shown in Table 4.
Table 1: Plant Availability Factor (%) in 2023
In 2023, the national grid’s gross generation was 36,710.38 GWh, averaging 4,190.68 MWh per hour as shown in Figure 4.
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 334
Figure 4: Average Hourly Generation (MWh per hour) in 2023. Source: NERC 2023
Most of the generation is thermal based (gas), with hydropower contributing 24.75% (9,086.90 GWh) to total generation.
Mechanical outages were the primary cause of plant unavailability in 2023, affecting about 38.04% (4,802.80 MW) of total
installed capacity. The aging infrastructure—where the average plant age in NESI was 21 years by December 2023—along with
maintenance issues, were key factors in these outages. Additionally, liquidity challenges due to underpayment of invoices by
distribution companies (DisCos) and unpaid government subsidies created financial constraints for generation companies
(GenCos). Gas supply challenges also plagued thermal GenCos, arising from limited gas infrastructure and a lack of effective Gas
Supply Agreements (GSA). The figure below shows the effect that the electricity sector finances from 2020 data from NERC.
Figure 5: Financial Inefficiencies of Nigeria Electricity Sector. Source: World Bank, 2021
NERC, 2023, recommended that to resolve the challenges of GenCo’s and improve their liquidity level, there is need to enforce
prompt payment on DisCos and also timely payment of subsidy by the government. These actions would enable the GenCos meet
their financial responsibilities in executing capital projects and other maintenance jobs to improve their capacity and plant
reliability.
Theoretical Framework:
The following are some economic theories that are pertinent to this study to enable us comprehend the relationships existing
between natural gas consumption by GENCOs and its impact on Nigeria’s GDP:
Energy-Led Growth Hypothesis (ELGH):
This theory maintains that consumption of energy is a vital driver of the growth of the economy. Hence, energy serves as an input
for production processes, enabling increased productivity and output across industries. The works of Apergis and Tang, 2013,
Narayan and Prasad, 2008 and Ozturk, 2010 supports ELGH. Considering GenCos reliance on natural gas for power generation in
Nigeria, it is expected to directly impacts industrial performance and, by extension, the GDP of the country.
Institutional Theory:
Institutions are indispensable in the development of energy policy and its utilization. The significance of regulatory frameworks,
governance structures, and policy interventions in energy resource management is underscored by Institutional Theory. The
performance of GenCos in Nigeria are significantly influenced by government policies regarding gas pricing, investment
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 335
incentives and pipeline infrastructure. Inefficiencies in energy distribution and increased economic losses as a result of power
theft and nonpayment of bills can be the result of weak institutional frameworks (North, 2012). In Nigeria, the Nigerian Upstream
Petroleum Regulatory Commission (NUPRC) is the regulator of the upstream sector of the petroleum sector while the
downstream and midstream is regulated Nigerian Midstream Stream and Downstream Petroleum Regulatory Authority
(NMDPRA). The Nigerian National Petroleum Company Limited (NNPC Ltd) became a commercial entity with government
stake in the company in accordance with the Petroleum Industry Act (PIA, 2021). The International Oil Companies (IOCs) are
major stakeholders in the upstream sector while Major Marketers and independent marketers are some of the major stakeholders
in the downstream sector (PIA, 2021). The electricity industry is regulated by NERC (Nigerian Electricity Regulatory
Commission)
Empirical Review:
This section will examine significant empirical studies that have analyzed the correlation between consumption of energy,
specifically natural gas, and growth of the economy, concentrating on developing nations such as Nigeria.
Electricity Consumption and Economic Growth:
Numerous individuals have investigated the correlation between electricity or consumption of energy and the growth of the
economy in both developing and developed nations. In Nigeria, natural gas consumption is a pivotal issue owing to its
significance in power generation and industrial operations. Research by Oyeleke and Akinlo (2019) indicates that consumption of
energy, encompassing gas, has a positive impact on GDP. Two-dimensional scatter plots featuring both linear and logarithmic
regressions are used to show overarching trends and patterns, facilitating the identification and elucidation of the relationship
between growth of the economy and gas-derived electricity generation in Nigeria. Figure 6 depicts the correlation between
economic growth and gas-generated power, with the trend line demonstrating a positive relationship between the two throughout
the study period.
Figure 6: Economic growth and electricity production from gas in Nigeria.
Source: Oyeleke and Akinlo (2019) Fi
Shengfeng et al. (2012) used the Vector Error Correction Model (VECM) to analyze the short-term interaction and long-run
connection between real GDP and consumption of energy in China from 1953 to 2009. The results indicated that the consumption
of electricity and the country’s real GDP are cointegrated, and that a unidirectional causal link exists between consumption of
electricity and the growth of the economy in both the short and long run. The research recommended that China should enhance
its efforts to reorganize its industrial framework and optimize its power supply system. The research was carried for China and
not Nigeria.
Nazlioglu et al. (2014) examined the causative relationship between consumption of electricity and growth in the economy in
Turkey from 1967 to 2007 using limits testing cointegration, non-linear Granger causality tests and linear Granger causality.
Their findings proved the long-term cointegration of economic growth and electricity use. The error correction model indicated
that the linear Granger causality findings demonstrated a two directional interrelationship between the variables were observed in
the short term and long periods. Turkey was the area of interest in their research. On the other hand, Gokten and Karatepe (2016)
revealed a unidirectional causal link between the growth of the economy and energy consumption using a bivariate Vector
Autoregressive (VAR) causality test for the years 1950 to 2010.
Similar to this, Shahbaz (2015) used the OLS (Ordinary Least Squares) approach to investigate how Pakistan's sectoral GDP
(agriculture, industry, and services) was affected by power outages between 1991 and 2013. The results showed that low levels of
electricity have a detrimental effect on industrial sector production, have an inverse association with agricultural output, and
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 336
exacerbate load-shedding in the service sector. The research suggested reducing needless energy use, encouraging the use of
electricity-saving gadgets, and encouraging a better receptiveness to energy-saving strategies. The scope of the work is limited to
Parkistan and the variables differ too.
Using cointegration and OLS techniques, Odedairo et al. (2013) investigated the link between energy consumption and the
development of economy of Nigeria from 1975 to 2010. Their results showed a long-term association between the variables and
showed that petrol use did not have a similar impact to that of petroleum and electricity consumption, which both had a
substantial positive correlation with GDP. The result is agreement with that of this work, but the technique used is different.
Similarly, Azlina (2012) examined the correlation between Malaysia's GDP and energy use with a VECM. The research identified
a causal association between GDP and consumption of energy, suggesting that economic expansion influences energy demand
rather than vice versa.
Okorie and Manu (2016) assessed the causal relationship between Nigeria's economic progress and electricity consumption from
1980 to 2014, in comparison to Onyeisi et al. (2016). Utilizing VAR-based methodologies and Johansen cointegration, they
successfully established a long-term relationship among the model's variables. The output indicated sustained link between the
growth of the economy and utilization of energy. The causality study revealed a unidirectional causal link between energy use
and growth in the country’s economy. They recommended that the government enhance daily energy output to meet the
increasing demand. The method used is different although their recommendation is in line with the recommendation of this
research paper.
Utilizing data from 1980 to 2011, Mustapha and Fagge (2015) reassessed the causal relationship between Nigeria's GDP and
energy consumption via variance decomposition, impulse response, and causality analyses. The causality test revealed no causal
relationship between the variables. Moreover, a variance decomposition analysis indicates that labour and capital have a more
substantial impact on output growth than energy use. The variables of this work are different from those of this work, hence, the
gaps need to be filled. Most of these researches concentrated on the linear connection, and ignored the possibility of asymmetric
effects due to variations in the energy supply.
Asymmetric Effects of Energy Consumption using the NARDL model:
Since its introduction by Shin et al. (2013), the NARDL (Nonlinear Autoregressive Distributed Lag) model has gained popularity
as a useful tool for analysing asymmetric interactions in economic research. This approach enables researchers to discriminate
between changes in an independent variable (natural gas consumption by GenCos in this example) that are positive or negative in
relation to a dependent variable (GDP). The NARDL model is used in a number of recent research that examined the asymmetric
impacts of energy consumption and how variations in energy use impact the economy. When Shin et al. (2013) used the NARDL
technique to investigate the asymmetric impacts of oil consumption on economic development in South Korea, they discovered
that the negative consequences of declining oil supply outweighed the benefits of rising supply. Using data from 1990 to 2014,
Farhani and Rahman (2020) researched on the natural gas usage versus economic development in France. To ascertain the long-
term link between the variables, they used ARDL testing technique. In order to determine the direction of causation, they also
used the Granger causality approach, which is a VEC model. Their results demonstrated a long-run cointegration among the
variables, with labour, capital, exports, and natural gas consumption all contributing to France's economic expansion. The
findings of the causality analysis validated the energy-led growth theory, indicating that economic expansion is driven by petrol
consumption. The area of concern to the authors were France although the result is supported by this research paper.
Galadima and Aminu (2018) employed the smooth transition regression (STR) technique to investigate Nigeria’s economic
growth versus the utilization of natural gas from 1981 to 2015. The results of this research demonstrated that there is an
asymmetry link connecting natural gas utilization and the economic advancement in Nigeria. The threshold for natural gas usage
in the nation was 9,085.36 standard cubic meters, but where consumption was below the optimum level. Additionally, it was
discovered that natural gas utilization, in both low and high regimes, had a favourable and significant effect on Nigeria GDP.
There is a variation in the variables and scope of their work from that of this paper.
Galadima and Aminu (2019a) analysed the shock effects of macroeconomic variables on Nigeria's natural gas consumption. They
used a structural VAR (SVAR) model with sign limitations to evaluate how shocks from macroeconomic variables, including:
money supply, inflation, real GDP, and exchange rate were transmitted onto the use of natural gas. The findings showed that
natural gas consumption responded considerably to shocks originating from the money supply and real GDP in the long run and
further in short run, but only in the short-term and not significantly in the case of inflation shock. In ordering the variance
decomposition, however, the money supply, real GDP, inflation, and exchange rate have all contributed to shocks to a greater
extent. 35 observations from annual time for the years 1981 to 2015 were used in this investigation. The scope as well as the
variables and techniques are different from those of this research paper.
Ekeocha, Penzin, and Ogbuabor (2020) used several model parameters to reassess the interaction of Nigeria’s economic growth
and her energy consumption from 1999 to 2016. This experiment used a nonlinear (or asymmetric) ARDL model, alongside an
ARDL-ECM framework that posits a linear correlation rather than a nonlinear one. Their analysis revealed that the impact of
energy use on growth was negligible. Granger causality tests demonstrated a unidirectional causal link between economic
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 337
development and energy use. This research examined the period from 1999 to 2016. The scope and the variables of their research
vary from those of this paper.
Yakubu et al. in 2022 investigated the influence of electrical power on Nigeria's economic development from 1981 to 2019. The
ARDL bounds cointegration test was used after they verified the order of integration among the variables (mixed order), as
established by the ADF and PP tests. The findings indicate that electricity consumption has a statistically significant and positive
influence on the growth of the economy for the short and long run. Henry et al. (2021) worked on the impact of transmission and
distribution losses, representing electric power shortages, on Nigeria's economic development, focusing on the real GDP of the
industrial and agricultural sectors. They found that a 1% rise in electric power transmission and distribution losses leads in a 3%
long-term decline in agricultural production, whereas the short-term impact is minimal. This analysis utilized data span from 1981
to 2017 and they used ARDL model. The variables and scope of their work is quite different from those of this paper.
Research Gaps and Contributions:
There are significant gaps in the literature, particularly when it comes to the asymmetric impacts of GenCos natural gas use and
the growth of Nigerian economy. Nevertheless, the body of current research offers a strong foundation for understanding the
relationship between them. The predominant body of research has concentrated on linear relationships, neglecting the potential
for varying impacts on economic output resulting from fluctuations in energy use. Moreover, despite its use in several economic
studies, the NARDL methodology has not been thoroughly investigated regarding this topic. This study aims to address these
gaps by:
1. analyzing the asymmetric impacts of natural gas consumption by Nigerian electric power plants on GDP;
2. utilizing the NARDL method to effectively capture the diverse implications of variations in natural gas consumption; and
3. offering policy recommendations based on the research findings to optimize the use of natural gas for power generation and
economic advancement in Nigeria.
III. Methodology
In this research study, the impact of natural gas usage on the Nigerian economy was assessed using NARDL model. The NARDL
method was used to examine quarterly data on the consumption of natural gas in power plants and its asymmetric effects on GDP.
Other variables considered include LPG consumption and Natural gas consumption by industries. The study utilized data from an
11-year period to explore the implications and validity of the long-term connections between the growth of the economy and
domestic gas consumption of power plants.
Introduction to NARDL
The NARDL model is a technique that aims to evaluate the long-term association among variables and consider non-linearity and
short-term dynamics. It is a variation of the linear ARDL model, which includes non-linear terms such as polynomials and
logarithms in the equations (Pesaran, 2017). The ARDL model evaluates the correlation among variables over both short-term
and long-term periods by incorporating both autoregressive and distributed lag components. The autoregressive component
examines the short-term dynamics of the relationship, including the potential for feedback effects among the variables.
Meanwhile, the distributed lag component analyzes the long-term dynamics, including the potential for equilibrium relationships
among the variables over the long-term. The steps taken to carry out NARDL analysis starts with identifying the variables of
interest. Next is to do stationary tests to ensure that the data is suitable for the analysis. If the variables are stationary at level and
first difference, then bounds test follows to ascertain for cointegration. The NARDL test is then done followed by diagnostics and
stability tests.
Research Design:
Quantitative research approach would be used, which is especially adept at examining the correlation between use of natural gas
by GenCos) and Nigeria's GDP. The research utilizes a quantitative methodology, applying statistical and econometric
approaches to evaluate the hypothesis and elucidate the link between the variables. The study used the NARDL model to detect
possible asymmetries among the variables.
Data Collection and Sources:
The research relied on secondary data sourced from credible references to guarantee precision and dependability. This study
examined quarterly data spanning from Q1 2010 to Q4 2020, offering a comprehensive perspective on the long-term correlation
between consumption of natural gas by GenCos and Nigeria's GDP. This period includes critical occurrences that have impacted
GDP, notably the COVID-19 pandemic. The study investigation used data for 11-year period to assess the relationship between
power plants gas consumption versus economic growth, while also identifying potential causal links. The quarterly data yields 44
data points, facilitating a comprehensive examination of economic performance over time, emphasizing long-term trends, and
providing enhanced reliability compared to shorter-term data by reducing noise. This renders it especially advantageous for policy
making and the formulation of business decisions.
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 338
Table 2: Description of variables
Variable
Abbreviation
Unit
Publisher
Real Gross Domestic Product
GDP
Billion Naira
Central Bank of Nigeria
LPG Consumption
GH
Metric Tons
United Nations data
Gas consumption for electric
power generation
GE
Billion Cubic
Meters
Gas Exporting Countries
Forum
Gas consumption by industries
GI
Billion Cubic
Meters
Gas Exporting Countries
Forum
Model Specification and Description
The research employs the NARDL technique, as established by Shin et al. (2013), to examine the asymmetric impacts of
consumption of natural gas by GenCos on Nigerias GDP. Its primary benefit is its capacity to differentiate between the effects of
fluctuations in the independent variable (GenCos - GE) on the dependent variable (GDP).
The linear relationship is expressed as follows: the linear functional form can be represented as:
GDP = f(GH, GE, GI) (3.1)
The equation stated represents the specified model. It can also be expressed in a non-linear functional form where the explanatory
variables are split into positive and negative partial forms:
GDP = f(GH
+
, GH
-
, GE
+
, GE
-
, GI
+
, GI
-
) (3.2)
expressed in form of coefficients as;
GDP = α
0
+ α
1
GH
+
+ α
2
GH
-
+ α
3
GE
+
+ α
4
GE
-
+ α
5
GI
+
+ α
6
GI
-
+ ϵ
t
(3.3)
The functional models at equation 3.3 is best expressed by taking the logarithm transform of each variable. This is to ensure that
the units of each variable are consistent, and this helps in addressing the non-stationarity of data.
Taking log of both side in equation 3.3
LGDP = α
0
+ α
1
LGH
+
+ α
2
LGH
-
+ α
3
LGE
+
+ α
4
LGE
-
+ α
5
LGI
+
+ α
6
LGI
-
+ ϵ
t
(3.4)
Where:
LGDP = Natural Logarithm of Real GDP
LGH = Natural Logarithm of LPG consumption
LGE = Natural Logarithm of Gas consumption for electric power generation
LGI = Natural Logarithm of Gas consumption for industries
α
0
= Intercept
α
1,
α
2…
α
6
are the parameter estimates for the dependent variables
ϵt = Error term
Estimation Technique (NARDL)
The developed model is articulated as double-logged model. A double log model serves as a robust econometric method for
examining the relationship among dependent and independent variables by transforming them into natural logarithms. Pesaran
and Shin (1999). The cointegration test applied with this model is the NARDL bounds test, which allows the evaluation of both
long-term and short-term dynamics, as noted by Narayan and Poi (2005). Other benefits of this method when compared to other
methods include: (i) Its capacity to show how the variables react to positive and negative changes among the variables and the
changes in these responses with time variation. (ii) it has the capacity to carry out short- and long run estimates of regressors with
respect to regressand. These benefits and more led to the choice of this technique.
Steps in conducting NARDL analysis
To ensure the accuracy and reliability of the NARDL model, several tests must be conducted. Firstly, the stationarity of the
variables must be tested using the ADF, KPSS tests or any other unit root test, as the ARDL model only works with I(0) and/or
I(1) variables (Emeka and Aham, 2016, Udoudo et al., 2023). Once stationarity is confirmed, the optimal lag for the ARDL model
must be determined using different criteria (AIC criteria was used) to account for any biases. The NARDL model can then be
built from the ARDL model, and cointegration tests is conducted to determine the presence of a long-run relationship between the
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 339
dependent and explanatory variables. The short-run and long-run coefficients can be generated, taking the error correction term
into account. Wald tests can then be used to confirm the long- and short-run asymmetries. To validate the results, tests for
autocorrelation, normality and heteroscedasticity must be conducted. Stability tests must also be performed to ensure that the
estimated coefficients are reliable, confirming the robustness of the model (Sohail et al., 2021). For this purpose, we considered
the Cumulative sum and Cumulative Sum of Squares to examine stability and reliability of the models.
Finally, the outcomes should be analyzed to see if they make intuitive sense.
The generalized asymmetry or nonlinear equation is expressed as:
Y
t
= f(Y
t-1
, X
t-1
, ε
t
) + ε
t
(3.5)
When using the non-linear ARDL to analyze Y and X, the equation would be represented as:
Y
t
= α
0
+ α
1
Y
t-1
+ α
2
X
t-1
+ β
1
Y
t-1
X
t-1
+ γ
1
(Y
t-1
)
2
+ δ
1
(X
t-1
)
2
+ ε
t
(3.6)
Where α
1
, α
2
, β
1
, γ
1
and δ
1
are parameters and ε
t
error term.
can be further decompose as:
=
+
+
(3.7)
(3.8)
(3.9)
The decomposed equations are substituted into an ARDL setting suggested by Pesaran et al. (2001):
(3.10)
where,
= dependent part, and
=independent part
and
are partial sum of
β
+
=
, and
β
-
associated parameters to
on
.
This is calculated by dividing both
and
are estimates of the increase and reduction of the short-run impacts of the independent variables.
Null hypothesis or no co-integration, is expressed as:
(3.11)
Here, q at 0,1,2,3,4 denotes the maximum lags for LGDP, LGH, LGE, and LGI, respectively.
Thus, for co-integration to exist, an error correction model (ECM) representation is generated as:
The estimation of the non-linear ECM in the short-run to evaluate the asymmetric effect among the variables is derived from
equation 3.13 as:
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 340
.
Where,
● Real GDP, LPG consumption (GH), Electric power plant gas consumption (GE), and Industry gas consumption (GI) are
variables expressed in logarithm.
● ϵ
t
= Error term
● ECT = Error correction term.
● q = best lag order in relation to the variable
● β
1i
to β
7i
,
and α
1
toα
i3
represents the asymmetric coefficients of the short-run.
● β
8
to β
13
and α
4
to α
5
represents the asymmetric coefficients.
● represents the long-run coefficient of the dependent term.
β
+
=
, and β
-
(3.15)
Assuming,
and
represents the positive and negative asymmetric coefficients of the independent variable, respectively and,
represents the long-run coefficient of the dependent variable.
Wald Test is then applied to confirm the presence of long-term and short-run asymmetries in NARDL modeling, the Wald Test is
employed (Shin, Yu, & Greenwood-Nimmo, 2014). The test is based on a comparison of the sums of square residuals between
the restricted and unrestricted models, resulting in a test statistic. If the calculated test statistic is higher than the critical value of
the Chi-square distribution, the null hypothesis is rejected, indicating that significant asymmetry exists. The null hypothesis
assumes that the coefficients are equal, indicating no long-term asymmetry, while the alternative hypothesis indicates the
presence of long-run asymmetry by indicating that the coefficients are not equal.
The test statistic is calculated as:
WLR = [g(γ) - g(-γ)]' [Var[g(γ) - g(-γ)]]
-1
[g(γ) - g(-γ)] (3.16)
Where, γ is the estimated coefficient of the ECM term, g(γ) is the vector of restrictions on the coefficients, and Var[g(γ) - g(-γ)] is
the covariance matrix of the limits.
The asymmetric dynamic multiplier equations: The dynamic multipliers describe the process by which the dependent variable
(LGDP) adapts to a new long-term equilibrium following the consequences of positive and negative shocks from the independent
variables.
Where, as q → ,
→
, and
and m is the asymmetric multiplier.
IV. Results and Discussion:
The NARDL model results is hereby presented and discussed.
Unit Root Tests:
The ADF test result reported in Tables 3 shows that the LGDP was seen to be stationary at level zero while LGE and LGH were
found to be stationary at order one and LGI maintained stationarity at both levels. According to Udoudo et al., 2023, Engel-
Granger and Johansen co integration method is not suitable for the analysis because of the distinct order in the stationary of the
variables. The results validate the use of the NARDL method in conducting co-integration test among the variables.
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 341
Table 3: ADF Unit Root Test result
Level I(0)
First difference I(1)
Variable
Test value
Probability
Test value
Probability
Stationarity
LGDP
ADF (-4.749873)
0.004
ADF (-1.877482)
0.3391
I (0)
1% (-3.605593)
1% (-3.610453)
5%(-2.936942)
5% (-2.938987)
10%(-2.606857)
10% (-2.607932)
LGE
ADF (-1.464311)
0.5408
ADF (-5.209183)
0.0001
I (1)
1% (-3.610453)
1% (-3.610453)
5% (-2.938987)
5% (-2.938987
10% (-2.607932)
10% (-2.607932)
LGH
ADF (-0.142578)
0.9374
ADF (-6.155365)
0.0000
I (1)
1% (-3.610453)
1% (-3.610453)
5% (-2.938987)
5% (-2.938987)
10% (-2.607932)
10% (-2.607932)
LGI
ADF (-3.289609)
0.0216
ADF (-6.521649)
0.0000
I (0) & I (1))
1% (-3.592462)
1% (-3.596616)
5% (-2.931404)
5% (-2.933158)
10% (-2.603944)
10% (-2.604867)
Source: Author’s Compilation from EVIEWS
NARDL Estimation:
Table 4: NARDL Estimation Output for the Model
Variable
Coefficient
Std. Error
t-Statistic
Prob.*
LGDP(-1)
-0.854321
0.307486
-2.778407
0.0141
LGDP(-2)
-1.185284
0.177767
-6.667629
0.0000
LGDP(-3)
-0.843448
0.298923
-2.821627
0.0129
LGE_POS
0.040427
0.083637
0.483367
0.6358
LGE_NEG
0.064744
0.102232
0.633306
0.5361
LGE_NEG(-1)
0.088743
0.135408
0.655378
0.5221
LGE_NEG(-2)
0.218898
0.131488
1.664777
0.1167
LGE_NEG(-3)
0.237984
0.140494
1.693908
0.1109
LGH_POS
0.058336
0.140547
0.415066
0.6840
LGH_POS(-1)
0.235205
0.125952
1.867415
0.0815
LGH_POS(-2)
0.071230
0.137125
0.519452
0.6110
LGH_POS(-3)
0.281587
0.117546
2.395552
0.0301
LGH_NEG
-0.486404
0.492703
-0.987214
0.3392
LGH_NEG(-1)
-0.955528
0.449937
-2.123692
0.0507
LGH_NEG(-2)
-0.393322
0.497657
-0.790348
0.4416
LGH_NEG(-3)
-1.282468
0.431013
-2.975476
0.0094
LGI_POS
0.289799
0.106718
2.715562
0.0160
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 342
LGI_POS(-1)
-0.177491
0.124046
-1.430844
0.1730
LGI_POS(-2)
-0.066860
0.131556
-0.508227
0.6187
LGI_POS(-3)
-0.231602
0.111227
-2.082240
0.0549
LGI_NEG
-0.214615
0.162028
-1.324560
0.2051
LGI_NEG(-1)
0.385519
0.188653
2.043538
0.0590
LGI_NEG(-2)
0.352028
0.209302
1.681912
0.1133
LGI_NEG(-3)
0.248092
0.193504
1.282102
0.2193
C
36.68142
6.475460
5.664682
0.0000
R-squared
0.974034
Mean dependent var
9.724505
Adjusted R-squared
0.932488
S.D. dependent var
0.095240
S.E. of regression
0.024746
AIC
-4.291119
Sum squared resid
0.009186
Schwarz criterion
-3.235569
Log likelihood
110.8224
Hannan-Quinn criter.
-3.909466
F-statistic
23.44482
Durbin-Watson stat
1.996801
Prob (F-statistic)
0.000000
Source: Authors Compilation from Eviews
The results in Table 4 displayed the regression of the model. The value of Durbin-Watson stat of 1.9968, F-statistic of 23.4448,
R-squared of 0.9740 and adjusted R squared of 0.9324 is positive indication to the quality of the model.
The Bound test:
According to Emeka and Aham, 2016, when the calculated F statistics value is higher than the upper bound I(1) of critical values,
then we confirm the presence of co-integration among the variables. Secondly, when the calculated value of F statistics is smaller
than the lower bound I(0) of the critical values, it means that no co-integration exists among the variables. Thirdly, where the
calculated F statistics value lies between the lower bound I(0) and upper bound I(1), the test is assumed inconclusive. In Table 5,
the F-statistic for the model is 7.083805. This value is compared against the I(0) and I(1) (lower and upper bounds respectively) at
significant levels of 1%, 5%, and 10%. The results demonstrate that the F-statistic exceeds both bounds at all levels, hence, null
hypothesis (H
0
) of no cointegration is therefore rejected. This suggests the presence of asymmetric cointegration among the
variables, underscoring the role of asymmetries and nonlinearities in analyzing Nigeria's GDP. The result of the F-bound test
displayed in Table 5 with F statistics value of 7.0838 which is far greater than the value of the upper bound critical value of 3.61
at 5% significance shows that we can proceed to carry out long run and short run estimation of the model (Sohail et al., 2021).
Table 5: F-bound co-integration results
TEST STAT
Value
Significant
I(0) Asymptotic: n=1000
I(1) Asymptotic: n=1000
F- statistics
7.083805
10%
2.12
3.23
K
6
5%
2.45
3.61
2.5%
2.75
3.99
1%
3.15
4.43
Source: Author’s Compilation from EVIEWS
The Long run estimation
Table 6: Long-run Estimations for the specified NARDL model
Variable
Coefficient
Std. Error
t-Statistic
Prob.
MODEL
LGE_POS
0.010411
0.020596
0.505492
0.0206
LGE_NEG
0.157188
0.022772
6.902544
0.0000
LGH_POS
0.166456
0.017231
9.660322
0.0000
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 343
LGH_NEG
-0.802905
0.071167
-11.28201
0.1200
LGI_POS
-0.047940
0.023763
-2.017394
0.0619
LGI_NEG
0.198561
0.034754
5.713354
0.0000
The long-run coefficients are 0.010411, and 0.157188, respectively, with a probability value of 0.0206, and 0.000 respectively.
This implies that for every 1% increase in LGE in positive direction, the LGDP increases by 1.04%, in the long-run. Moreso, the
LPG consumption (LGH) in the positive, and negative direction had a positive, and negative impact on LGDP, respectively. The
long-run coefficients are 0.166456, and -0.802905, respectively, with a probability value of 0.000, and 0.1200 respectively. This
implies the impact of LGH in positive direction is statistically significant but insignificant in the negative direction, i.e for every
1% rise in LGH in positive direction, the LGDP increases by 16.64%. There is no effect on the negative direction of LPG
consumption on GDP because it is not statistically significant.
Furthermore, the gas consumption in Industries (LGI) have significant impacts on LGDP in the negative direction but
insignificant in the positive direction (LGI_POS, and LGI_NEG). The long-run coefficients are -0.047940, and 0.198561,
respectively, with a probability value of 0.0619, and 0.000, respectively. This implies the impact of LGI in positive direction is
statistically insignificant, but significant in the negative direction at a 5% probability threshold. This means that for every 1%
drop in industries gas consumption, the GDP decreases by 19.85%.
The asymmetric cointegrating level equation for the model can be represented as:
(4.1)
The model can be used to predict the estimates for the parsimonious NARDL equation at different lag periods.
Table 7: Short-Run Estimations for the Specified NARDL Model
Variables
Coefficient
Std. Error
t-Statistic
Prob.
MODEL
C
36.68142
4.398251
8.340001
0.0000
D(LGDP(-1))
2.028732
0.253667
7.997624
0.0000
D(LGDP(-2))
0.843448
0.197796
4.264237
0.0007
D(LGE_NEG)
0.064744
0.077558
0.834782
0.4169
D(LGE_NEG(-1))
-0.456882
0.096083
-4.755051
0.0003
D(LGE_NEG(-2))
-0.237984
0.090506
-2.629491
0.0190
D(LGH_POS)
0.058336
0.091121
0.640206
0.5317
D(LGH_POS(-1))
-0.352817
0.100490
-3.510971
0.0032
D(LGH_POS(-2))
-0.281587
0.083726
-3.363218
0.0043
D(LGH_NEG)
-0.486404
0.335591
-1.449394
0.1678
D(LGH_NEG(-1))
1.675790
0.389395
4.303573
0.0006
D(LGH_NEG(-2))
1.282468
0.309275
4.146684
0.0009
D(LGI_POS)
0.289799
0.070417
4.115445
0.0009
D(LGI_POS(-1))
0.298462
0.092527
3.225682
0.0057
D(LGI_POS(-2))
0.231602
0.079656
2.907528
0.0108
D(LGI_NEG)
-0.214615
0.120714
-1.777878
0.0957
D(LGI_NEG(-1))
-0.600120
0.151935
-3.949861
0.0013
D(LGI_NEG(-2))
-0.248092
0.117290
-
2.115203
0.0516
CointEq(-1)*
-3.883053
0.466044
-
8.331944
0.0000
R-squared
0.967810
Mean dependent var
0.006847
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 344
Adjusted R-squared
0.940219
S.D. dependent var
0.085539
S.E. of regression
0.020914
Akaike info criterion
-4.591119
Sum squared resid
0.009186
Schwarz criterion
-3.788901
Log likelihood
110.8224
Hannan-Quinn criter.
-4.301063
F-statistic
35.07658
Durbin-Watson stat
1.996801
Prob(F-statistic)
0.000000
Source: Author’s Compilation from EVIEWS
From the estimates, the model shows that the gas consumption by GenCos in the negative direction D(LGE_NEG) has a positive
impact on LGDP in the short run. The short-run coefficient for D(LGE_NEG) is 0. 064744 and its respective probability value is
0. 4169. This indicates that the impact of D(LGE_NEG) in the current period is statistically insignificant. This also means that for
every 1% decrease in (D(LGE_NEG)) during the current period, the LGDP decreases by 45.68%. %. Additionally, the coefficient
of the co-integration equation during the previous year (CointEq(-1)) denotes the Error Correction Mechanism (ECM). The model
generates a negative and statistically significant coefficient of the ECM at -3.883053. It indicates that approximately 388% of the
short-run disequilibrium from the previous year’s disturbance converges back to a long-run equilibrium in the present year. The
adjusted R-Square revealed that 94.02% variation in the LGDP is explained by the changes in logged value of gas consumption in
industries, electric plants, and LPG consumption. The F-statistic value of 35.07658 and p-value of 0.0000 proves the relevance
and adequacy of the NARDL model.
The ECM equation for short run can be expressed as:
(4.3)
Wald Tests for the Long-run and Short-run asymmetries.
Tables 8 present the results of the Wald test, which examines the existence of both long-run and short-run asymmetries. At a 0.05
significance level, the results reveal that the effect of LGE on LGDP is asymmetric in the long run, as indicated by the Chi-square
probability of 0.0072. This leads to the rejection of the null hypothesis, suggesting asymmetries in the long term. In the short term
however, the probability value of 0.0590 confirm symmetry.
Table 8: Wald Tests
Variable
Analysis
Long run
Test stat
Value
df
Prob
Inference
LGE
T-stat
-2.862723
33
0.0072
Asymmetric
F-stat
8.195183
1, 33
0.0072
Chi- square
8.195183
1
0.0042
Short run
T-stat
-1.951442
35
0.0590
Symmetric
F-stat
3.808125
1, 35
0.0590
Source: Author’s Compilation
NARDL Dynamic Multipliers
To analyze how variables impact GDP over time, dynamic multipliers are used to track their effects, particularly identifying any
asymmetries caused by changes in independent variables. A positive change in the dependent variable follows the solid black
line, while a negative shock follows the dotted black line. The light-dotted red line shows the asymmetric plot, with upper and
lower bands, and the dark red dotted line indicates the difference between positive and negative changes. If the x-axis zero line
falls within this interval, it indicates no asymmetric effect from the explanatory variable.
Figure 9 shows that electric power plants (GE) gas consumption impact on GDP is asymmetric in the long run, aligning with the
results of the long run Wald tests. The value of the negative shock at long run is about (-7) while the positive shock stood below
(6) as represented by the tiny, dotted lines, the upper one for positive and the lower for negative response of GDP to gas
consumption by power plants. This is in agreement with the result of the Wald test that confirmed asymmetry in long run. The
effect of negative shock is higher in agreement to the NARDL output where the negative effect of gas consumption had higher
effect on the Nigerian GDP. In figure 9 legend, the black solid line represents positive while the black dotted line represents
negative shocks, we see that at the short run, there is symmetry between the positive shocks and negative shocks as the
asymmetric plot lie along the zero line and around the 95% significance. This graph support the result of the Wald test that
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 345
affirmed symmetry in the short term. The thick red dotted line is the asymmetric plot and in the long run, it tilted towards the
negative part of the graph thereby supporting the NARDL output of higher negative impact in the long run and Wald test
asymmetric result in the long run.
-8
-6
-4
-2
0
2
4
6
1 3 5 7 9 11 13 15
Multiplier for LGE(+)
Multiplier for LGE(-)
Asymmetry Plot (with C.I.)
Figure 9: LGE Multiplier
Diagnostic Tests
Diagnostic tests were further carried out to confirm the robustness of the model such as normality of residuals, heteroskedasticity,
Ramsey Reset and serial correlation tests. The results indicate that the model lack autocorrelation issues because the estimated
LM statistic is statistically insignificant, the result of the Ramsey RESET test statistic confirms that the model is well-specified.
Furthermore, the output from the ARCH tests confirmed that there are no heteroscedasticity issues with the model. Finally, the
normality test with Jaque-Bera of 2.8184 and probability value of 0.244 confirms that the residual series are normally distributed.
The results of these tests are summarized in the Table 9.
Table 9: Diagnostic Tests
TEST
STATISTICS
Breusch-Godfrey Serial
Correlation LM Test
F-STAT
0.180544
Prob. F(2,13)
0.8369
Obs*R-squared
1.081014
Prob. Chi-Square (2)
0.5825
The ARCH
heteroscedasticity test
F-STAT
1.846035
Prob. F(3,33)
0.1580
Obs*R-squared
5.317073
Prob. Chi-Square(3)
0.1500
The Ramsey RESET
Df
Prob
T-STAT
1.867059
14
0.0830
F-STAT
3.485911
(1, 14)
0.0830
Normality Test
Jaq-Bera
2.8184
PROB
0.2443
Stability Diagnostics
The Cumulative Sum (CUSUM) Test and CUSUM of Square Test indicates the stability of the findings generated from the
estimations of the long-run and short-run parameters for NARDL model.
-12
-8
-4
0
4
8
12
II III IV I II III IV I II III IV I II III IV
2017 2018 2019 2020
CUSUM 5% Significance
Figure 7: Lgdp Cusum Test
Source. Author’s compilation
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 346
The CUSUM and CUSUMSQ statistic for LGDP in model one falls inside the critical bands of the 5% significance interval of
parameter stability. The finding validates the lack instability in the coefficients of the estimates of the model throughout the
period of the investigation.
Figure 8: Lgdp Cusumsq Test
Source. Author’s compilation
V. Conclusions and Recommendations:
This paper used NARDL approach to examine the asymmetric relationship of GenCos natural gas consumption and the growth of
Nigerian economy from year 2010 to year 2020. To ascertain the reliability of the results, we performed ADF unit root test. The
bounds test for co-integration was done to ascertain the features of the model. The short-term and long-term characteristics of the
model were then carried out. Wald tests to confirm asymmetry were performed on both short-run and long-run coefficients.
Statistical diagnostic tests and stability diagnostic tests, such as the Cumulative Sum (CUSUM) Test, CUSUM Square tests were
also conducted to confirm the robustness of the model.
The empirical findings of the research revealed that the natural gas consumption by power generating companies in Nigeria is
asymmetric with the growth of Nigerian economy over long period and symmetric in short duration. It further indicates that
increase in gas consumption by GenCos increases Nigeria GDP and decrease in their consumption decreases the country’s GDP.
Therefore, the need for a stable, sustainable and increasing supply of natural gas to power plants is vital to facilitate the
development of the economy of Nigeria.
We recommend that the government and policy makers, which include NUPRC and NMDPRA should implement measures to
enhance the infrastructure for gas production, storage and transmission to power plants to stop the recurrence inadequacy of gas
supply to power plants, thereby increasing gas consumption by electric power plants. The government should encourage private
sector participation in the gas industry and electricity generation to boost gas production and consumption through liberalization
of the sector. This will lead to increased power supply, gas production and consumption, thereby contributing positively to the
country’s economic growth. This recommendation agrees with that of Sohail et al, 2021 for Pakistan that concluded that energy
security is important in sustaining the growth of the economy especially in developing countries that are starved of energy and
further recommended enhanced consumption of relatively clean energy resources.
References
1. Akinlo, A. E. (2009). Electricity consumption and economic growth in Nigeria: Evidence from cointegration and co-
feature analysis. Journal of Policy Modeling, 31(5), 681–693. https://doi.org/10.1016/j.jpolmod.2009.03.004
2. Apergis, N., & Tang, C. F. (2013). Is the energy-led growth hypothesis valid? New evidence from a sample of 85
countries. Energy Economics, 38, 24–31. https://doi.org/10.1016/j.eneco.2013.02.007
3. Azlina, A. A. (2012). Energy Consumption and Economic Development in Malaysia: A Multivariate Cointegration
Analysis. Procedia - Social and Behavioral Sciences, 65, 674–681. https://doi.org/10.1016/j.sbspro.2012.11.183
4. Ekeocha, P. C., Penzin, D. J., & Ogbuabor, J. E. (2020). Energy consumption and economic growth in nigeria: A test of
alternative specifications. International Journal of Energy Economics and Policy, 10(3), 369–379.
https://doi.org/10.32479/ijeep.8902
5. Emeka Nkoro, and Aham Kelvin Uko, “Autoregressive Distributed Lag (ARDL) cointegration Technique: Application
and Interpretation,” Journal of Statistical and Econometric Methods, vol. 5, no. 4, pp. 63-91, 2016.
6. Energy Agency, I. (2022). World Energy Outlook Special Report Africa Energy Outlook 2022. www.iea.org
7. Energy Information Administration (2018), Energy Efficiency and Conservation [online]
https://www.eia.gov/energyexplained/index.php?page=about_energy_efficiency
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 347
8. Farhani, S. & Rahman, M., M. (2020). Natural gas consumption and economic growth nexus: an investigation for
France, International Journal of Energy Sector Management, 14(2), 261-284. http://dx.doi.org/10.1108/IJESM-07-2019-
0005
9. Galadima, D. M., & Wambai Aminu, A. (2019a). Positive and negative impacts of natural gas consumption on economic
growth in Nigeria: a nonlinear ARDL approach. In African J. Economic and Sustainable Development (Vol. 7, Issue 2).
10. Galadima, M. D. and Aminu, A. W. (2018). Application of smooth transition regression (STR) model on the relationship
between natural gas consumption and economic growth in Nigeria. International Journal of Energy Sector Management,
12(2), 284-296.
11. Gokten, S., & Karatepe, S. (2016). Electricity consumption and economic growth: A causality analysis for Turkey in the
frame of import-based energy consumption and current account deficit. Energy Sources, Part B: Economics, Planning
and Policy, 11(4), 385–389. https://doi.org/10.1080/15567249.2012.666332
12. Henry, J. T., Ndem, B. E., Ujong, O. L., & Ihuoma, C. M. (2021). Electric power deficit and economic growth in
Nigeria: A sectoral analysis. International Journal of Energy Economics and Policy, 11(6), 508–516.
https://doi.org/10.32479/ijeep.11491
13. Malami Mohammed, Nwosi-Anele, A. S., and O. Iledare. "Petroleum Industry Act 2021 and Natural Gas Development
in Nigeria: Matters Arising." Paper presented at the SPE Nigeria Annual International Conference and Exhibition,
Lagos, Nigeria, August 2024. doi: https://doi.org/10.2118/221708-MS
14. Mustapha, A. M., & Fagge, A. M. (2015). Energy consumption and economic growth in Nigeria: A causality
analysis. Journal of Economics and Sustainable Development, 6(13). Energy Consumption and Economic Growth in
Nigeria: A Causality Analysis - CORE Reader
15. Narayan, P. K., & Prasad, A. (2008). Electricity consumption-real GDP causality nexus: Evidence from a bootstrapped
causality test for 30 OECD countries. Energy Policy, 36(2), 910–918. https://doi.org/10.1016/j.enpol.2007.10.017
16. Narayan, P.K. (2005) ‘The saving and investment nexus for China: evidence from cointegration tests’, Applied
Economics, Vol. 37, pp.1979–1990.
17. Nigerian Electricity Regulatory Commission (NERC), 2023_Annual_Report.
18. North, D. C. (2012). An introduction to institutions and institutional change. In Institutions, Institutional Change and
Economic Performance (pp. 3–10). Cambridge University Press. https://doi.org/10.1017/cbo9780511808678.003
19. Nazlioglu, S., Kayhan, S., & Adiguzel, U. (2014). Electricity consumption and economic growth in Turkey:
Cointegration, linear and nonlinear granger causality. Energy Sources, Part B: Economics, Planning and Policy, 9(4),
315–324. https://doi.org/10.1080/15567249.2010.495970
20. Nwaoha, C., & Wood, D. A. (2014). A review of the utilization and monetization of Nigeria’s natural gas resources:
Current realities. In Journal of Natural Gas Science and Engineering (Vol. 18, pp. 412–432). Elsevier.
https://doi.org/10.1016/j.jngse.2014.03.019
21. Nwabueze, G., Joel, O., & Nwaozuzu, C. (2022). Analysis of Nigerian Natural Gas Consumption (1990 – 2020). A
Vector Error Correction Model Approach. International Journal of Engineering Technologies and Management
Research, 9(1), 7-19 https://doi:10.29121/ijetmr.v9.i1.2022.1075.
22. Obada, D. O., Muhammad, M., Tajiri, S. B., Kekung, M. O., Abolade, S. A., Akinpelu, S. B., & Akande, A. (2024). A
review of renewable energy resources in Nigeria for climate change mitigation. Case Studies in Chemical and
Environmental Engineering, 9. https://doi.org/10.1016/j.cscee.2024.100669
23. Onyeisi, O., Odo, S., Ogbonna, O. S., Idenyi, O. S., & Nick, A. (2016). Power Generation Capacity and Economic
Growth in Nigeria: A Causality Approach. In European Journal of Business and Management www.iiste.org ISSN (Vol.
8, Issue 32). Online. https://www.researchgate.net/publication/333376282
24. Odedairo, B., Babatunde Onakoya, A., Olatunde Onakoya, A., Adejuwon Jimi -Salami, O., & Omoniyi Odedairo, B.
(2013). Energy consumption and Nigerian economic growth: An empirical analysis. 9(4), 1857–7881.
https://www.researchgate.net/publication/277306368
25. Owebor, K., Diemuodeke, E. O., Briggs, T. A., & Imran, M. (2021). Power Situation and renewable energy potentials in
Nigeria – A case for integrated multi-generation technology. In Renewable Energy (Vol. 177, pp. 773–796). Elsevier
Ltd. https://doi.org/10.1016/j.renene.2021.06.017
26. Okorie, David Iheke and Sylvester, Manu Adasi, Electricity Consumption and Economic Growth: The Nigerian Case
(December 2016). international Journal of Current Research, Vol 8: 12, 2016, Available at
SSRN: https://ssrn.com/abstract=3003553
27. Oyeleke, O. J., & Akinlo, T. (2020). Energy generation and economic growth: empirical evidence from Nigeria.
Environment, Development and Sustainability, 22(8), 7177–7191. https://doi.org/10.1007/s10668-019-00476-4
28. Ozcan, B., & Ozturk, I. (2019). Renewable energy consumption-economic growth nexus in emerging countries: A
bootstrap panel causality test. Renewable and Sustainable Energy Reviews, 104, 30–37.
https://doi.org/10.1016/j.rser.2019.01.020
29. Ozturk, I. (2010). A literature survey on energy-growth nexus. Energy Policy, 38(1), 340–349.
https://doi.org/10.1016/j.enpol.2009.09.024
30. Petroleum Industry Act 2021. https://pia.gov.ng/petroleum-industry-act/
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XIII, Issue XII, December 2024
www.ijltemas.in Page 348
31. Pesaran, M.H., Shin, Y. and Smith, R.J. (2001) ‘Bounds testing approaches to the analysis of level relationships’, J.
Appl. Econom., Vol. 16, pp.289–326.
32. Shahbaz, M. (2015). Munich Personal RePEc Archive Measuring Economic Cost of Electricity Shortage: Current
Challenges and Future Prospects in Pakistan Measuring Economic Cost of Electricity Shortage: Current Challenges and
Future Prospects in Pakistan.
33. Shahid, M., Mehmood, R. K., & Anjum, M. N. (2021). Empirical Analysis of Poverty Reduction and Governance in
Pakistan: A Qualitative Comparison with Sub-Continent. Global Economics Review, VI(II), 87–102.
https://doi.org/10.31703/ger.2021(vi-ii).08
34. Shin, Yongcheol and Yu, Byungchul and Greenwood-Nimmo, Matthew, Modelling Asymmetric Cointegration and
Dynamic Multipliers in a Nonlinear ARDL Framework (October 21, 2013). Festschrift in Honor of Peter Schmidt, W.C.
Horrace and R.C. Sickles, eds., Forthcoming, Available at
SSRN: https://ssrn.com/abstract=1807745 or http://dx.doi.org/10.2139/ssrn.1807745
35. Sohail, .M., Li, Z., Murshed, M. et al. An analysis of the asymmetric effects of natural gas consumption on economic
growth in Pakistan: A non-linear autoregressive distributed lag approach. Environ Sci Pollut Res 29, 5687–5702 (2022).
https://doi.org/10.1007/s11356-021-15987-9
36. Solarin, S.A. and Ozturk, I. (2016) ‘The relationship between natural gas consumption and economic growth in OPEC
members’, Renewable and Sustainable Energy Reviews, Vol. 58, pp.1348–1356.
37. Udoudo, K. J., Kalu, I. E., and Oduola, K. "Effect of Liquefied Natural Gas Exports on the Nigeria Economy: An
ARDL Model Approach," SSRG International Journal of Economics and Management Studies, vol. 10, no. 5, pp. 35-46,
2023. https://doi.org/10.14445/23939125/IJEMS-V10I5P105
38. Ummalla, M., & Samal, A. (2019). The impact of natural gas and renewable energy consumption on CO2 emissions and
economic growth in two major emerging market economies. Environmental Science and Pollution Research, 26(20),
20893–20907. https://doi.org/10.1007/s11356-019-05388-4
39. World Bank Report; Resilience through Reforms. (2021). www.worldbank.org
40. Yakubu, M., Chindo, S., & Bala, U. (2022). Impact of Electricity Power on Economic Growth in Nigeria: Evidence from
ARDL Bounds Approach. In CONFLUENCE JOURNAL OF ECONOMICS AND ALLIED SCIENCES. CJEAS.
https://www.researchgate.net/publication/366275744
