




































American Research Journal of Economics, Finance and Management 

Volume 12 Issue 1, January-February 2024 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https:// https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 

 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

1 | P a g e  

FROM ENERGY TO ECOLOGY: NAVIGATING ENVIRONMENTAL 

CONSEQUENCES IN NIGERIA'S ECONOMIC LANDSCAPE 
 

 
1Chijioke Emmanuel Okonkwo, 2Ngozi Chioma Nwachukwu 

1,2Department of Economics, Alvan Ikoku University of Education, Owerri Imo State, Nigeria 

DOI: https://doi.org/10.5281/zenodo.10619098 

 

Abstract: Energy commodities, spanning a spectrum from nuclear and chemical to mechanical, 

thermal, radiation, and electrical forms, play a pivotal role in fostering economic growth through 

heightened productivity and increased employment opportunities. Despite the diversity of these 

energy manifestations—liquids, solids, and gases—their environmental ramifications present a 

complex landscape. Notably, the combustion of fossil fuels, a primary source of energy, is implicated 

in contributing to the global warming phenomenon. This study delves into the multifaceted 

dimensions of energy production, transportation, and consumption, unraveling the intricate web of 

environmental consequences that invariably result from these processes. 

Acknowledging the indispensable role of energy in economic development, the research explores the 

nexus between energy commodities and their environmental impact. Fossil fuel combustion, a major 

driver of energy generation, is scrutinized for its substantial role in climate change. The study 

investigates the interplay of various energy forms, analyzing their distinct environmental footprints 

and the broader implications for sustainable development. By comprehensively examining the 

production, transportation, and consumption of energy commodities, this research aims to provide a 

nuanced understanding of the environmental challenges posed by these essential components of 

modern economies. 

Keywords: Energy Commodities, Environmental Consequences, Fossil Fuel Combustion, 

Sustainable Development, Economic Growth  

 

 

INTRODUCTION   

Energy commodities, encompassing nuclear, chemical, mechanical, thermal, radiation, and electrical 

energy, contribute to economic growth by enhancing productivity and employment. They exist in 

various forms—liquids, solids, and gases—yet their environmental impact is intricate, particularly with 

the combustion of fossil fuels contributing to global warming. The production, transportation,    and     

consumption     of     energy    almost invariably  result  in  significant  environmental 

consequences.  

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 12 Issue 1, January-February 2024 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https:// https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 

 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

2 | P a g e  

The consumption of fossil fuels can lead to localized air pollution and climate change (Han et al., 2019). 

Recent research by the WEF (2022) indicates that certain pollutants related to fossil fuels actually have 

a cooling effect. According to Bölük and Mert (2015), natural gas is less aggressive than oil, accounting 

for only half of the CO₂ emissions compared to coal. On average, the combustion of oil (in the form of 

petroleum) releases approximately 33% less carbon dioxide (CO₂) per unit of energy produced 

compared to the combustion of coal. 

  

 

   
 Figure 1. Carbon dioxide emissions, energy consumption and economic growth.  

In many emerging nations like Nigeria, fossil fuels remain the primary source of energy (Sugiawan and 

Managi, 2019). Despite their numerous benefits, such as providing thermal power plants with more 

precise operational control and monitoring (Vincent and Ezaal, 2022), these systems face various 

challenges that have been extensively studied. Robinson et al. (2007) note that Nigeria is not an 

exception to the escalating environmental concerns. Okafor and Joe-Uzoegbu (2010) emphasize the 

environmental impact of urbanization in Nigeria, where rural communities rely on traditional biomass 

for energy, resulting in greenhouse gas emissions. This imbalance contributes to global warming and 

environmental degradation, with Nigeria experiencing some of the highest CO₂ emissions worldwide.  

Elevated carbon dioxide emissions are primarily associated with economic growth, as proposed by the 

Environmental Kuznets (1955) Curve (EKC) hypothesis and supported by studies such as those 

conducted by Han et al. (2018), Acheampong (2018), Abbas et al. (2019), and Esmaeili et al. (2023). 

The EKC theory posits that income contributes to environmental degradation in the early stages of 

development but diminishes once certain income levels are reached. However, the EKC exhibits diverse 

shapes, suggesting different policy implications. The validity of this hypothesis is debated due to 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 12 Issue 1, January-February 2024 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https:// https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 

 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

3 | P a g e  

variations in methodology, independent variables, examined sectors, and modeling. Few empirical 

studies focus on Nigeria, and no recent dataset has been employed to analyze Nigerian CO₂ emissions, 

unlike the studies by Chuku (2011), Ogundipe (2013), Alege and Ogundipe (2013), and Okon (2021). 

To our knowledge,  as  of  the  time  of  writing,  no paper concentrating on Nigerian CO₂ emissions has 

utilized a more recent dataset.  

Nigeria's economy remains susceptible to the risks associated with climate change due to the country's 

escalating energy consumption and the ensuing CO₂ pollution. Figure 1 illustrates the shift in energy 

consumption from negative to positive after 1995, resulting in a 7.4% increase in CO₂ emissions in 1996 

compared to the negative rates of 24.1% and 6.7% in 1994 and 1995, respectively. Concurrently, due to 

the rising energy consumption, GDP per capita rose from 18.9% in 1995 to 27.03% in 1996, and the 

overall economy expanded from -0.1% in 1995 to 4.2% in 1996. The year 2010 marked the most 

significant change in CO₂ emissions (46%) and the highest GDP per capita (19.3%) during the research 

period, with an 8% expansion in the economy. Conversely, the lowest changes occurred in 1989, with a 

-0.35% change in energy consumption and a -41.7% change in CO₂ emissions. Although both CO₂ 

emissions and GDP deviated from their 1981 values, the growth rate of CO₂ emissions exceeded that of 

GDP. This suggests limited evidence of absolute decarbonization, indicating that the nation's CO₂ 

emissions were not proportional to economic growth. Figure 1 demonstrates that from 1981 to 2021, 

Nigeria's economy did not follow a low-carbon trajectory. Policymakers need to comprehend the 

directional and causal relationship between energy commodities, economic growth, and the 

environment.  

The increase in CO₂ emissions over the past 70 years is also attributed to the expansion of the human 

population. A growing population results in increased demands for commodities, energy, and food, 

leading to higher emissions from transportation, industry, and agriculture. However, to minimize 

emissions per person, additional measures such as enhancing energy efficiency, transitioning     to      

renewable      sources,     and     altering consumption habits must be implemented alongside population 

policies (The Conversation, 2023). Globally, CO₂ emissions are distributed unevenly, with high- and 

upper-middle-income countries, housing slightly less than half of the world's population, responsible 

for over 80% of global CO₂ emissions. According to Our World in Data (2023), the average person in 

high-income countries emits over ten times as much CO₂ as the average person in lowincome countries.  

This study addresses several gaps in existing literature. Firstly, it focuses on Nigeria from 1981 to 2021, 

as data before the 1980s are incomplete. Secondly, it employs the Vector Error Correction Model 

(VECM) to explore longitudinal cointegration and causal links between the environment, energy 

commodities, and economic growth, providing fresh empirical data for the ongoing discussion about 

their relationship.   

 LITERATURE REVIEW   

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 12 Issue 1, January-February 2024 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https:// https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 

 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

4 | P a g e  

The links between energy consumption, CO₂ emissions, and economic growth are the subject of three 

broad genres of literature. The first discusses whether the relationship between economic growth and 

CO₂ emissions is consistent with the environmental Kuznets curve (EKC) theory. According to this 

theory, some pollutants and percapita income have an inverted U-shaped connection (Grossman and 

Krueger, 1995). The Environmental Kuznets Curve (EKC) is a relationship between income change and 

environmental quality, based on Kuznets' work. It suggests that rapid industrialization leads to 

increased pollution and resource use, putting pressure on the environment. As income increases, people 

value the environment more, leading to a decline in pollution levels. The EKC hypothesis reveals how 

environmental quality changes as a country's fortunes change, with an inverted U-shaped curve when 

pollution indicators are plotted against income per capita (Dinda, 2004). Richer consumers put more 

pressure on lawmakers to enact environmental laws and regulations, in addition to being prepared to 

spend more money on eco-friendly goods. The majority of the examples where emissions have 

decreased while income has increased can be attributed to institutional reforms at the local and national 

levels, including environmental laws and market-based incentives aimed at halting environmental 

degradation.  

In their study, Özokcu and Özdemir (2017) verified the "inverted U shape theory." However, Friedl and 

Getzner (2003) hypothesize a long-term link that takes the form of an N or another shape rather than 

an inverted U between CO₂ emissions and per-capita income. Although He and Richard (2010) and 

Agras and Chapman (1999) maintain that there is no correlation between CO₂ emissions and economic 

growth in their non-existence theory, the primary issue with these early investigations on the EKC 

hypothesis   is   that   they   may    be    biased    by   missing variables. This happens when one or more 

independent variables that correlate with one or more of the included independent variables and have 

an impact on the dependent variable are excluded from a statistical model (Tong et al., 2020).  

Recently, the Granger causality test, an econometric technique particularly well-suited for time series 

and panel data analyses, was proposed to examine the connection between economic growth and 

carbon emissions. For example, Hossain (2012) discovered that in newly industrialized nations, there 

was unidirectional short-run causality between economic growth and carbon dioxide emissions, as well 

as between urbanization and economic growth. Wang et al. (2016) discovered that economic growth 

was a Granger cause of CO₂ emissions in China between 1995 and 2012, and Hamit-Haggar (2012) 

found a unidirectional causality relationship between the economy and greenhouse gas emissions in 

both the short and long runs in their investigation of the Canadian industrial sector. In contrast to Omri 

(2013), who discovered only a one-way Granger causation linking CO₂ emissions to economic growth 

in some European, Central Asian, Latin American, and Caribbean countries, Salahuddin and Gow 

(2014) showed a two-way Granger causal association between the two components.  

Abubakar and Cudjoe (2021) estimate the short-run and long-run impacts of energy consumption on 

Nigeria's environment through total CO₂ emissions using error correction models and normalized 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


American Research Journal of Economics, Finance and Management 

Volume 12 Issue 1, January-February 2024 

ISSN: 2836-9416 

Impact Factor: 5.57 

Journal Homepage: https:// https://americaserial.com/Journals/index.php/ARJEFM, 

Email: contact@americaserial.com 

Official Journal of America Serial Publication 

 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

5 | P a g e  

estimations. Results show that GDP has a significant long-run tendency to reduce total CO₂ emissions 

in Nigeria, confirming the Kuznets curve hypothesis for climate. The research also supports the 

suggestion that environmental destruction increases with per capita income during early economic 

development stages and decreases with an increase after reaching a plateau. Rafindadi (2016) modeled 

the relationships between economic development, energy use, and emissions. The model had 

collinearity issues because the study takes into account CO₂ emissions as a function of income, income 

squared, and income cubed in addition to other explanatory variables like energy consumption.  

In his paper, Okon (2021) used the auto-regressive distributed lag approach to investigate the 

applicability of the Environmental Kuznets Curve in Nigeria from 1970 to 2018. According to the 

bounds test, there exists an equilibrium relationship over a long period of time between the gross 

domestic product per capita, the square of the GDP per capita, waste, combustible renewable energy, 

alternative and nuclear energy, adjusted savings, or net forest depletion. However, neither short-run 

nor long-run results are consistent with the Environmental Kuznets Curve hypothesis, nor there is no 

evidence of an inverse U-shaped link between growth and fluorinated greenhouse gas emissions in 

Nigeria. Omisakin (2009) tested the Environmental Kuznets Curve (EKC) hypothesis in Nigeria finding 

no long-term causal relationship between carbon emissions and income. The regression line shows an 

"U-shaped" pattern, suggesting that income increases carbon emissions before rising again. Other 

studies have shown a long-term relationship between environmental pollution indicators, per capita 

income, institutional variables, and trade. Alege and Ogundipe (2013) found no EKC in Nigeria due to 

its early development stages. Egbetokun et al.'s (2020) study found that SPM and CO₂ have an EKC, 

while other environmental contamination measures did not significantly affect economic development. 

mailto:contact@americaserial.com
mailto:contact@americaserial.com


Ihugba et al.          6  

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

6 | P a g e  

Table 1. Variables measurement and sources of data.  

S/N  Variable  Measurement  
Expected 

sign  
Sources of data  

1  

carbon 

 dioxid

e CO₂ 

 emissi

ons per 

capital  

  

Annual CO₂ emissions per capital 

measures how much Nigeria emits 

from fossil fuels and industry divided 

by its population in a given year  

  

  

  

https://data.worldbank.org/ 

indicator/EN.ATM.CO₂E.P 

C?locations=NG  

  

2  

Gross 

 domest

ic product 

 per  

capita 

(GDPPC)  

It analyzes Nigeria's GDP per capita 

and gauges the prosperity of 

Nigerians by looking at our GDP 

growth. It is computed by dividing the 

nation's GDP by its total population.  

We expect a positive relationship 

between the variables  

+  

https://data.worldbank.org/ 

indicator/NY.GDP.PCAP. 

KN?locations=NG  

  

  

3  

  

  

Gross  fixed  

capital 

formation  

  

  

Measure capital stock, this study 

employed the Gross fixed capital 

formation, which is essential to any 

country’s economic growth. We 

expect a positive relationship between 

the variables  

  

  

+  

  

  

Central bank of Nigeria 

(CBN) statistical bulletin 

volume 32, December 2021  

  

4  

  

Fossil fuel 

energy 

consumption 

(% of total)  

  

It refers to the use of petroleum, 

natural gas, and coal as sources of 

energy. We expect a positive 

relationship between the variables  

  

+  

  

https://data.worldbank.org/ 

indicator/EG.USE.COMM. 

FO.ZS?locations=NG  

  

5  Population  

The total population of Nigeria during 

the different study years, expressed in 

millions. An increase in population 

will result in more land being cleared 

for agriculture, business, or other 

uses, as well as more energy use 

(fossil fuel). Global CO₂ emissions are 

greatly increased by these activities. 

Therefore, a positive correlation 

between the variables is what we 

anticipate  

+  

https://data.worldbank.org/ 

indicator/SP.POP.TOTL?v 

iew=chartandlocations=N 

G  

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

7 | P a g e  

 Source: Researcher’s compilation, 2023.  

METHODOLOGY  

Granger causality, a linear regression model, and cointegration tests will all be used to examine the 

collected data. Using the unit root test, the initial step will be to determine the stationarity and order of 

integration among the study variables. Using cointegration tests, the second phase will look at the long-

term relationship between the research variables. In the third, regression analysis will be used to 

examine how the independent study variables affect CO₂ emissions. The Granger causality test will be 

used in the fourth step to determine the causal relationship between the research variables.  

Data and sources   

 The Central Bank of Nigeria's (CBN) statistical bulletins and the World Bank 2023 Development 

Indicators (WDI) provided the data used in the study's empirical analysis. Table 1 lists the variables' 

names, meanings, and measurements.  

Model specification  

Based on the empirical literature in energy economics, it makes sense to write the long-term 

relationship between CO₂ emissions, energy commodities, and economic growth in the form of a linear 

logarithmic quadratic. This will allow us to test the EKC hypothesis in the following way (Equation 1):  

 CO2 = (FFC, GDPPC, GFCF, POP)                                                              (1)   

In this case, POP stands for population, GFCF for gross fixed capital formation, GDPPC for gross 

domestic product per capita, FFC for fossil fuel energy consumption, and C02 for carbon dioxide 

emissions per capita. Equation 2 can be expressed as follows in the natural log form for C02, GDPPC, 

GFCF, and POP as well as in the econometric model:  

 LCO2𝑡� = 𝛽�0 + 𝛽�1FFC𝑡� + 𝛽�2LGDPPC𝑡� + 𝛽�3LGFCF𝑡� + 𝛽�4LPOP𝑡� + 𝜀�𝑡�     (2)  

 Stationarity test  

 Since it can affect a series' behavior, stationarity is a significant phenomenon. Regressing X on Y in 

Equation (3) will result in spurious or gibberish regression if X and Y are two non-stationary series 

(Yule, 1926).  

 𝑌�𝑡� = 𝛽�0 + 𝛽�1𝑋�𝑡� + 𝜀�𝑡�                                                                                            (3)   

The series is considered non-stationary if it has a unit root. The series is stationary if it doesn't have a 

unit root. The purpose of the stationarity test is to determine if an autoregressive model has a unit root 

or not. To  ascertain  the  sequence  of  the  variables' integration, the unit root test is helpful. To verify 

if the provided series is stationary, the augmented Dickey-Fuller test (ADF) and PhillipsPerron test (PP) 

have been employed.  

Co-integration test   

 To determine whether there is a co-integration relationship between the two variables' non-stationary 

series, the Johansen-Juselius test is used. We can determine whether there is co-integration between 

two non-stationary series using the Johansen-Juselius co-integration procedure. This indicates that 

0<rank (π) = r <n, which is the maximum rank that the matrix π can have. In terms of the vector or 

matrix of adjustment parameters 𝛼� and the vector or matrix of cointegrating vectors 𝛽�′, π can be 

expressed as =𝛼�𝛽�′, where (r) is the number of co-integration vectors and (n) is the number of variables. 

Based on a likelihood ratio test (LR), this procedure uses the Trace test and the Maximum Eigenvalues 

test (λ𝑚�𝑎�𝑥�) to calculate the number of co-integration vectors between variables. The definition of a 

trace test is:  n ˆ ) 

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

8 | P a g e  

trace(r) =−T i=r+1log(1− i   

 In contrast to the alternative hypothesis, which states that there are r co-integration vectors, the null 

hypothesis states that there are ≤r co-integration vectors.   

 The  Maximum  Eigenvalues  test ( max )is  defined  as:  

max (r,r+1) =−T log(1− r+1) The null hypothesis that the number of co integration vectors = r 

against the alternative those they r+1.  

Granger-causality  

 The results of the stationarity and co-integration tests will determine the Granger-Causality test's 

application in the following ways: The following vector auto-regression (VAR) should be estimated in 

order to perform the conventional Granger-Causality test to determine whether the series (FFC), 

(LGDPPC), (LGFCF), (LPOP), and (LCO₂) are stationary.  

  𝑚� 𝑚� 

𝐹�𝐹�𝐶�𝑡� = 𝛼� + ∑ 𝛽�1𝑖�𝐹�𝐹�𝐶�𝑡�−1 + ∑ 𝛽�2𝑖�𝐿�𝐶�𝑂�2𝑡�−𝑖� + 𝜀�                                 (4)  

 𝑖�=1 𝑖�=1 

  𝑚� 𝑚� 

𝐿�𝐶�𝑂�2𝑡� = 𝛼� + ∑ 𝛽�3𝑖�𝐿�𝐶�𝑂�2𝑡�−1 + ∑ 𝛽�4𝑖�𝐹�𝐹�𝐶�𝑡�−𝑖� + 𝜀�                              (5)  

 𝑖�=1 𝑖�=1 

   𝑚� 𝑚� 

𝐿�𝐺�𝐷�𝑃�𝑃�𝐶�𝑡� = 𝛼� + ∑ 𝛽�5𝑖�𝐿�𝐺�𝐷�𝑃�𝑃�𝐶�𝑡�−1 + ∑ 𝛽�6𝑖�𝐿�𝐶�𝑂�2𝑡�−𝑖� + 𝜀�                (6)  

 𝑖�=1 𝑖�=1 

  𝑚� 𝑚� 

𝐿�𝐶�𝑂�2𝑡� = 𝛼� + ∑ 𝛽�7𝑖�𝐿�𝐶�𝑂�2𝑡�−1 + ∑ 𝛽�8𝑖�𝐿�𝐺�𝐷�𝑃�𝑃�𝐶�𝑡�−𝑖� + 𝜀�                    (7)  

𝑖�=1 𝑖�=1 

  𝑚� 𝑚� 

𝐿�𝐺�𝐹�𝐶�𝐹�𝑡� = 𝛼� + ∑ 𝛽�9𝑖�𝐿�𝐺�𝐹�𝐶�𝐹�𝑡�−1 + ∑ 𝛽�10𝑖�𝐿�𝐶�𝑂�2𝑡�−𝑖� + 𝜀�                    (8)  

𝑖�=1 𝑖�=1 

    𝑚� 𝑚� 

𝐿�𝐶�𝑂�2𝑡� = 𝛼� + ∑ 𝛽�11𝑖�𝐿�𝐶�𝑂�2𝑡�−1 + ∑ 𝛽�12𝑖�𝐿�𝐺�𝐹�𝐶�𝐹�𝑡�−𝑖� + 𝜀�                    (9)  

 𝑖�=1 𝑖�=1 

  𝑚� 𝑚� 

𝐿�𝑃�𝑂�𝑃�𝑡� = 𝛼� + ∑ 𝛽�13𝑖�𝐿�𝑃�𝑂�𝑃�𝑡�−1 + ∑ 𝛽�14𝑖�𝐿�𝐶�𝑂�2𝑡�−𝑖� + 𝜀�                     (10)  

 𝑖�=1 𝑖�=1 

  𝑚� 𝑚� 

𝐿�𝐶�𝑂�2𝑡� = 𝛼� + ∑ 𝛽�15𝑖�𝐿�𝐶�𝑂�2𝑡�−1 + ∑ 𝛽�16𝑖�𝐿�𝑃�𝑂�𝑃�𝑡�−𝑖� + 𝜀�                     (11)  

 𝑖�=1 𝑖�=1 

 In models (Equations 4 and 11), the subscripts denote time periods and 𝜀� is a white noise error. The 

constant parameter 𝛼� represents the constant growth rate of FFCt in Equation 4;  LCO2t in Equation 5; 

LGDPPCt in Equation 6; LGFCFt in Equation 8 and POPt in Equation 10. We can obtain eight tests from 

this analysis: the first examines the null hypothesis that the FFCt does not Grangercause LCO2t and the 

second test examine the null hypothesis that the LCO2t does not Granger-cause FFCt . The third 

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

9 | P a g e  

examines the null hypothesis that the LGDPPCt does not Granger-cause LCO2t and the fourth test 

examines the null hypothesis that the LCO2t does not Granger-cause LGDPPCt 

. The fifth examines the null hypothesis that the LGFCFt does not Granger-cause LCO2t and the sixth 

test examines the null hypothesis that the LCO2t does not Granger-cause LGFCFt . The seventh 

examines the null hypothesis that the LPOPt does not Granger-cause LCO2t and the eight test examines 

the null hypothesis that the LCO2t does not Granger-cause POPt.    

  Vector error correction model (VECM)  

 The conventional VECM is written compactly as (Equation 12):  

 𝑘�−1 𝑘�−1 𝑘�−1 

𝛥�𝑌� = 𝛼� + ∑ 𝛾�𝛥�𝑌�𝑡�−1 + ∑ 𝜂�𝛥�𝑋� + ∑ 𝜑�𝛥�𝑅� + 𝜆�𝐸�𝐶�𝑇� + 𝜇�                    (12)  

 𝑖�=1 𝑖�=1 1=𝑖� 

 where ECTt−1= OLS residual with a lag derived from the long-run cointegrating formula (Equation 13):  

 𝑌�𝑡� = 𝜎� + 𝜂�𝑗�𝑋�𝑡� + 𝜉�𝑚�𝑅�𝑡� + 𝜇�𝑡�                                                                          (13)  

  And express as (Equation 14):  

 𝐸�𝐶�𝑇�𝑡�−1 = [𝑌�𝑡�−1 − 𝜂�𝑗�𝑋�𝑡�−1 − 𝜉�𝑚�𝑅�𝑡�−1]                                                         (14)  

 = coefficient of the ECT and the speed at which changes to X and R cause Y to stabilize. The specific 

VECM for this study is as follows (Equations 15 to 19):  
 𝑘�  𝑘�  𝑘�  𝑘�

 𝑘�  

𝛥�𝐿�𝐶�𝑂�𝛼�𝑖�𝛥�𝐿�𝐶�𝑂�𝜑�𝑚�𝛥�𝐹�𝐹�𝐶�𝑡�𝜂�𝑛�𝛥�𝐿�𝐺�𝐷�𝑃�𝑃�𝐶�𝑡�𝜙�𝑝�𝛥� 𝐿�𝐺�𝐹�𝐶�𝐹�𝑡�𝜉�𝑞�𝛥�𝐿�𝑃�𝑂�𝑃�𝑡�−1 + 𝜆�𝐸�𝐶�𝑇� + 𝜇�                                 (15)  

 𝑖�  𝑚�  𝑛�  𝑝�  𝑞�  

  

 𝑘�  𝑘�  𝑘�  𝑘�

 𝑘�  

𝛥�𝐹�𝐹�𝐶�𝑡�𝛼�𝑖�𝛥�𝐿�𝐶�𝑂�𝜑�𝑚�𝛥�𝐹�𝐹�𝐶�𝑡�𝜂�𝑛�𝛥�𝐿�𝐺�𝐷�𝑃�𝑃�𝐶�𝑡�𝜙�𝑝�𝛥� 𝐿�𝐺�𝐹�𝐶�𝐹�𝑡�𝜉�𝑞�𝛥�𝐿�𝑃�𝑂�𝑃�𝑡�−1 + 𝜆�𝐸�𝐶�𝑇� + 𝜇�                                    (16)  

 𝑖�  𝑚�  𝑛�  𝑝�  𝑞�  

  

 𝑘�  𝑘�  𝑘�  𝑘�

 𝑘�  

𝛥�𝐿�𝐺�𝐷�𝑃�𝑃�𝐶�𝑡�𝛼�𝑖�𝛥�𝐿�𝐶�𝑂�𝜑�𝑚�𝛥�𝐿�𝐺�𝐷�𝑃�𝑃�𝐶�𝑡�𝜂�𝑛�𝛥�𝐹�𝐹�𝐶�𝑡�𝜙�𝑝�𝛥� 𝐿�𝐺�𝐹�𝐶�𝐹�𝑡�𝜉�𝑞�𝛥�𝐿�𝑃�𝑂�𝑃�𝑡�−1 + 𝜆�𝐸�𝐶�𝑇� + 𝜇�                            (17)  

 𝑖�  𝑚�  𝑛�  𝑝�  𝑞�  

  

 𝑘�  𝑘�  𝑘�  𝑘�

 𝑘�  

𝛥�𝐿�𝐺�𝐹�𝐶�𝐹�𝑡�𝛼�𝑖�𝛥�𝐿�𝐶�𝑂�𝜑�𝑚�𝛥�𝐿�𝐺�𝐹�𝐶�𝐹�𝑡�𝜂�𝑛�𝛥�𝐿�𝐺�𝐷�𝑃�𝑃�𝐶�𝑡�𝜙�𝑝�𝛥� 𝐹�𝐹�𝐶�𝑡�𝜉�𝑞�𝛥�𝐿�𝑃�𝑂�𝑃�𝑡�−1 + 𝜆�𝐸�𝐶�𝑇� + 𝜇�                                (18)  

 𝑖�  𝑚�  𝑛�  𝑝�  𝑞�  

  

 𝑘�  𝑘�  𝑘�  𝑘�

 𝑘�  

𝛥�𝐿�𝑃�𝑂�𝑃�𝑡�𝛼�𝑖�𝛥�𝐿�𝐶�𝑂�𝜑�𝑚�𝛥�𝐿�𝑃�𝑂�𝑃�𝑡�𝜂�𝑛�𝛥�𝐹�𝐹�𝐶�𝑡�𝜙�𝑝�𝛥� 𝐿�𝐺�𝐷�𝑃�𝑃�𝐶�𝑡�𝜉�𝑞�𝛥�𝐿�𝐺�𝐹�𝐶�𝐹�𝑡�−1 + 𝜆�𝐸�𝐶�𝑇� + 𝜇�                                   (19)  

 𝑖�  𝑚�  𝑛�  𝑝�  𝑞�  

Table 2. ADF and PP unit root test results.  

 Variable  

 ADF test statistic    PP test statistic   

Constant  
Constant 
and trend  

None  
First 
difference  

Constant  
Constant 
and trend  

None  
First 
difference  

LCO₂  -1.04  -2.14  0.56  -6.59*  -1.03  -2.11  0.62  -6.61*  

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

10 | P a g e  

= -2.94); P-value= Probability value, * signifies stationarity.  

Source: Researcher’s calculations from Eviews 10, 2023.   

This model was selected because empirical studies show that when economic variables show individual 

cointegration, or a strong longterm relationship, the VECM performs well for model estimation. 

Another advantage is its capacity to integrate the short-run dynamic and long-run equilibrium models 

into a single, efficient system. It also guarantees accuracy, conceptual rigor, and data integrity 

(Abubakar and Cudjoe, 2021).  

RESULTS AND FINDING  

 The empirical estimation result and a suitable justification are presented in this section to support the 

study's argument.  

 Stationarity test  

 The study initiated by examining the stationarity of relevant variables using tests detailed in the 

methodology section. The Augmented Dickey-Fuller (ADF) and Phillips and Perron (PP) unit root tests 

were employed to determine if the variables are stationary.  Table  2  presents  compelling evidence that 

all our variables are integrated at order one (that is, I(1)). The data reveals that for each variable, at 

least one of the tests does not reject the null hypothesis of the unit root at levels, indicating non-

stationarity. In contrast, it is found that every variable in the first difference is stationary. The 

subsequent step involves confirming whether our variables of interest exhibit a long-run relationship 

since all the variables in our model are integrated of order one, according to at least one of the tests 

employed.   

  Endogeneity analysis   

 In order to ascertain whether variables are exogenous or endogenous, Endogeneity analysis is 

necessary. To verify it, apply the paired Granger causality test. Table 3 displays the outcomes of the 

pairwise Granger causality tests. The null hypothesis is rejected at F-statistic critical values of 1, 5, and 

10%. First, the study indicates that fossil fuel consumption does not Granger-cause CO₂ emissions in 

Nigeria based on the pairwise Granger causality test.   

*indicates lag order selected by the criterion. Source: Researcher’s calculations from Eviews 10, 2023.  

 Secondly, CO₂ emissions in Nigeria, a proxy for the environment, do not have any feedback from gross 

domestic product per capita, a proxy for economic growth, and instead Granger-cause it. Third, gross 

fixed capital formation is a proxy for investment without feedback and population growth, Granger-

cause CO₂ emissions in  

FFC  -2.82  -3.23  -0.41  -6.55*  -2.85  -3.32  -0.60  -8.08*  

LGDPPC  -1.20  -1.88  1.72  -4.00*  -0.48  -3.06  0.66  -4.00*  

LGFCF  -0.75  -2.01  1.29  -4.96*  -0.87  -0.74  2.37  -4.94*  

LPOP  -1.64  0.33  2.73  -3.18*  -1.64  0.33  12.88  -3.18*  
  

ADF:  Test critical values at 5% (At level: constant = -2.94, Constant and trend = -3.54, none = -1.95 while at First difference 

= -2.95); P-value= Probability value, * signifies stationarity. PP:  Test critical values at 5% (At level: constant = -2.94, 

Constant and trend = -3.53, none = -1.95 while at First difference  

 
 
 
 

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

11 | P a g e  

Table 3. Pairwise granger causality test (Lags: 3).  

 Null Hypothesis:  Obs  F-Statistic  Prob.  

 FFC does not Granger Cause LCO₂  38  1.5097  0.2314  

 LCO₂ does not Granger Cause FFC    1.37356  0.2691  

 LGDPPC does not Granger Cause 

LCO₂  

38  0.55583  0.6481  

 LCO₂ does not Granger Cause 

LGDPPC  

  3.17565  0.0378**  

 LGFCF does not Granger Cause LCO₂  38  2.0334  0.1296  

 LCO₂ does not Granger Cause LGFCF    1.4502  0.2472  

 LPOP does not Granger Cause LCO₂  38  3.28918  0.0336**  

 LCO₂ does not Granger Cause LPOP    0.51413  0.6756  

 LGDPPC does not Granger Cause FFC  38  2.00407  0.1339  

 FFC does not Granger Cause LGDPPC    0.75542  0.5277  

 LGFCF does not Granger Cause FFC  38  1.96521  0.1397  

 FFC does not Granger Cause LGFCF    1.04618  0.386  

 LPOP does not Granger Cause FFC  38  1.68859  0.1898  

 FFC does not Granger Cause LPOP    0.11114  0.9529  

 LGFCF does not Granger Cause 

LGDPPC  

38  1.5466  0.2221  

 LGDPPC does not Granger Cause 

LGFCF  

  5.68459  0.0032*  

 LPOP does not Granger Cause 

LGDPPC  

38  0.44513  0.7225  

 LGDPPC does not Granger Cause 

LPOP  

  0.62673  0.6032  

 LPOP does not Granger Cause LGFCF  38  15.0869  0.00*  

 LGFCF does not Granger Cause LPOP    1.36982  0.2703  

 *Causality at 1 % critical level; ** Causality at 5 % critical level. Source: Researcher’s calculations 

from Eviews 9, 2023.  

   Table 4. VAR Lag order selection criteria.  

  Lag  LogL  LR  FPE  AIC  SC  HQ  

0  36.54388  NA  1.31e-07  -1.660204  -1.444732  -1.583541  

1  246.0650  352.8776  8.03e-12  -11.37184  -10.07901*  -10.91186*  

2  267.2214  30.06440  1.06e-11  -11.16955  -8.799356  -10.32625  

3  308.1256  47.36277*  5.55e-12*  -12.00661*  -8.559060  -10.78000  
 

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

12 | P a g e  

Nigeria. Fourth, gross fixed capital formation, a stand-in for investment that lacks feedback, Granger-

causes gross domestic product per capita (a proxy for economic growth).  

Lag selection  

 The Vector Error Correction Model (VECM), the Phillips and Perron (PP), the Augmented Dickey-

Fuller (ADF), and the co-integration tests are sensitive to the number of lags when they are run. Thus, 

the Schwarz Information Criterion (SIC) and Akaike Information Criterion (AIC) were used to 

determine the actual amount of lags used. Table 4 displays the proper lag length for each variable. Table 

4 shows that the AIC value at lag 3 is the lowest and is likewise lower than the SIC value at lag 1. To 

estimate Equation (1), the model (Lag 3) is selected as a result. Below is the cointegration result. 

Cointegration test  

The relevant hypothesis is that there is no long-run relationship in order to ascertain whether the 

variables are cointegrated over the long term, such as:  

Hypothesis: λ1 = λ2 = λ3 = λ4 = 0 (no long-term association exists).  

Hypothesis 1: λ1 ≠ λ2 ≠ λ3 ≠ λ4 ≠ 0 (a long-term  

Relationship exists)  

 The next step is to run a cointegration test after confirming that all variables are integrated to order 

one and that I(1) cannot be refused. Johansen (1988) and Johansen and Juselius (1990) suggested the 

multivariate cointegration technique, which is used with multivariate time series to find stable long-

term links between carbon dioxide emissions, GDP per capita, gross fixed capital formation, energy use 

from fossil fuels, and population. Because the cointegration vectors will be used for the subsequent 

vector error correction model (VECM), it should be emphasized that the cointegration test is conducted 

before the VECM.   

Table 5. Cointegration results.  

 

Hypothesiz

ed  

Trace  0.05    Hypothesized  Max-

Eigen  

0.05  Prob.**  

No. of CE(s)  Statistic  Critical 

Value  

Prob.**  No. of CE(s)  Statistic  Critica

l Value  

  

None *  115.2717  69.8188

9  

0.0000  None *  46.24334  33.876

87  

0.0011  

At most 1 *  69.02839  47.8561

3  

0.0002  At most 1 *  31.44267  27.584

34  

0.0151  

At most 2 *  37.58572  29.7970

7  

0.0052  At most 2 *  22.25237  21.131

62  

0.0347  

At most 3  15.33335  15.4947

1  

0.0529  At most 3  12.49434  14.264

60  

0.0934  

At most 4  2.839018  3.84146

6  

0.0920  At most 4  2.839018  3.8414

66  

0.0920  

Source: Researcher’s calculations from Eviews 9, 2023. * Denotes rejection of the null hypothesis at 

the 0.05 level.  

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

13 | P a g e  

The Johansen test, for instance, rejects the existence of one or fewer cointegrating relations but fails to 

reject the existence of at most two in model 1. This suggests that there are two cointegrating equations 

in model 1.  

Using trace test statistics, the null hypothesis is rejected because the probability value is less than 5% 

(P-value = 0.00) and the trace statistic value is greater than the critical value (115.2717>69.81889). This 

suggests the existence of at least one cointegrating vector. A second evaluation states that since the trace 

statistic value is higher than the essential values, we reject the null hypothesis for asterisks ranked one 

through two. Every associated probability value is less than five percent. In summary, the results 

indicate that two equations are cointegrated to order one (1) at the 0.05 critical level, and there is at 

least one cointegrating vector.  

Based on the Max-Eigen results, the null hypothesis that there are no cointegrating equations is likewise 

rejected. This is because the probability value is less than 5% (Pvalue = 0.00) and the Max-Eigen 

Statistic is bigger than the important value (46.24334>33.87687). This suggests the existence of at least 

one cointegrating vector. Based on the trace statistical test and the Max-Eigen test, the series are 

cointegrated to the same order (1), as seen in Table 5. This study also takes advantage of a series that 

has a long-standing relationship. This study will estimate the VECM using trace value statistics since it 

offers a more accurate alternative hypothesis that specifies the number of cointegrating vectors. 

Consequently,  one  may  contend that there is a long-term relationship between the variables and that 

both their short- and long-term dynamics can be found using the VECM model.  

Vector Error Correction Model (VECM) Estimation  

Using the same variables, two distinct vector autoregression models (VAR and VEC) were made to 

determine which one more accurately captured the relationship between Nigeria's economic growth, 

energy commodities, and environmental factors in the real world. Despite not being as structural as the 

VAR, the VEC model functioned well as a limited substitute. Meanwhile, as Table 5 illustrates, the 

cointegration relationship between the variables made the VAR ineffective. The optimum model to 

apply in this situation is the Vector Error Correction Model (VECM). Table 6 displays the outcomes of 

the vector error correction model (VECM) for the cointegrated series' first, second, and third 

differences. It also includes the error-correction terms from Equation 20. The results are displayed in 

two sections: the first section displays the cointegrating equations, and the second section displays the 

outcomes of the vector error correction models. Table 7 displays the regression's result. The target 

equations D(LCO₂), D(FFC), and D(LGFCF) have error correction terms that are negative (-0.26), 

0.62), and (-0.29), respectively, according to Table 6 above, but D(LGDPPC) and D(LPOP) have 

positive (0.01) and (0.01) error correction terms, respectively. You can see that the VEC model can 

explain about 69% of the changes in the variables that you can depend on, 43% of the changes in the 

target variable D(LCO₂), and 48% of the changes in the D(FFC), D(LGDPPC), D(LGFCF), and D(LPOP) 

equations. This suggests that all five models fit the data.  

Using the VECM approach, VAR generated and computed a simultaneous equation in Table 6. On the 

other hand, the simultaneous equation computed under VAR using the VECM technique only yields 

coefficients, standard errors, and t-statistics; probability values are absent. Therefore, in order to 

evaluate the relationship between the environment, energy commodities, and economic  growth  in  

Nigeria,   the   simultaneous  equation  must be evaluated. This is due to the fact that within-group 

designs and two samples are the ideal settings for the tstatistic's application. This makes the 

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

14 | P a g e  

simultaneous model of t-statistic-based result interpretation insufficient. Second, t-statistics are 

inappropriate for sample sizes greater  than   or   equal   to   30  (n ≥ 30).  The  independent variables 

are not homogeneous; they have variations for both groups (Engle and Granger, 1987). The impact of 

the explanatory factors on Nigeria's carbon dioxide emission is estimated using the simultaneous 

equation in the study using ordinary least squares (OLS).  

The error  correction term (ECT) in Table 7 indicates the  

 

 

 

 

Table 8. Breusch-Godfrey serial correlation LM test.  

 Source: Researcher’s calculations from Eviews 9, 2023.  

Figure 2. Normality test.  

Rate of correction of the disequilibrium between the longrun and short-run estimations. The value 

indicates that only about 26% of errors generated in the previous period are corrected in the current 

period for the equation. With a p-value of 0.06 at a 5% confidence level and a standard error of 

0.133009, this value is significant.  

Model checking  

The null hypothesis (H0) is accepted if the probability is more than 5%, indicating no serial correlation 

in the longrun model in Table 8. The normality test shows a kurtosis of 2.81 and skewness of 0.36, 

indicating normal  

Figure 3. Plot of CUSUM. distribution. The heteroscedasticity test shows continuous   variance, 

indicating continuous variance. The stability test  shows the Cusum of squares plots do not pass the 5%  

critical line, indicating the model is stable and suitable for economic study.  

 Autocorrelation Residual LM Test  

  

  

 

 

 

 

 

 

 

 

Test for normality  

 A normal model is indicated by residual skewness and kurtosis, and confirmed by JB test (Figure 2).  

Test for stability     

Figure 4. Plot of CUSUMSQ.  The  Figures  3  and  4 show the results of the stability tests Ihugba et 

al.           

Table 9. Breusch-Pagan-Godfrey tests for Heteroscedasticity.  

 F-statistic  0.892884  Prob. F (20,16)  0.6001  

F-statistic  

Obs*R-squared  

0.205142  

1.292661  

Prob. F (3,17)  0.8914  

Prob. Chi-Square 

(3)  

0.7309  

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

15 | P a g e  

Obs*R-squared  19.51504  Prob. Chi-Square 

(20)  

0.4886  

Scaled explained 

SS  

4.024741  Prob. Chi-Square 

(20)  

1.0000  

Source: Researcher’s calculations from Eviews 9, 2023.  

Table 10. ARCH tests for Heteroscedasticity.  

F-statistic  

Obs*R-squared  

0.318784  

1.047440  

Prob. F (3,31)  0.8117  

Prob. Chi-Square 

(3)  

0.7898  

Source: Researcher’s calculations from Eviews 9, 2023.  

Table 11. Wald tests and short-run test.  

 Dependent variable: 

DLCO₂  

  

Variables  Chi-square 

test  

Prob.  Relationship  

D(FFC)  11.58  0.00  Short-run 

causality  

D(LGDPPC)  5.84  0.05  Short-run 

causality  

D(LGFCF)  5.20  0.07  No Short-run 

causality  

D(LPOP)  6.77  0.03  Short-run 

causality  

ALL  25.41  0.00  Short-run 

causality  

 Source: Researcher’s calculations from Eviews 9, 2023.  

 (CUSUM and CUSUMSQ). They show that the estimates, variance, residuals, and square residual are 

stable because they are all within the 5% critical boundaries for both the CUSUM and the CUSUMSQ.  

Both the CUSUM and CUSUMSQ tests accept parameter stability as one of their null assumptions 

(Tables 9 and 10).  

 Simultaneous equation short-run simulation and analysis  

 The results of the short-run test are presented in Table 11. The Chi-square joint statistics probability 

values show that, aside from LGFCF, there is a short-run relationship between the explanatory variables 

and the independent variable according to our findings in Table 11. If the p-value of the chi-square test 

for (FFC) fossil fuel energy consumption, (LGDPPC) gross domestic product per capita, and (LPOP) 

annual total population is less than 0.05, the null hypotheses (𝐻�0): β5=0 will be rejected, therefore 

they cause LCO₂ in the short run, while (LGFCF) gross fixed capital formation as a proxy for investment 

does not cause LCO₂ in the short run. The VECM systems Granger causality tests results dos not 

conform to the Pairwise Granger causality tests except for LGDPPC and  

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

16 | P a g e  

LPOP.  The   next   step   is  to  conduct exante  forecasting involving impulse response and variance 

decomposition tests.  

Impulse response function   

According to Table 12, Nigeria's carbon dioxide emissions forecast are on the positive side, with 

sporadic variations brought on by innovations and shocks. The findings demonstrate that (FFC) fossil 

fuel energy consumption, (LGDPPC) gross domestic product per capita, and (LGFCF) gross fixed 

capital formation will all contribute to explaining the country's increased carbon dioxide emissions. A 

one-standard deviation positive own shock will result in a short-term change from 0.14 to 0.07 and a 

long-term increase at a decreasing rate to 0.037. Second, projections indicate that the energy 

consumption of fossil fuels (FFC) has a short-term negative impact on carbon dioxide emissions (-0.01) 

and a long-term positive impact (0.01). This indicates that FFC has a long-term beneficial effect on 

carbon dioxide emissions.  

Third, the simulation shows that in the short run, carbon dioxide emissions will rise by 0.015 in 

response to a onepositive standard deviation shock from (LGDPPC) gross domestic product per capita. 

The shocks will ultimately be negative (-0.021). Fourth,  innovations  for  (LGFCF)  gross   fixed capital 

formation result in higher carbon dioxide emissions over a five-year period. Simulations show that 

carbon dioxide emissions rise by 0.034 in the short term and 0.032 in the long term for every standard 

deviation increase in LGFCF. Accordingly, the amount invested has a significant impact on carbon 

dioxide emissions. Fifth, projections indicate that, despite both short- and long-term declines, Nigeria's 

annual population will not be a cause for concern when it comes to carbon dioxide emissions. 

 Variance decomposition  

 To predict the error variance effects for each endogenous variable in a system, variance decomposition 

is used. Any change in time in a simple linear equation corresponds to a change in the dependent 

Table 12. Impulse response analysis.  

 Response of LCO₂  

 Period  LCO₂  FFC  LGDPPC  LGFCF  LPOP  

 1  0.143419  0  0  0  0  

 2  0.065792  -0.01005  0.014749  0.033842  -0.0955  

 3  0.03158  -0.02308  -0.04123  0.050493  -

0.15446  

 4  0.049118  0.001002  -0.03575  0.008647  -

0.09152  

 5  

  

0.037665  0.011291  -0.02109  0.031796  -

0.08607  

Source: Researcher’s calculations from Eviews (2023).  

Table 13. Variance decomposition of LCO₂.  

  
Period LCO₂ FFC LGDPPC LGFCF LPOP Short-run 85.1 0.1 0.3 1.6 12.9  

 Medium-term  105.0  0.6  1.8  4.5  38.2  

 Long-run  140.2  1.5  5.9  9.6  92.7  

 
Source: Researcher’s calculations from Eviews (2023).  

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

17 | P a g e  

variable (Wickremasinghe 2011). This study's forecast consists of three time periods: short-term (two 

years), medium-term (five years), and long-term (ten years), all based on the Monte Carlo method and 

Cholesky's ordering. LCO₂, FFC, LGDPPC, LGFCF, and LPOP are the outcomes of the variance 

decomposition forecast for endogenous variables.  

In the short run, impulses, innovations, or shocks to carbon dioxide emissions account for 85.1% of 

fluctuations in carbon dioxide emissions. However, carbon dioxide emissions own shock fluctuations 

continuously increase to 140.2% in the long run. Meanwhile, shocks to fossil fuel energy consumption 

account for 0.1% of fluctuations in carbon dioxide emissions in the short run. The fluctuations in carbon 

dioxide emissions due to fossil fuel energy consumption increase in the long run to 1.5%. In the short 

run, shocks to gross domestic product per  capita  account for 0.3%, gross fixed capital formation 

accounts for 1.6%, and annual population accounts for 12.9%. In the long run, shocks to gross domestic 

product per capita increase to 5.9%, gross fixed capital formation increases to 9.6%, and the annual 

population accounts for 92.7%. Shocks to carbon dioxide emissions will account for the highest 

fluctuations in Nigeria’s carbon dioxide emissions, followed by its own shock (Table 13).  

DISCUSSION   

The primary objective of this paper was to investigate the relationship between CO₂ emissions, energy 

commodities, and economic growth between 1981 and 2021 in Nigeria. The vector error correction 

model (VECM) method was utilized to estimate Equation 2 using annual data. Both the Phillips and 

Perron (1988) and Dickey and Fuller (1981) Augmented Dickey-Fuller (ADF) unit root-testing methods 

are used to check the time series properties of the variables in Equation 1. In Equation 1, all the series 

seem to have a unit root in their levels, but their first differences show that they are stationary.  

The analysis of the data collected has revealed some notable findings that make this study a significant 

contribution to knowledge in the area of carbon dioxide emissions in Nigeria. Firstly, The findings does 

not support the Kuznets Curve (EKC) hypothesis for climate change by demonstrating that economic 

growth as measured by GDP per capita has a significant negative long-run tendency to reduce total CO₂ 

emissions in Nigeria. The adjustment term (-2.11356) in the second year from the estimated  result  in  

Table  7  is statistically significant. The findings indicate that when income rises, emissions fall, and 

vice versa. The World Bank (2018), divides Nigeria's income distribution into five quintiles. In 2018, 

the second 20% of the population held an income share of 11.60%, while the third 20% of the population 

held an income share of 16.20 percent. 22.70 percent of the population, or the fourth 20%, had an 

income share. The richest 20% of the population owned 42.40% of the total income. Nigeria's 2022 

Gini coefficient for nations with high levels of wealth inequality was 35.1%. Nigeria is ranked 100th out 

of 163 countries worldwide and 11th in West Africa with this score (Harmon, 2023).  

This can be as a result of changes in consumption patterns, energy efficiency, technology, or income 

inequality. The findings can also be attributed to institutional reforms at the local and national levels, 

including environmental laws and market-based incentives aimed at halting environmental 

degradation not necessarily increase in income. The finding of the negative effect of income on CO₂ 

emissions agrees with the findings of Friedl and Getzner (2003), He and Richard (2010), Agras and 

Chapman (1999) and Alege and Ogundipe (2013) but disagrees with the findings of Dinda (2004), 

Özokcu and Özdemir (2017), Abubakar and Cudjoe (2021) and Okon (2021). The pairwise Granger 

causality test also indicates that CO₂ emissions in Nigeria, a proxy for the environment, do not have 

any feedback from gross domestic product per capita, a proxy for economic growth, and instead 

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

18 | P a g e  

Granger-cause it. This finding is line with the study of Omisakin (2009), Omri (2013) and Wang et al. 

(2016). Salahuddin and Gow (2014) findings do not agree with our pairwise Granger causality test 

result.  

Secondly, this study found that, in the long run, fossil fuel consumption had a negative effect on CO₂ 

emissions. The results show that fossil fuel use and carbon dioxide emissions are inversely correlated; 

a decrease in fossil fuel use corresponds to a decrease in atmospheric carbon dioxide emissions. Even 

when it is the biggest cause of air pollution in developed countries, It impacts positively in the current 

year but negatively after a year of fossil fuel consumption. The findings suggest that the quantity of 

fossil fuel used for energy production, transport, or industrial processes is still low in the country. In 

2014, the nation's share of global energy consumption from fossil fuels was 18.9%, lower than the 79.4% 

global average for the same year. The pairwise Granger causality test indicates that fossil fuel 

consumption does not cause CO₂ emissions in Nigeria. The findings of the negative effect of fossil fuel 

consumption on CO₂ emissions do not agree with the findings of Abubakar and Cudjoe (2021).  

Thirdly, investments proxied by gross fixed capital formation have the long-term possibility of reducing 

total CO₂ emissions if increased, and finally, the total annual population has a significant positive effect 

on total CO₂ emissions in the long run. This signifies that an increase in total annual population has a 

possibility of rising Nigerian total CO₂ emissions in the future.  Also, the  implication  of  the finding is 

that increasing total annual population threatening Nigeria’s effort to meet the global goal for O2 

emission reduction as outlined in the 2015 Paris Climate Change Conference. This finding agrees that 

the increase in CO₂ emissions during the past 70 years has also been attributed to the expansion of the 

human population.  

Conclusion  

 This paper empirically analyzes the dynamic relationships between CO₂ emissions, energy 

commodities, and economic growth in Nigeria, using CO₂ emissions as a proxy for the environment. 

The long-run relationship, with CO₂ emissions as the dependent variable, is examined to test the short-

run and long-run elasticities of CO₂ emissions with respect to explanatory variables.  

Contrary to the typical positive correlation between income and emissions, indicating higher emissions 

per capita in wealthier nations, our findings reveal the opposite trend. However, this relationship is not 

constant, suggesting that emissions rise at varying rates based on income levels. In high-income 

countries, consumptionbased emissions tend to exceed production-based emissions, while the reverse 

is observed in low-income countries. This implies that high-income countries are net importers of 

emissions, while low-income countries are net exporters.  

The significant impact of fossil fuel usage on the environment is acknowledged in our study. 

Surprisingly, our findings indicate a significantly negative impact on the environment of Nigeria, as 

fossil fuel usage influences the amount of CO₂ emissions. This relationship mirrors the social and 

economic development of the nation. Given the high prices and limited supply of fossil energy in 

Nigeria, insufficient to meet the demands of its over 200 million inhabitants and expanding economy, 

there is a pressing need for a substantial increase in energy efficiency. Additionally, the creation of new 

energy consumption structures, particularly those based on affordable renewable sources like solar 

energy, is essential for sustainable growth over time.  

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

19 | P a g e  

Furthermore, our empirical results challenge the Environmental Kuznets Curve (EKC) theory of climate 

change, which posits that higher income can lower a nation's environmental pollution once a certain 

threshold is reached.  

RECOMMENDATIONS AND POLICY CONSEQUENCES The study's conclusions lead to the 

recommendation that, to mitigate environmental degradation in Nigeria, governments should support 

initiatives educating and training rural residents to use fewer non-renewable energy sources. Despite 

nonrenewable resources being widely utilized   for     fuel,   industrial   production,   and   residential 

energy consumption in Nigeria without currently causing substantial environmental harm, the 

suggestion is for the nation to prioritize energy sources causing minimal environmental damage. 

Policymakers, serious about preventing long-term environmental damage, should enact policies 

promoting the use of environmentally friendly machinery, vehicles, utilities, and equipment.  

Considering fossil fuel consumption has not yet reached a point where CO₂ emissions are increasing, 

Nigerian policymakers should focus more on adopting renewable energy sources to reduce emissions 

from other sources. All energy-related investments and developments in the country should prioritize 

renewable energy and include it as a key performance indicator in investment appraisal considerations.  

While concerns about economic growth are valid, this study suggests a bi-directional causal 

relationship between GDPPC and CO₂ in Nigeria. Policymakers should consider both factors when 

making decisions, emphasizing the need for a comprehensive strategy to boost renewable energy 

investments. This includes creating a stable policy environment, setting ambitious targets for 

renewable energy capacity, providing financial incentives, and implementing feed-in tariffs. The 

government should invest in research and development, workforce development, and public-private 

partnerships, encouraging private sector participation. Risk mitigation instruments should be 

introduced to reduce perceived risks associated with renewable energy projects. Infrastructure 

development should incorporate grid integration and energy storage. Sustainable finance initiatives, 

such as green bonds, public investment, community engagement, and awareness campaigns, should be 

established.  

International support can be leveraged through climate finance, technology transfer, and streamlined 

permitting processes. Performance monitoring should be instituted to ensure projects meet their 

objectives, attracting investors and accelerating project development.  

The study's general conclusions propose that, to reduce poverty and lower CO₂ emissions in Nigeria, 

the government should directly deliver goods and services, including free medical services, subsidized 

housing, and education. Implementing negative income taxes to supplement the earnings of the poor 

and providing a guaranteed income are additional measures suggested.  

CONFLICT OF INTERESTS  

The authors have not declared any conflict of interests.  

 REFERENCES  

 Abbas M, Dalia S, Fausto C, Nanthakumar L, Masoumeh K (2019). Carbon dioxide (CO2) emissions 

and economic growth: A systematic review of two decades of research from 1995 to 2017. The 

Science of the Total Environment. 649(1):31-49.  

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

20 | P a g e  

Abubakar   SI, Cudjoe D   (2021).  The Environmental Impact of energy  consumption in Nigeria: 

Evidence from CO₂ Emissions. Retrieved from https://www.researchsquare.com/article/rs-

420727/v1  

Acheampong AO (2018). Economic growth, CO2 emissions and energy consumption: What causes what 

and where? Energy Economics 74:677-692.  

Agras J, Chapman D (1999). A dynamic approach to the environmental Kuznets curve hypothesis. 

Ecological Economics 28(2):267-277.  

Alege PO, Ogundipe AA (2013). Environmental quality and economic growth in Nigeria: A fractional 

co-integration analysis. International Journal of Development and Sustainability 2(2):20-29.  

Bölük G, Mert M (2015). The renewable energy, growth and environmental Kuznets curve in Turkey: 

An ARDL approach. Renewable and Sustainable Energy Reviews 52:587-595.  Chuku AC (2011). 

Economic development and environmental quality in Nigeria: Is there an environmental 

Kuznets curve? MPRA Paper No. 30195, posted 18 Apr 2011 11:38 UTC pp. 83-98. 

https://mpra.ub.unimuenchen.de/30195/Cooperation   

Dickey D, Fuller W (1981). Likelihood ratio statistics for autoregressive time series with a unit root. 

Econometrica 49:1057-1072.  

Dinda S (2004). Environmental Kuznets curve hypothesis: A survey. Ecological Economics 49(4):431-

455.  

Egbetokun S, Osabuohien E, Akinbobola T, Onanuga OT, Gershon O, Okafor V (2020). Environmental 

pollution, economic growth and institutional quality: Exploring the nexus in Nigeria. 

Management of Environmental Quality 31(1)18-31.  

Engle R, Granger W (1987). Co-Integration and error correction:representation, estimation, and 

testing. Econometrica 55:251-276.   

Friedl B, Getzner M (2003). Determinants of CO₂ emissions in a small open economy. Ecological 

Economics 45(1):133-148.  

Grossman GM, Krueger AB (1995). Economic growth and the environment. Quarterly Journal of 

Economics 60(2):353-375. http://www.jstor.org/stable/2118443   

Hamit-Haggar M (2012). Greenhouse gas emissions, energy consumption and economic growth: A 

panel cointegration analysis from Canadian industrial sector perspective. Energy Economics 

34(1):358-364.  

Han J, Tianyi D, Chao Z, Xuepeng Q (2018). Correlation analysis of CO₂ emissions, material stocks and 

economic growth nexus: evidence from Chinese provinces. Journal of Cleaner Production 

180:395-406.  

mailto:contact@americaserial.com
https://www.researchsquare.com/article/rs-420727/v1
https://www.researchsquare.com/article/rs-420727/v1
https://www.researchsquare.com/article/rs-420727/v1
https://www.researchsquare.com/article/rs-420727/v1
http://www.jstor.org/stable/2118443
http://www.jstor.org/stable/2118443


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

21 | P a g e  

Han S, Lin C, Zhang B, Farnoosh A (2019). Projections and recommendations for energy structure and 

industrial structure development in China through 2030: A system dynamics model. 

sustainability 11(18):1-20.   

Harmon C (2023). Gini coefficient shows progress in Nigeria’s wealth distribution under democracy. 

Accessed from  

https://nairametrics.com/2023/03/21/gini-coefficient-shows-progressin-nigerias-wealth-

distribution-under-democracy.  

He J, Richard P (2010). Environmental Kuznets curve for CO₂ in Canada, Ecological Economics 

69(5):1083-1093.  

Hossain S (2012). An econometric analysis for CO₂ emissions, energy consumption, economic growth, 

foreign trade and urbanization of Japan. Low Carbon Economy 3(3A):92-105.  

Johansen S, Juselius K (1990). Maximum likelihood estimation and inference on cointegration — with 

applications to the demand for money. Oxford Bulletin of Economics and Statistics 52(2):169-

210.  

Johansen S (1988). Statistical analysis of cointegrating vectors. Journal of Economic Dynamics and 

Control 12(2/3):231-54. Kuznets S (1955). Economic growth and income inequality. American 

Economic Review 45(1):1-28.  

Okafor ECN, Joe-Uzoegbu CKA (2010). Challenges to development of renewable energy for electric 

power sector in Nigeria. International Journal of Academic Research 2(2):211-216.  

Okon EO (2021). Nigeria: Is there an environmental Kuznets Curve for fluorinated gases?, Open 

Economics, ISSN 2451-3458, De Gruyter, Warsaw 4(1):57-71.  

Omisakin OA (2009). Economic growth and environmental quality in Nigeria: Does environmental 

Kuznets curve hypothesis hold? Environmental Research Journal 3(1):14-18.  

Omri A (2013). CO₂ emissions, energy consumption and economic growth nexus in MENA countries: 

Evidence from simultaneous equations models. Energy Economics 40:657-664.  

Özokcu   S,   Özdemir   Ö (2017).    Economic     growth,     energy,    and environmental Kuznets curve. 

Renewable and Sustainable Energy Reviews 72:639-647.  

Phillips PC, Perron P (1988). Testing for a unit root in time series regression. biometrika, 75(2):335-

346.  

Rafindadi AA (2016). Revisiting the concept of environmental Kuznets curve in period of energy 

disaster and deteriorating income: Empirical evidence from Japan. Energy Policy 94:274-284.  

mailto:contact@americaserial.com


 

American Research Journal of Economics, Finance and Management 
americaserial.com/Journals/index.php/ARJEFM, Email: contact@americaserial.com 

22 | P a g e  

Robinson P, Shaheen EI, Shaheen EI (2007). Environmental pollution control. In ‘Practical Advances 

in Petroleum Processing’ pp. 395-447.  

Salahuddin M, Gow J (2014). Economic growth, energy consumption and CO₂ emissions in Gulf 

Cooperation Council countries. Energy 73:4458.  

Sugiawan Y, Managi S (2019). New evidence of energy-growth nexus from inclusive wealth. Renewable 

and Sustainable Energy Reviews 103:40-48.  

Tong T, Ortiz J, Xu C, Li F (2020). Economic growth, energy consumption, and carbon dioxide 

emissions in the E7 countries: a bootstrap ARDL bound test. Energy, Sustainability and Society 

10(1):1-17.  

Vincent M, Ezaal O (2022). Fossil fuel consumption, economic growth, and environmental degradation: 

is the 'energy consumption- growth' nexus sustainable in Nigeria? IOSR Journal of Economics 

and Finance 13(2):51-68.  

Wang S, Zhou C, Li G, Feng K (2016). CO₂, economic growth, and energy consumption in China's 

provinces: Investigating the spatiotemporal and econometric characteristics of China's CO₂ 

emissions. Ecological Indicators 69:184-195.  

  Wickremasinghe G (2011). The Sri Lankan stock market and the macro  economy: an empirical 

investigation. Studies in Economics and Finance 28(3):179-195.  

Yule U (1926). Why do we sometimes get non-sense correlations between time-series? A Study in 

Sampling and the Nature of TimeSeries. Journal of the royal statistical society 89(1):1-63. 

  

  

  

   

  

  

  

  

  

  

mailto:contact@americaserial.com

