1
Chapter 1: Introduction to the Study
The research on the relationship between electricity and economic development is
vast, but there is no consensus on the direction or type of relationship. Various studies in
Nigeria have yielded different results. Researchers have indicated that electricity
consumption leads to economic development in Nigeria (Alley et al., 2016; Nwankwo &
Njogo, 2013; Ubi & Effiom, 2013). Further, electricity and economic growth have a
mutually reinforcing effect in Nigeria; economic development drives electricity
consumption, and increased electricity consumption drives economic development
further (Ogundipe & Apata, 2013; Ologundudu, 2015). Other researchers have
discovered a unidirectional relationship that runs from economic development to
electricity consumption in Nigeria (Ogundipe et al., 2016; Ugwoke et al., 2016).
However, research has also shown no effects between electricity consumption and
economic development in Nigeria (Ologundudu, 2015). These studies have policy
implications, but the diverse research outcomes confuse rather than support policy
development for electric power development and economic growth.
Most researchers in the literature I reviewed on the electricity economic
development nexus treated electricity as a unitary system without consideration for the
subsystems that constitute the power delivery system and how they could affect Nigeria’s
industrial output. A systems approach offers new perspectives on the relationship and
interdependencies between available power generation capacity, available transmission
capacity, distribution capability, and how they affect Nigeria’s industrial output. It was
difficult to locate published studies where researchers applied the systems approach in
2
studying the interactions between the elements that make up the power delivery system
and their effect on industrial output in Nigeria.
In this study, I evaluated the relationships, interdependencies, and the coordinated
joint effort required between power generation capacity, transmission capacity, and
distribution capability in delivering power and how these elements affect industrial
output in Nigeria. Secondary data from the Nigeria Electricity System Operator (NESO)
website that houses the operational data of electric utilities in Nigeria was used for the
analysis. The outcome of this study could advance positive social change by providing
new knowledge about the unique roles of various elements in the power delivery value
chain in the relationship between electricity and economic development. This new
knowledge may inform government policy to support economic development. The rest of
this chapter consists of the background to the study, problem statement, purpose of the
study, research questions and hypotheses, theoretical foundations for the study, nature of
the research, assumptions, and delimitations, including limitations of the study.
Background of the Study
Since Nigeria gained independence from Britain, lack of access to electricity, both
in quantity and quality, has been a recurring problem. In the 2017 state of electricity
access report in the World Bank website (https://www.worldbank.org), Nigeria was
ranked as the second country, behind India, with the largest electricity access gap, with
an estimated 75 million people without access to electricity. In the World Bank 2021
energy progress report, Nigeria had topped the list of countries with the largest electricity
access gap in 2019, with 90 million people without access to electricity. Research on the
3
effect of electricity on economic development in sub-Saharan African (SSA) countries
indicates that poor state of infrastructure and services negatively impact economic growth
in the region (Azolibe & Okonkwo, 2020; Chakamera & Alagidede, 2018; Kodongo &
Ojah, 2016; Owusu-Manu et al., 2019). Lack of electricity access impedes economic
development, health care, and education (Zhang et al., 2019).
More than half of the world’s population now lives in urban centers, with a
projection that this figure could reach 75% by 2050, and SSA is mainly regarded as the
region with the highest urbanization rate (Saghir & Santoro, 2018). It is difficult to
transition from a low-income country to a medium or high-income country without
urbanization (Saghir & Santoro, 2018). But economic growth in SSA countries may be
challenged by poor electricity infrastructure and services. Urbanization usually puts
pressure on existing electricity supply infrastructure leading to power supply shortages
(Zhang et al., 2019). This situation is further compounded by diverse research outcomes
on the region’s electricity and economic development that do not support coherent policy
development.
It was difficult to locate published studies where researchers used the systems
approach to investigate the relationship between electricity and economic development.
In this regard, it is rare to find researchers that have viewed electricity as a natural system
to understand the interplay between the subsystems that constitute the electricity delivery
systems and explore how these subsystems could impact economic development
independently and collectively. There are about eight studies in Nigeria on the effect of
electricity on economic development with growth hypothesis (n = 3), feedback
4
hypothesis (n = 2), neutrality hypothesis (n = 1), and conservation hypothesis (n = 2).
With this type of mixed or diverse research outcomes, it would be difficult for the
government of Nigeria to develop a coherent policy to drive Nigeria’s power sector and
economic development.
To address gaps in research, in this study, I used a systems approach to investigate
the interface relationships between power generation, transmission, and distribution, as
elements of the power delivery system and how these variables affect industrial output in
Nigeria. Von Bertalanffy (1968) noted the continuous relationship between elements of a
system and the need to understand the interdependencies among these elements. Laszlo
(1996) also identified the hypothetico-deductive method as an approach to promote
understanding natural systems as being made up of systems in layers, superimposed on
one another, to constitute a whole. Laszlo’s hypothetico-deductive method allows for
examining the interface relationships between systems and the joint effort required to
maintain a dynamic equilibrium. The systems approach in the electricity and economic
development literature can allow a deeper understanding of the power (electricity) supply
dynamics and how electricity could affect economic development in different ways. This
understanding will support a more informed and targeted policy development approach.
Problem Statement
The literature on the effect of energy resources on economic development
indicates a correlation between electricity consumption and gross domestic product
(GDP). Researchers have used the auto-regressive distributed lag (ARDL) cointegration
test to establish the existence of a long-run correlation between electricity consumption
5
per capita and real GDP per capita as well as a relationship between industrialization and
economic development in Nigeria (Acaravci, 2010; Odeleye & Olunkwa, 2019).
Electricity, through industrialization, has positive effects on economic development
(Allley et al., 2016). But from 2015 onwards, Nigeria witnessed a steady decline in the
production growth rate of most industrial sub-sectors (Odeleye & Olunkwa, 2019). In
2018, the International Monetary Fund and the World Bank ranked Nigeria 140 and 136
in the world, respectively, in terms of GDP per capita. Despite the identified relationship
between electricity and industrialization and their impacts on economic development in
Nigeria, it was difficult to locate studies that established how the elements or variables
that constitute the electricity supply system affect industrialization or economic
development. Researchers tend to view electricity as a unitary system, not considering the
constitutive and interacting elements like power generation capacity, transmission
capacity, and distribution capability that make up the electric power delivery chain and
how they affect industrial output in Nigeria.
The general management problem is that there is a gap in the literature regarding
the interrelatedness between the elements that constitute the power delivery value chain
and industrial output in Nigeria. The specific management problem is the lack of
understanding about the relationship between available monthly power generation
capacity, available monthly transmission capacity, monthly distribution capability, and
industrial output in Nigeria. Most research on the effects of electricity on industrial
output points to correlations and causal linkages between electricity and industrialization,
industrialization and economic development, and electricity and economic development,
6
without explaining the relationship between various elements that constitute the
electricity supply value chain and industrial output in Nigeria. Explaining the relationship
between elements that make up the power supply value chain will serve to disaggregate
all variables involved in the generation, transmission, and distribution of electricity and
provide an understanding regarding how these variables interact to affect electricity
supply and hence industrial output in Nigeria.
Purpose of the Study
The purpose of this quantitative correlational study was to examine the
relationship between industrial output in Nigeria and available monthly power generation
capacity, available monthly transmission capacity, and monthly distribution capability. In
this study, I used secondary data on available monthly electric power generation capacity,
available monthly power transmission capacity, monthly distribution capability, and
industrial output in Nigeria to examine the relationships among these variables. Available
monthly power generation capacity, available transmission wheeling capacity, and
monthly power distribution capability constitute the independent variables, and industrial
output in Nigeria was the dependent variable. The focus of multiple regression analysis
was 6 years of monthly electricity sector operational data in Nigeria from 2015 to 2020.
Examining the relationships between the power sector variables and industrial output can
support the development of a conceptual model to explain how these variables
individually and collectively contribute to the industrial output trajectory in Nigeria.
7
Research Question and Hypotheses
The primary aim of this study was to examine how the elements that make up the
power delivery system affect industrial output in Nigeria. The elements that constitute the
power delivery subsystems are (a) available monthly power generation capacity, (b)
available monthly power transmission capacity, and (c) monthly power distribution
capability. The research question and hypotheses that guided this study are as follows:
Research question: Is monthly industrial output in Nigeria related to available
monthly power generation capacity, available monthly power transmission capacity, and
monthly power distribution capability?
H0: Monthly industrial output in Nigeria is not related to available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability.
Ha: Monthly industrial output in Nigeria is related to available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability
The independent variables are the available monthly power generation capacity,
available monthly transmission capacity, and monthly power distribution capability, and
the industrial output in Nigeria is the dependent variable. The level of measurement of all
the variables is at the ratio scale.
Theoretical Foundation
Von Bertalanffy’s (1968) systems theoretical perspective was the theoretical
framework for this study. From the seminal works of von Bertalanffy (1968), Laszlo
8
(1996), Passmore (1988), and Oshry (2007), it is evident that to gain knowledge about the
emergent nature of a system, there is a need to understand the interrelated and dynamic
relationships that exist between elements of that system, including the effect of the
environment on the system. The four propositions Laszlo used to investigate
organizational invariances was used in this study to establish the Nigerian power supply
value chain as a natural open system that is a whole with irreducible properties, how it
maintains itself within a changing environment, responds to the self-creativity of other
organizations or systems, and serves as a coordinating interface in the Nigerian super
system or holarchy.
The electric power delivery value chain in Nigeria involves power generation,
transmission, and distribution segments. Using a systems theoretical approach in this
study enabled understanding how the different components that make up the power
delivery value chain affect industrial output in Nigeria. The various components are the
individual technological elements and heterogeneous technologies that comprise the
power sector as a system and their impact on industrial output in Nigeria. Using this
theoretical framework aided in understanding the interactions between these elements,
their interdependencies, and the coordinated joint effect required at the interface points to
achieve dynamic equilibrium, thus illuminating the complexities of multiple interactions
resulting from the embedded connectedness of elements that make up the power supply
system. The generation and distribution segments of the power delivery value chain in the
Nigerian power sector are privatized, meaning they will face many risks, including
technological, institutional, regulatory, political, and environmental risks. Adopting a
9
systems view approach can provide details that support the development of practical and
integrative models that may promote a more in-depth understanding of the relationship
between power sector variables and industrial output in Nigeria.
Nature of the Study
The nature of the study is quantitative with a correlational design. Most studies on
the effect of electricity on industrialization and economic development are mostly either
correlational or causal, and the researchers treat electricity as a unitary system. In this
study, I used a correlational design to examine relationships between the variables that
make up the electric power supply value chain in Nigeria. Researchers use correlational
design to measure the strength of association between variables (Gujarati & Porter,
2009).
Secondary data from government institutions that play critical roles in Nigeria’s
power sector was the source of data for the study. These institutions collect data,
including available monthly power generation capacity, available monthly transmission
capacity, and monthly distribution capability. Another source was secondary data
published by the Nigerian Bureau of Statistics (NBS), which archives critical national
economic statistical data. The Nigerian GDP was the source of data for industrial output,
the independent variable in this study. The data points for this study were monthly
operational data for 6 years from 2015 to 2020 for the three segments of the power
delivery value chain and industrial output in Nigeria. I used Pearson’s correlation to
examine the correlation between power generation capacity, transmission capacity, and
distribution capability to analyze the efficacy of the interface (ability) relationships
10
between the variables. Multiple regression analysis was used to test the interaction effects
between monthly industrial output, the dependent variable, and available monthly
generation capacity, available monthly transmission capacity, and monthly distribution
capability, the independent variables, and to estimate the extent to which each
independent variable explains variations in the dependent variable.
Definitions
Available power generation capacity: The reported daily available Nigeria
national power generation capacity from NESO website (https://nsong.org/). This figure
is measured in megawatts (MW) and is usually different from the installed capacity of the
power plants.
Available power transmission capacity: The reported daily available Nigeria
power transmission capacity reported as transmission wheeling capacity from the website
of NESO (https://nsong.org/). This capacity could vary from day to day and is measured
in MW.
Power distribution capability: Total energy the distribution companies in Nigeria
distribute, reported as daily energy utilized from the NESO website (https://nsong.org/).
The energy the distribution companies distribute may vary from day to day.
Industrial output: GDP component of the industrial sectors as reported in the NBS
website (https://www.nigerianstat.gov.ng/).
Growth hypothesis: A postulation that electricity consumption leads to economic
development without feedback (Amin & Murshed, 2017). Under this hypothesis, there is
a causal relationship between electricity consumption and economic development that
11
runs from electricity consumption to economic growth. This hypothesis supports
electricity expansion policies (Mawejje & Mawejje, 2016)
Conservation hypothesis: Under this hypothesis, improvements in economic
development drive electricity consumption (Guan et al., 2015). There is a unidirectional
relationship between economic development and electricity consumption, which is
usually prevalent in energy-sufficient nations (Guan et al., 2015). Under this scenario,
energy conservation policies would not hurt economic development (Mawejje &
Mawejje, 2016)
Feedback hypothesis: A postulation that indicates a mutual and interdependent
effect between energy consumption and economic development (Hasan et al., 2018). In
this case, causation is bidirectional and reinforcing.
Neutrality hypothesis: Under this hypothesis, there is an absence of a causal
relationship between electricity consumption and economic development (Mawejje &
Mawejje, 2016). In this case, economic growth is independent of electricity consumption.
System thinking: System thinking is a broad knowledge approach to problem-
solving that professionals in diverse domains use in analysis, synthesis, and inquiry
regarding a problem of interest (Boardman et al., 2009). Thus, systems thinking
represents a holistic approach to problem-solving that involves examining the bigger
picture.
Assumptions
A major assumption in this study is that the population size would be adequate to
avoid validity issues. I used G*Power software to determine 77 as the minimum sample
12
size, but the available population size was 72 due to limiting data from 2015 to 2020. I
assumed that the difference was not enough to affect the validity of the study. Second,
because of the use of a systems approach in the study, the assumption was that examining
the relationships between the independent variables would reveal interface issues within
the power delivery system and expose possible infrastructural deficits that may further
explain the relationship between the independent and dependent variables. Third, there
was an assumption that the data for the study would satisfy all assumptions for multiple
regression analysis to avoid issues of invalid inferences. Last, I used secondary data for
this study, sourced from government agencies that hold the data; the assumption was that
the data would be of high quality, minimizing issues of applicability, fit for purpose,
availability, relevance, accuracy, and sufficiency.
Scope and Delimitations
In this study, I examined available generation capacity, available transmission
capacity, and distribution capability to study the effect of power system variables on
industrial output in Nigeria. Researchers in electricity and economic development
literature have focused mainly on electricity consumption as the electricity variable. In
this study, the focus was on the elements of the power delivery value chain and the
potential to affect electricity consumption, especially in energy-challenged countries.
This explanation goes beyond electricity consumption to include factors that affect
electricity consumption and provides a path to convergence in the electricity and
economic development literature.
13
A significant delimitation was the cut-off date for data collection. The
government of Nigeria effectively privatized the Nigerian electricity market in 2015
through Order 136 of the Nigeria Electricity Regulatory Commission for commencement
of the transitional electricity market, which represented a significant structural change in
the electricity sector. Collecting data earlier than 2015 for this study could have
introduced structural breaks in data analysis that may affect research validity (external
validity).
Another delimitation was the choice of generation capacity, transmission
capacity, and distribution capability as electricity delivery variables, including industrial
output as the economic variable. The choice of the three power delivery variables was
based on studies on the effect of poor infrastructure on economic development in SSA
countries (Azolibe & Okonkwo, 2020; Chakamera & Alagidede, 2018; Kodongo & Ojah,
2016; Owusu-Manu et al., 2019). Studies on the effect of electricity on industrialization
informed the choice of industrial output as the economic variable (Ologundudu, 2015;
Zhang & Broadstock, 2016). Seventy-one percent of industries in Nigeria provide their
electricity, further supporting the choice of industrial output as a dependent variable in
the study (Osakwe, 2017).
From the literature on electricity and economic development, most scholars
applied econometric theories and statistical methods to assess and analyze economic
theories. The use of these theoretical models does not enable the investigation of
electricity as a natural system. Von Bertalanffy’s (1968) general systems theory and
Laszlo’s (1996) hypothetico-deductive systems method promote the understanding of
14
electricity as a system, made up of layers, in a coordinated joint effort to deliver
electricity for consumption. It was difficult to locate literature on electricity and
economic development where researchers used the systems approach to the phenomenon.
Because this is a nonexperimental design study using secondary data, the threats
to internal and external validity are minimal (see Drew et al., 2008). An external validity
issue that may impact the generalizability of this study is the population size, which is
somewhat less than the minimum sample size. The findings of this study could support
the development of coherent energy and economic development policies through a better
understanding of the dynamics between electricity and economic development in energy-
challenged countries like Nigeria.
Limitations
The population size is a limitation in this study. The a priori power analysis at
80% power (1-β error probability) determined 77 as the minimum sample size; however,
the available population size was 72. This limitation arose because 2015 was the data cut-
off date, and monthly operational data of electric utilities and GDP data constitute the
study population. During data analysis, the implementation of first differencing further
reduced the sample size to 71. Small sample size issues introduce threats to statistical
conclusion validity (SCV; Busk, 2010).
The second limitation is the omission of gas as a variable in the study. In 2015,
thermal generation constituted up to 82% of the power generated in Nigeria (Osakwe,
2017). Natural gas, a fuel source for thermal generation in Nigeria, a covariate in this
study, was not included as a variable in the study. Extraneous variables that could affect a
15
study, other than the identified independent and dependent variables, could threaten
internal validity (Drew et al., 2008). Excluding this variable could be a threat to internal
validity as it could be interpreted as an omitted variable. The use of available generation
capacity as an independent variable addresses this issue. The reason for this assertion is
because gas availability could be one of the reasons for the level of available generation.
Another limitation of the study is the use of secondary data, which presents the
issues of applicability and fit for purpose. In this regard, the issues of availability,
relevance, accuracy, and sufficiency become veritable concerns. Though the Nigerian
system’s operator was the sole aggregator of operational data for the electricity industry
in Nigeria, there was no means of cross-validating the data posted by this agency. In this
regard, the issue of data quality remains a veritable concern. Given the geographically
dispersed nature of the institutions that house data for the study, there may be issues of
cost and time to travel to collect data.
Significance of the Study
Many scholars have studied the role electricity plays in industrialization and how
industrialization leads to the economic development of countries. In these studies, the
researchers identified electricity as a unitary system. In this study, I sought to present
electricity as a natural open system that consists of available monthly power generation
capacity, available monthly transmission capacity, and monthly power distribution
capability as independent variables and industrial output as the dependent variable.
Reducing this gap in the literature could afford a deeper understanding of the relationship
16
between the power sector variables and how they individually and collectively impact
industrial output in Nigeria.
Significance to Theory
In most of the articles reviewed in the electricity economic and development
literature, researchers viewed electricity as a unitary system rather than acknowledging
the interactive and interdependent nature of the constituent elements of the power
delivery system. It was difficult locating studies in this field where researchers viewed
electricity as a natural open system. By using von Bertalanffy’s (1968) general systems
approach and Laszlo’s (1996) hypothetico-deductive method, it was possible to identify
the elements that constitute the power delivery system and elucidate the interrelationships
and interdependencies among them as well as the joint effort required to maintain a
dynamic equilibrium. A systems approach enables an understanding of what happens at
the interface points of these subsystems to offer a robust explanation of the effect of
electricity on economic development and may explain the divergence in the literature.
The various possible interface disconnect issues, especially in power deficit countries,
include (a) available transmission capacity being lower than available generation
capacity, (b) distribution capability being lower than available transmission capacity, and
(c) available generation capacity being higher than distribution capability. Using
generation, transmission, or power consumption (distribution capability) as an
independent variable in the electricity and economic development research becomes
problematic under any of these scenarios. A systems approach brings clarity to these
17
issues and contributes to theory, especially in studying the effect of electricity on
economic development in SSA and other power deficit countries.
Significance to Practice
Poor-quality power supply has negatively affected Nigeria’s industrial output
(Chete et al., 2014). Further, poor electricity access, low quality power supply, and high
electricity cost were the three main reasons for low capacity utilization, lack of
competitiveness, and lack of growth in Nigeria’s industrial sector (Osakwe, 2017). In this
study, I used a systems perspective to identify the power system variables and analyze the
effect of electricity on industrial output in Nigeria, which can enable insights into the
relationship between power sector variables and how they affect Nigeria’s industrial
output. Insights of this nature may support power sector practitioners in Nigeria to
identify bottlenecks in the power delivery value chain. Identifying these bottlenecks
could aid the development of practical and integrative models that support improved
power supply to the industrial sector and lead to growth in industrial output in Nigeria.
The diverse nature of research outcomes in the electricity and economic growth literature
has made policy development difficult, especially for countries experiencing power
supply shortages. The system thinking approach promotes a better understanding of the
dynamics of the electricity supply subsystems, leading to convergence and a targeted
approach to policy development.
Significance to Social Change
Because of the systems approach to this study, I used secondary data on available
monthly generation capacity, available monthly transmission capacity, monthly
18
distribution capability, and monthly industrial output in Nigeria to develop a predictive
model that aids in the development of an integrated power sector policy to drive effective
and efficient power delivery to the industrial sector. An integrated power sector policy
may aid the development of infrastructural and institutional imperatives for improved
production output in Nigeria. Because industrialization leads to economic growth and
electricity has a positive impact on industrialization, the results of this study could lead to
the development of effective policies that drive improvement in power supply to the
industrial sector in countries where power supply has been identified to be deficient. Such
improvements could catalyze enhanced production output that would lead to economic
growth and eradication of poverty through increased economic activities that lead to
gainful employment and positive social change.
Summary and Transition
The literature on electricity and economic development is diverse, with multiple
studies in the same country yielding different results. Most researchers in electricity and
economic development literature view electricity as a unitary system and use electricity
consumption as a key independent variable in their studies. The purpose of this
correlational study, anchored on a systems approach, was to examine the relationship
between the power delivery subsystems and their effect on industrial output in Nigeria. In
this chapter, I addressed delimitations, assumptions, limitations, and significance of the
study, including social change implications and theoretical foundations of the study.
Chapter 2 includes the theoretical foundation of the study and an examination of the
literature related to key variables. Chapter 3 will include a description of the dependent
19
and independent variables for the study, including a detailed explanation of the research
design and methodology.
20
Chapter 2: Literature Review
There is a lack of research acknowledging the multiple elements of the electric
power delivery chain such as power generation capacity, transmission capacity, and
distribution capability despite documentation of the impact of electricity on
industrialization and economic development (Acaravci, 2010; Alley et al., 2016; Odeleye
& Olunkwa, 2019). The purpose of this quantitative correlational study was thus to
examine the relationship between industrial output in Nigeria and available monthly
power generation capacity, available monthly transmission capacity, and monthly
distribution capability. In this chapter, I review relevant literature on the impact of
electricity on industrialization and economic development. The focus of this literature
review was Nigeria, though relevant studies in other countries were used to deepen
understanding of the topic of the study. In this regard, there is a review of literature on
electricity and industrial output, the relationship between electricity access and GDP,
electricity and industrialization, and the impact of the quality of power on industrial
output in Nigeria and other countries. This chapter also includes an exploration of the
concept of systems theory, the foundational model in this study, and a description of the
generation, transmission, and distribution of electricity as a natural open system.
Literature Search Strategy
I conducted literature searches using Thoreau Multi-Database Search, SAGE
Journals database, and Google Scholar. Walden University Library served as an anchor
for these searches. The key search terms and combinations were electricity access and
development, electricity consumption and development, energy utilization and
21
development, quality of power and industrial output, quality of power and
competitiveness, industrialization and GDP, and industrialization and development.
Because Nigeria is the focus of the study, these searches included the word “Nigeria.”
The searches were primarily limited to the last 5 years, from 2015 to 2020, but included
some historical articles published more than 5 years due to their significance and
relatedness to the study. The resources related to the theoretical framework were much
older as they are seminal in nature and provided a foundational perspective. Of the more
than 250 articles reviewed, I selected 100 for this study. I used Zotero to organize, store,
and manage the references.
Theoretical Foundation
Most of the literature on systems theory contains references to the seminal works
of Ludwig von Bertalanffy as the core proponent of general systems theory. I grounded
this literature review on von Bertalanffy’s (1968) foundations of general systems theory
to provide a basic theoretical understanding of systems and support the conceptualization
of electric power delivery as a system. Additionally, scholars like Laszlo (1996),
Passmore (1988), and Oshry (2007) had approaches which hold elements that supported
the elucidation of my research question and hypotheses. Laszlo presented the
hypothetico-deductive method as a means of analyzing the organizational features of
natural systems. This method involves hypothesizing about likely features of natural
systems and using such a hypothesis as a guide to study phenomena to confirm or
disconfirm them as systems. Oshry focused on real-life applications and likely
implications of the concept of system theory. Oshry promoted the concept of seeing
22
systems with the understanding that patterns of issues, either at the personal or
organizational levels, experienced in human organizational settings, are systemic. Such
patterns are mirrored in other organizations, making it a phenomenon that cannot be
solved at the micro level (Oshry, 2007). Passmore (1988) described socio-technical
research as the study of the interaction that exists between social and technical systems.
Passmore noted a need first to understand how the interaction between humans and
machines impacts organizational and environmental outcomes to understand the dynamic
interaction and issues of emergence in organizations, including the impact of the
environment on organizations.
The two systems viewpoints from Oshry (2007) and Passmore (1988) are relevant
to this study because they support understanding of the interaction between the dependent
and independent variables in this study. Von Bertalanffy (1968) and Laszlo (1996)
presented systems from theoretical perspectives, but Passmore and Oshry presented real-
life manifestations of systems in human organizations. Combining these works enabled
insights into the emergent and complex nature of systems by providing an understanding
of the dynamic nature of the interrelationships that exist between elements of a system,
including the impact on the environment on systems. In this regard, the general systems
theoretical foundations and systems thinking models provided direction for the analysis
in this study.
Origins of General Systems Theory
Von Bertalanffy (1968) was the first to use the phrase general systems theory to
describe system thinking approaches to phenomena in different fields in his book General
23
Systems Theory. The difficulties encountered using the logic of cause and effect to
explain theoretical problems led to the development of the systems approach to explain
phenomena, especially with a focus on factors related to the biosocial sciences and
modern technology (von Bertalanffy, 1968). Shortcomings in the methods of the physical
sciences gave rise to the systems approach to explaining phenomena.
Von Bertalanffy (1968) noted that the systems approach is not limited to studies
within specific fields; rather, it involves consideration of several factors that generate
systems, including the relationship between humans and machines. Von Bertalanffy
noted that technology, for instance, does not refer to single machines working in
isolation. Instead, technology should be viewed with consideration of various
technologies, including the relationship between humans and machines and other
intervening economic, political, and social factors that combine to give rise to the
technology (von Bertalanffy, 1968).
According to von Bertalanffy (1968), there is a continuous relationship between
parts or elements of a system, and these relationships are nonlinear. He further noted that
this understanding facilitates the explanation of the issues of organization and order in
terms of the dynamic interaction between parts of a system. In this regard, order and
organization pertain to the differences in the behavior elements of a system exhibit when
acting either as parts or wholes. Von Bertalanffy noted that the sum of the parts of a
system is not equal to the whole. Further, the relationship between elements of a system
and the nature of dynamic interaction between these elements in different fields tend to
exhibit general trends; however, the causal factors that drive these relationships and
24
interactions are different. Von Bertalanffy stated that there is a need for a theoretical
framework to explain these trends to facilitate communication across disciplines. Core to
these trends is the emergence and similarities or isomorphism of structures in different
fields (von Bertalanffy, 1968). In von Bertalanffy’s general systems theory, isomorphism
is context-specific in terms of concrete and unique problems and their corresponding
abstractions.
Closed and Open Systems
Von Bertalanffy (1968) described closed systems as those where the elements,
under specific conditions, are isolated from their environment. Under this scenario, it is
possible to measure the reaction rates between elements, including the attainment state of
equilibrium. Living organisms represent open systems, as there is a constant exchange
between the organisms and their environment arising from internal metabolism that leads
to the generation and breakdown of energy. By interacting with the environment, living
organisms attain a steady state, different from the state of equilibrium, as in
thermodynamics laws that apply to closed systems.
Von Bertalanffy (1968) noted that in closed systems, the status of initial
conditions impacts the final state; that is, a change in the initial condition tends to impact
the final state. For open systems, the final state could arise from different initial
conditions and processes. Von Bertalanffy referred to this phenomenon as equifinality.
According to von Bertalanffy, these differences between open and closed systems, and
hence the differences between physical and living systems, present difficulties in
applying the general laws of physics to open systems.
25
Developments in System Theory
Von Bertalanffy’s (1968) objective in promoting the general systems theory was
to attain a unified language for analysis and conveyance of information/knowledge about
a phenomenon across disciplines. But literature on systems research indicates a
fragmentation in the field of systems research. This fragmentation is apparent in scholars’
adaptation of system principles for their field of interest, thus presenting multiple views
in systems research (Adams et al., 2014; Demetis & Lee, 2017; Rousseau, 2017; Stroh,
2015). The issue is a lack of cohesion or consensus in systems research. The trends and
present-day reality of systems theory remain largely a work in progress as systems
scholars are yet to achieve the unity of language that von Bertalanffy sought.
The different schools of thought within the systems field share the same
fundamentals of the general systems theory; what differs is the methodology for
analyzing systems and improving systems (Stroh, 2015). There is a multiplicity of
definitions of systems theory in the systems research literature as a result of differences in
the methods of inquiry in the different fields (Adams et al., 2014; Caws, 2015; Demetis &
Lee, 2017; Rousseau, 2015, 2017; Stroh, 2015; Verhoeff et al., 2018). The failure of a
standard definition for general systems theory may have arisen due to the inability of von
Bertalanffy to provide a construct for theory or the lack of required axioms or
prepositions (Adams et al., 2014). Others have pointed to the need for removing
“general” from the original general systems theory given the yet-to-be explained
trajectory of evolutionary processes of natural systems, including the unaccounted
accidents of faith that happened in the process making it impossible to assume any form
26
of generality (Caws, 2015). Though these generalization issues exist in systems research,
the literature reviewed indicates the emergence of pluralism in systems approaches and
methodologies.
The inherent diversity in the definition of systems theory is also present in
systems thinking (Bonnema & Broenink, 2016; Cabrera et al., 2015; Chan, 2015; Clancy,
2018; Grohs et al., 2018; Monat & Gannon, 2015; Mononen, 2017; Whitehead et al.,
2015). The disciplinary area impacts the view or definition of systems thinking given the
propensity for system scholars to anchor system thinking on their guiding systems theory
and corresponding method(s) of inquiry (Verhoeff et al., 2018). Researchers in the
natural sciences view systems thinking from an objective/positivist approach that
includes systematic processes that may consist of models (Verhoeff et al., 2018). System
thinking is multidisciplinary, and theory is critical to applying system thinking (Monat &
Gannon, 2015; Sibo-Ingrid et al., 2018; Verhoeff et al., 2018; Weissenberger-Eibl et al.,
2019). The diverse definitions have made it difficult to achieve a common language for
analyzing, understanding, and communicating research in system thinking across
disciplines.
System Thinking Approaches
System thinking is both a worldview and a process used to solve complex system
problems and develop and understand systems (Moldavska & Welo, 2016). Barry
Richmond (as cited in Bonnema & Broenink, 2016) developed the term system thinking
to press for the need to understand the complex ways systems behave to support decision
making. The systems thinking approach means seeing systems within the context of their
27
environment and that systems thinking enables the perception of complexity
(Weissenberger-Eibl et al., 2019). Systems thinking can also be described as an approach
for examining the relationships among elements or components, how they affect system
outcomes, and their roles within their larger environment (Amissah et al, 2020). Systems
thinking is also a collection of knowledge that assists system professionals in multiple or
various domains to conduct analysis, synthesis, and inquiry regarding a phenomenon or
system of interest (Boardman et al., 2009). When a problem is systemic, it means that it
pervades all aspects of a system and therefore requires a holistic approach, rather than
specific piecemeal efforts or solutions to deal with the issues; this is the bedrock of
systems thinking (Boardman et al., 2009). Systems thinking involves seeing the bigger
picture, both in terms of analyzing a problem and proffering a solution (Arnold & Wade,
2015).
Systems thinking is primarily a philosophy without a methodology; systems
thinking is sometimes referred to as a multidisciplinary approach involving multiple
perspectives (Sibo-Ingrid et al., 2018). The need to synthesize data from different sources
into a single source remains an important underlying aspect of a systems approach (Sibo-
Ingrid et al., 2018). Systems thinking introduces a problem recognition and analysis
frame different from the more linear reductionist approaches that focus more on cause
and effect type of relationships and analysis (Castelle & Jaradat, 2016). Systems thinking
supports the understanding of wholes beyond the components that constitute it to include
considerations for the connectedness of subsystems with the whole systems and the inter-
relationships among the components that make up the systems and other systems
28
(Kordova et al., 2018). The reductionist approach poses the danger of working in silos
that prevent the view of the full organizational context leading to limited potency in
resolving complex problems (Vemuri & Bellinger, 2017). Though different approaches to
systems thinking exist, there is no bad or good system thinking approach depending on
the discipline (Castelle & Jaradat. 2016). This statement reinforces the need for pluralism
in terms of the approaches used in systems research.
A review of systems literature reveals duplicity of efforts and lack of coherence in
defining systems thinking and developing a standard methodology for applying systems
theory to real-life/practical situations. Different authors have made attempts at attaining a
unified definition or common language for system thinking (Boardman et al., 2009;
Bonnema & Broenink, 2016; Cabrera et al., 2015; Clancy, 2018; Grohs et al., 2018;
Harris & Caudle, 2019; Moldavska & Welo, 2016; Monat et al., 2020; Monat & Gannon,
2018; Mutingi et al., 2017; Serrano et al., 2018; Stroh, 2015; White, 2015; Whitehead et
al., 2015; Zare et al., 2017). The most prominent of these approaches is system
archetypes and conceptagon in the analysis of systems. Archetypes in systems thinking
are a means of understanding the complex and dynamic relationships in organizations
(Clancy, 2018; Moldavska & Welo, 2016; Zare et al., 2017). The conceptagon is a system
thinking tool and a framework for applying system thinking in multiple domains of
specialization (Boardman et al., 2009; Moldavska & Welo, 2016). The conceptagon is
used to apply system thinking ideas such as emergence, relationships, interconnectedness,
nonlinearity, feedback, delays, and system of systems in the analysis of systems
(Moldavska & Welo, 2016). No matter the approach to systems thinking or methodology
29
applied in the study of systems, the common denominator is the use of system thinking as
a structured analytical process for understanding complex systems.
The systems research literature is dotted with multiple methods and approaches to
system thinking and is mainly driven by ontological and epistemological differences. But
literature is scarce about how to conduct system thinking (Bonnema & Broenink, 2016).
The combination of different system thinking methodologies introduces a variety of
worldviews that could enable, support, or facilitate the ability to analyze complex
systems (Castelle & Jaradat, 2016). Further, the systems research literature is lacking on
how electric power delivery system elements affect industrialization and economic
development. In most literature reviewed for this study, researchers treat electricity as a
unitary system.
System Thinking Approaches in the Energy Growth Nexus Literature
From the literature that was reviewed for this study, many scholars presented
electricity as a unitary system. None of these studies presented or considered electricity
as an open system. The systems theoretical perspective was used to explore electricity as
an open system, consisting of multiple interdependent variables, and how these variables
impact industrial output in Nigeria. I used von Bertalanffy’s (1968) general systems
theory as a foundational perspective on systems theorizing and the guiding framework for
this study. Laszlo’s (1996) hypothetico-deductive method was used to analyze the
organizational features of natural systems. The electric power delivery value chain in
Nigeria involves the power generation, power transmission, and power distribution
segments. In this regard, Laszlo’s four propositions were used to investigate
30
organizational invariances to establish the Nigerian power supply value chain as a natural
system with irreducible properties, how it maintains itself within a changing
environment, responds to the self-creativity of other organizations (or systems), and
serves as a coordinating interface in the Nigerian super system or holarchy. In this study,
the elements that make up the Nigerian power delivery system are power generation,
transmission, and distribution segments.
There are specialists in all disciplines; however, Laszlo (1996) noted that
specialization leads to depth and not breadth of knowledge, which results in restricted
perspectives. According to Laszlo, real-life scenarios manifest as varied and
simultaneous influences of different factors or elements that interrelate in complex ways,
including the biology of humans that present in multiple forms. Making sense of this
complex, yet common, scenario can only be through a broad knowledge perspective that
individual specialists in different fields may not be able to offer. Using such broad
knowledge perspectives enables the understanding of the multiple factors and influences
that interrelate as a complex whole rather than as single factors and influences.
Though elements that constitute a system may have unique characteristics, it is
the type of relationships among them and the contingent interdependencies that bind them
together that generate their characteristics as a system (Laszlo, 1996). Skyttner (as cited
in Önday, 2018) stated that a system is where the behavior of two or more elements
affects the whole in an interdependent manner and that though subgroups of elements
may have an effect on the whole, these effects are not independent. In this regard,
summing up the characteristics of the constituent units cannot characterize the whole
31
without considering the mediating roles of the interdependencies and relationships among
the constituent units. In this regard, the structure of the whole (system) informs the
unique relationships between the constituent parts of the whole. Laszlo further stated that
studying the organization of the structures of different systems by way of theorizing
enables the understanding of the commonalities, and these commonalities represent the
nonvarying aspects of organizations are known as organizational invariances.
Laszlo (1996) presented the hypothetico-deductive method as a means of using
hypotheses to study phenomena to confirm or disconfirm them as systems. Laszlo
developed four propositions, which represent four instances of organizational invariances
with which to investigate and verify the existence of systems. Laszlo’s four propositions
state that natural systems: (a) are wholes with irreducible properties (p.25), (b) maintain
themselves in a changing environment (p.30), (c) create themselves in response to the
self-creativity of other organizations (p.39), (d) are coordinating interfaces in nature’s
holarchy (p.53).
Natural Systems are Wholes with Irreducible Parts
Wholes differ from heaps due to the relationships and interdependencies between
elements that make up a whole system (Laszlo, 1996). The existence of a proper structure
and interdependency between elements of wholes characterize the existence of wholes.
The Nigerian electric power delivery system consists of generation, transmission, and
distribution segments for supplying power to the people of Nigeria. It is difficult to
reduce the properties of wholes to the sum of the elements that make up the whole. The
reason for this assertion is that wholes exhibit properties based on the peculiar structure
32
of the wholes and tend to retain these properties even with the gradual replacement of the
elements that make up the wholes. Laszlo’s assertion is based on the logic that it is the
unique structure of wholes and the attendant specific interdependencies and relationships
between the elements of the wholes that differentiate wholes. The expectation is that the
Nigerian power sector, guided by its peculiar structure, will exhibit specific
characteristics, and this is the object of this study.
Natural Systems in a Changing Environment
Open systems import energy from their environment to sustain operations and
maintain a steady state (Laszlo, 1996). Natural systems are open systems because of their
ability to maintain steady states. The main characteristic of open systems is the transport
or exchange of energy and resources, including waste across systems boundaries.
Another system characteristic is maintaining equilibrium under a changing external
environment. Core to Laszlo’s supposition is the embeddedness of subsystems within
larger systems that may constitute an external environment. The generation subsector of
the Nigerian power sector, an independent variable in this study, imports fuel from the
external environment to generate power and hence possesses a core characteristic of
systems.
Natural Systems Respond to the Self-Creativity of Other Organizations
A principal characteristic of natural systems is maintaining a steady-state or
dynamic equilibrium by responding to environmental changes (Laszlo, 1996). Systems
attain dynamic equilibrium in response to changes in the environment through self-
creativity that involves creating structures and new functions. Differentiation is part of
33
the reality of systems, given the existence of different structures, functions, and roles. In
this regard, differentiation enables the joint action that systems require to attain a steady
state and respond to emergent threats to survival. The interdependency between the
generation sector, the transmission sector, and the Nigerian distribution sector represents
a joint action needed to deliver power to consumers. The issues of increased demand for
energy or changes in gas availability for power generation represent veritable changes in
the environment and to continue to supply power, the power delivery system needs to
respond to these changes. How the Nigerian electric power delivery system responds to
these changes represents a critical component of this study. This assertion is in terms of
how the independent variables in this study affect the dependent variable.
Natural Systems as Coordinating Interfaces
Laszlo (1996) presented natural systems are layered systems and these layers are
made up of sub organic (physical sciences), organic (life sciences), and supra organic
(social sciences). Simple structures occupy the lowest layers in this layered structure, the
natural systems occupy the intermediate layers, and the more complex systems occupy
the top layers. In this model, the natural systems coordinate interactions between the top
and bottom layers. The constituent parts of each layer are wholes at that level and, at the
same time, a part of the layer above them. This arrangement means that wholes at any
layer or level serve as coordinating interfaces with other parts or wholes within the
system and represent the joint efforts systems need to maintain dynamic equilibrium.
Laszlo described holarchic duality as the way elements of a system interact as both parts
and wholes in a coordinated joint effort to maintain systems. Holarchic systems respond
34
to emergent environmental threats or changes dynamically as they continually interface
with their environment. The cycle of interface and coordination between elements at the
different layers of a system mirrors the structure and the attendant relationships and
interdependencies between parts of a complex organization. The role of the generation
segment in the Nigerian power delivery systems, both as sources of power generation and
interconnection to the transmission system, represents a holarchic duality. The same
argument holds for the interconnection between the power transmission and distribution
segments, with both the distribution and generation segments in direct interaction with
the external environment.
Literature Review
Research is scarce on the electricity and economic development nexus using a
systems approach. Ordinarily, the constructs of interest regarding this study would be
generation capacity and industrial output, transmission capacity and industrial output, and
distribution capability and industrial output. Due to the limited availability of literature, I
reviewed the literature on the impact of electricity on economic development, the impact
of electricity on industrialization, and the impact of electricity access on economic
development as related constructs of interest.
Electricity and Economic Development
Electricity consumption in developing countries tends to grow as the economy
grows, hence the need to continuously study how electricity consumption rate affects
economic development (Kunda & Chisimba, 2017). Four main hypotheses drive energy
growth research: (a) the growth hypothesis, (b) the feedback hypothesis, (c) the neutrality
35
hypothesis, and (d) the conservation hypothesis (Abokyi et al., 2018; Amin & Murshed,
2017; Guan et al., 2015; Hasan et al., 2018; Istaiteyeh, 2016; Mawejje & Mawejje, 2016;
Ogundipe et al., 2016; Rashid & Yousaf, 2016; Samu et al., 2019; Sankaran et al., 2019;
Sekantis & Motlokoa, 2015). The growth hypothesis points to a dependency of
economies on energy consumption and how energy consumption positively impacts such
economies’ GDP (Sankaran et al., 2019). Under the growth hypothesis, there is a
unidirectional causal relationship that runs from electricity consumption to economic
growth such that an increase in electricity consumption would lead to a rise in economic
development (Ogundipe et al., 2016). The feedback hypothesis is used to focus on
situations where there is a bidirectional causal relationship between electricity
consumption and economic development. Under the feedback hypothesis, energy use and
economic growth are mutually reinforcing, such that a dip in energy consumption causes
a drop in GDP and vice versa (Mawejje & Mawejje, 2016). The neutrality hypothesis
indicates no relationship between electricity consumption and economic development and
that policies aimed at energy conservation will not impact economic development
(Ogundipe et al., 2016). The conservation hypothesis points to a unidirectional
relationship that runs from economic growth to electricity consumption with the
implication that economic growth drives electricity consumption (Mawejje & Mawejje,
2016). The divergence in research outcomes that drive these hypotheses implies that the
literature on the role of electricity in economic growth remains varied.
From the literature reviewed (n = 100), the research results in the energy growth
nexus fall under the four energy growth hypotheses with different policy implications for
36
energy consumption and economic growth: growth hypotheses (n = 13), feedback
hypothesis (n = 10), neutrality hypothesis (n = 5), and conservation hypothesis (n = 13).
From Table 1, the studies conducted in Nigeria, Pakistan, Uganda, Bangladesh, and
China showed conflicting results. For instance, out of the eight studies in Nigeria, three
indicated growth hypothesis, two indicated feedback hypothesis, one showed neutrality
hypothesis, and two indicated conservation hypotheses (see Table 2). These conflicts
could be a result of differences in the methods, variables, and time frames for the studies.
Current Trends in the Electricity and Growth Literature
Most research in the electricity growth nexus is quantitative. Many researchers in
the field used different instruments in their studies that tend to suggest methodological
pluralism in energy and economic growth research. Most of the researchers in the
literature I reviewed under the energy growth nexus utilized multiple regression methods
with a diversity of statistical instruments. For studies in Nigeria, various authors
implemented different statistical and econometric tools and different variables to arrive at
different results. Table 1 showcases countries with conflicting results in the energy
growth literature. Table 2 shows studies in Nigeria with different energy growth
hypotheses, variables, and methods. As Table 1 shows, the country effect may not be the
reason for the diversity in results obtained in the literature. A review of Table 2 indicates
that authors using different variables and methods output different results for the same
country. As shown in Table 2, when the variables and methods differ, the results may not
support existing literature; instead, they create new perspectives.
37
Table 1
Countries with Conflicting Results in the Energy Growth Literature
Country
Hypothesis
Author
Bangladesh
Growth hypothesis
Amin & Murshed (2017)
Conservation hypothesis
Hasan et al. (2018)
China
Growth hypothesis
Guan et al. (2015)
Feedback hypothesis
Zhang & Broadstock (2016)
Ha et al. (2018)
Conservation hypothesis
Chol (2020)
Nigeria
Growth hypothesis
Alley et al. (2016)
Nwankwo & Njogo (2013)
Ubi & Effiom, 2013)
Feedback hypothesis
Ogundipe & Apata, (2013)
Ologundudu (2015)
Neutrality hypothesis
Ologundudu (2015) a
Conservation hypothesis
Ogundipe et al. (2016)
Ugwoke et al. (2016)
OECD Countries
Growth hypothesis
Salahudin & Alam (2016)
Feedback hypothesis
Baloch et al. (2019)
Conservation hypothesis
Baloch et al. (2019) b
Pakistan
Growth hypothesis
Ali et al. (2020);
Conservation hypothesis
Rashid & Yousaf (2016)
Mawejje & Mawejje (2016)
Feedback hypothesis
Sekantis & Motlokoa (2015)
Conservation hypothesis
Sekantis & Motlokoa (2015) c
Note. a Neutrality hypothesis with respect to industrial output and not economic
development; b Conservation hypothesis with respect to an indirect relationship through
environmental pollution; c Conservation hypothesis with regard to short-run causality
38
Table 2
Variables and Methods Used in Studies in Nigeria
Hypothesis
Author
Variables
Methods
Growth
hypothesis
Alley et al.
(2016)
Labor, capital formation, industrial
output, and GDP
Three-stage least squares
(3SLS) estimation
technique, Johansen
cointegration tests, and
Growth
hypothesis
Nwankwo &
Njogo (2013)
Electricity, Gross fixed capital
formation, industrial production,
population variables, and RGDP Per
capita.
Ordinary Least Square
(OLS) regression analysis.
Growth
hypothesis
Ubi &
Effiom,
(2013)
GDP, electricity supply, capital stock,
and technology
OLS, the Johansen
cointegration test, and
ECM
Feedback
hypothesis
Ogundipe &
Apata, (2013)
Electricity consumption and economic
development
The Johansen
cointegration test, Juselius
Maximum Likelihood
approach, and Wald block
endogeneity causality test.
Feedback
hypothesis
Ologundudu
(2015)
Electricity supply, industrialization, and
economic development
Granger causality test and
ARDL bounds test.
Neutrality
hypothesis
Ologundudu
(2015) a
electricity supply, industrialization, and
economic development
Granger Causality test and
ARDL bounds test.
Conservation
hypothesis
Ogundipe et
al. (2016)
GDP per capita, electricity
consumption, stock of capital available
in the economy, total labor force,
government effectiveness, structure of
Nigeria economy, state of technology,
and measure of
environmental degradation
The Johansen
cointegration
Test, Juselius maximum
Likelihood approach, and
Wald block endogeneity
causality test.
Conservation
hypothesis
Ugwoke et al.
(2016)
Electricity supply and industrial output
Johansen cointegration
and Augmented Dickey-
Fuller test
Note. a Neutrality hypothesis with respect to industrial output and not Economic
development
39
The question remains: how many views or perspectives will be enough to
understand the relationship between electricity and economic growth? As indicated in
Table 2, Nigeria represents a case in which the results of eight studies represent each of
the four hypotheses in the energy growth nexus research. The same issues are observable
in research in China (see Table 3).
Table 3
Variables and Methods Used in Studies in China
Hypothesis
Author
Variables
Methods
Growth hypothesis
Guan et al.
(2015)
Energy consumption,
urbanization, and economic
development
ARDL; ECM.
Feedback
hypothesis
Zhang &
Broadstock
(2016)
energy consumption,
economic growth,
industrialization, and
urbanization
VAR; Directed Acyclic
Graphs (DAG); Zivot and
Andrews (ZA) test for
structural breaks.
Ha et al. (2018)
Energy consumption and
economic development
Ð-wavelet analysis;
Modified Wald test
(MWALD) for linear
causality test.
Conservation
hypothesis
Chol (2020) a
Per capita national income,
energy consumption, and
environmental pollution
ARDL
Note. a Conservative hypothesis with respect to an indirect relationship through
environmental pollution
Infrastructure and Economic Development
A review of literature reveals that new considerations, which represent gaps in the
literature, introduce new variables that impact results, further increasing the diversity of
research outcomes in the energy growth nexus. It was difficult to locate studies where
researchers considered electricity as a system or introduced generation capacity,
transmission capacity, and distribution capability as new variables in the energy growth
research. Introducing these variables can offer possible clarifications or a path to
40
convergence in the energy and economic growth research outcomes. In the literature
reviewed, only Ali et al. (2020), Azolibe and Okonkwo (2020), Chakamera and
Alagidede (2018), Kodongo and Ojah (2016), and Owusu-Manu et al. (2019) studied the
effect of infrastructure on economic development with a focus on electricity
infrastructure. By viewing the power sector (electricity supply chain) as a system, the
objective was to include considerations for the effects of infrastructure deficits, especially
related to the interface relationships between the independent variables (generation
capacity, transmission capacity, and distribution capability).
Many scholars have identified poor infrastructure as one of the leading factors
responsible for the lack of economic growth in SSA countries (Azolibe & Okonkwo,
2020; Kodongo & Ojah, 2016). In 43 SSA countries, improvements in infrastructure
stock and quality drive economic growth (Chakamera & Alagidede, 2018). The
researchers also determined that poor electricity services slow economic growth
(Chakamera & Alagidede, 2018); these negative effects are only associated with the
quality index of the infrastructure stock, which weights electricity high. Azolibe and
Okonkwo (2020) tested the impact of infrastructure on industrial output in 17 SSA
countries and noted that poor electricity infrastructure and services had a negative effect
on industrial output, but in another study, Kodongo and Ojah (2016) found weak and
indirect correlations between infrastructure development and economic development
indices. By investigating the effects of generation capacity, transmission capacity, and
distribution capability in this study, I sought to provide a new perspective, a systems
41
perspective, on the impact of power delivery systems infrastructure on industrial output
or economic development in Nigeria.
Electricity Access and Economic Development
A review of extant literature on the relationship between electricity access and
economic development revealed some interesting dynamics. Zhang et al. (2019)
implemented Bayesian model averaging (BMA) instead of the more common
econometric models in energy growth literature that establish causality and long-run
relationships between economic factors and electricity access. Zhang et al. discovered
that urbanization did not significantly affect electricity access, in contrast with most
results in the literature, and noted that lack of access to electricity limits economic
development and modern services, including health care and education. The researchers
focused on long-run relationships rather than causality factors and used BMA to identify
significant predictor variables that impact electricity access in China. Though electricity
access supports economic development, urbanization does not improve electricity access
Different Approaches and Their Strengths and Weaknesses
The literature on electricity consumption and economic development nexus is
mixed. Different scholars have identified various reasons for the mixed results.
Urbanization, industrialization, and improved electricity access are all drivers of energy
consumption or demand and form significant considerations for electricity and economic
growth research. A review of the energy growth nexus literature indicated that GDP and
level of economic development tend to position countries into the different energy growth
hypotheses. Most scholars in the energy growth literature did not factor in the effect of
42
inadequate power (electricity) infrastructure in the electricity and economic development
research. The conflicting outcomes notwithstanding, the different sociopolitical,
sociocultural, socioeconomic, political, and geographic factors mean that each study will
be unique and the issue will no longer be that of conflict; instead, it could be an issue of
context (Akinwale & Muzindutsi, 2019; Kwakwa, 2017; Sankaran et al., 2019).
In a study of 10 nations, Sankaran et al. (2019) classified countries into the three
broad categories of (a) developed, (b) developing, and (c) newly industrialized to account
for the structural and socioeconomic differences among nations and to explain the
differences in results in the energy growth nexus research, but the results were varied.
The researchers noted the issues of differences in natural resources, population size,
technology level, nature of labor market, the functioning of governments’ machinery, and
the variety of industrial production as likely exogenous factors. Economic development
and increased urbanization lead to increased electricity consumption in Malaysia
(Ridzuan et al., 2020). Rashid and Yousaf (2016) studied the implication of economic
development as a mediator on the energy growth nexus research and noted that increased
urbanization amidst energy shortages leads to a negative correlation between electricity
use and economic development. The inability to account for structural breaks could
explain some of the diversity in the energy growth literature (Zhang & Broadstock,
2016).). Further, energy consumption patterns might have a bearing on the outcome of
tests, especially high consumption by nonindustrial sectors (Ha et al., 2018). Overall, the
differences in research outcomes remain varied, and the reasons for the variations are
contentious.
43
Each researcher in the electricity growth literature claims to fill a gap in the
literature. Scholars introducing new variables or considerations could be another source
of diversity in the energy growth nexus literature. From the research I reviewed for this
study, every new perspective contributes to the diversity in the results within the
electricity and economic growth literature. The paucity of evidence, inadequacies in
existing methodology, neglect of certain important variables in existing research,
expansion of the scope of a study (variables), the introduction of new methodology or
models, types of relationship between variables (linear vs. nonlinear), decomposition of
scope into time frames, and types of data are some of the diverse perspectives that
scholars bring to the energy growth nexus research.
Instances of the variations in the focus of study in the energy and growth literature
abound. Ali et al. (2020) noted the focus of existing literature has been on the energy
growth nexus instead of Pakistan's electricity consumption growth nexus and bifurcated
electricity consumption into electricity generation and shortage as variables in their study
and noted that these gaps could compromise the findings of previous studies. Amin and
Murshed (2017) claimed that no prior research involved a multivariate method to
investigate the relationship between electricity consumption and economic growth in
Bangladesh to explain the disparities in previous studies. Baloch et al. (2019) found that
most prior researchers investigated linear relationships between variables without
considering the nonlinear relationships. Baloch et al. also pointed to the variations in the
use of series and panel data in energy and growth research. All these approaches
represent different perspectives and may account for the diversity in results.
44
Literature is scarce on the systems approach to the energy (electricity) growth
nexus. Literature abounds on the use of econometric and multiple regression methods in
the study of the electricity growth nexus. Ha et al. (2018) and Zhang and Broadstock
(2016) implemented a time-varying framework to account for structural breaks that could
cause data instability, leading to conflicting results. Ha et al. implemented the D-wavelet
analysis to decompose time series data into time domains to account for possible
structural breaks in the data. Zhang and Broadstock incorporated the vector
autoregression into a time rolling framework to account for time variation in a system
framework. At different times in the economic development of the United States, the
relationship between energy use and economic development showed different trends
(Hirsh & Koomey, 2015). From 2007 onwards, the United States recorded GDP growth
with little or no net growth in electricity use, and since 1996, the United States had
required less electricity to increase GDP (Hirsh & Koomey, 2015). Structural breaks are
important considerations in the energy and economic growth literature and may account
for some of the diversity in research outcomes.
From literature on the energy growth nexus, the relationship between electricity
and economic development may vary in a nonlinear manner. Nazlioglu et al. (2014)
implemented a nonlinear Granger causality test and found a bi-directional relationship
between electricity consumption and economic growth in both the long run and short run
in Turkey. Electricity consumption and economic growth nexus might vary in the short
and long run within the same economy (Ali et al., 2020). In Thailand, there was a long
run unidirectional relationship between GDP and electricity consumption, running from
45
GDP to electricity consumption, and a bidirectional relationship between the two
variables in the short run (Jiranyakul, 2016). In a study of 17 industries in Taiwan, there
was a bidirectional relationship between electricity consumption and GDP in both the
long run and short run (Lu, 2017). These approaches emphasize considerations for
correlational analysis, both in the short and long run, to understand relationships among
variables in the electricity economic growth nexus.
A review of literature revealed that the issue of context is germane to energy
growth nexus research. There is a need for new approaches in electricity and economic
growth research to deal with issues of mixed or conflicting results (Ha et al., 2018).
Owusu-Manu et al. (2019) studied quantum and quality of power sector infrastructure as
key independent variables but did not incorporate the interface relationships between the
various subsectors within the electric power delivery chain. By focusing on different
subsystems that make up the power delivery chain and how they impact industrial output
in this study, I introduced new approaches and considerations in the electricity and
economic growth literature. Most researchers in the energy growth nexus used variables
like electricity price, electricity consumption, national income, population, urbanization,
foreign direct investments, industrialization, GDP, and inflation.
Rationale for Selection of the Variables
Many scholars have explored the complex interactions between economic
variables in the electricity and economic growth literature. The purpose of this study was
to explore the complex interactions between power sector variables and how they impact
industrial output in Nigeria. The aim was to deepen understanding and present robust
46
perspectives in the energy growth literature. Using the four propositions in Laszlo’s
(1996) hypothetico-deductive method can provide clarity on how failures in the interface
relationships (infrastructure) between subsectors within the electricity delivery system
could lead to conflicting results in the energy growth nexus research. Laszlo’s (1996)
hypothetico-deductive model focused on the unique characteristics of elements that make
up a system, the interrelationship between the system’s components, and the contingent
interdependencies that bind them together. The sum of the constituent parts of a system’s
characteristics would not depict or describe the whole system without considerations for
the mediating roles of the type of interdependencies and relationships among that
system’s elements (Laszlo, 1996). This logic holds for the different parts of the power
delivery system that form the independent variables in the current study: power
generation, power transmission, and power distribution systems.
The structure of the whole informs the type or unique relationships between the
constituent parts of systems (wholes) (Laszlo, 1996). Studying the organization of
different systems enables the understanding of the commonalities that exist and issues
that characterize the systems. It is these commonalities that represent organizational
invariances (Laszlo, 1996). Using Laszlo’s systems concept to analyze the variables in
the current study to answer the research question requires a correlational study of the
power generation, transmission, and distribution data. This approach will promote an
understanding of the unique relationship and interdependency between the subsystems
and how these variables impact industrial output in Nigeria, both as a system and as
independent variables.
47
Adopting Laszlo’s hypothetico-deductive method implies that the generation,
transmission, and distribution subsystems of the power delivery system require a joint
coordinated effort at each subsystem’s interface points for efficient power delivery. Still
relying on the logic of the hypothetico-deductive method, a capacity mismatch at the
various interface points of the power delivery systems could result in the inefficient
transmission of power generated or inadequate distribution of power transmitted. In this
regard, the risk could be that the electricity available for consumption may not truly
reflect the capacity of the power delivery systems. Theoretically, these elements could
impact industrial output differently when analyzed independently or as a unitary system.
Review, Synthesis, and Justification for Study and Variables
The independent variables in this study are generation capacity, transmission
capacity, and distribution capability, while the dependent variable is the monthly
industrial output in Nigeria. There was difficulty during literature search identifying
studies where researchers had examined the relationship between the independent and
dependent variables used in this study. The most closely related studies addressed the
effects of electricity infrastructure and the quality of electricity services on
industrialization and economic development. Other researchers investigated the impact of
electricity on economic growth and treated electricity as a unitary system without
considering the subsystems that constitute the electricity delivery system. Infrastructure
plays a significant role in the development of SSA countries and the poor state of
electricity infrastructure and electricity services impeded economic growth in these
countries (Azolibe & Okonkwo, 2020; Chakamera & Alagidede, 2018; Kodongo &
48
Ojah, 2016). In these studies, the researchers assumed that the different subsystems
(generation, transmission, and distribution) are in an equal state of disrepair and that poor
services run across all the subsystems; what is yet to be studied is how the different
power supply systems contribute to the infrastructure deficit and quality of service. In this
study, I used a systems approach to investigate each power delivery subsystem, the
interface relationships between the systems, the interdependencies that exist, and how
they independently impact monthly industrial output in Nigeria. Results may provide
broader perspectives that could explain some of the diversity in studies in Nigeria’s
electricity growth nexus.
Summary and Conclusions
There appears to be a lack of consensus or convergence in the electricity and
economic growth literature on the energy and electricity growth nexus. The use of
electricity offers more advantages over other energy sources as it enables more efficient
utilization in the telecommunication and manufacturing industries, including lighting
(Stern et al., 2019). In this regard, I focused on electricity and economic development in
the literature review for this study. Different scholars have implemented different
methodologies, levels of analysis, types of analysis, and variables to arrive at divergent
results.
By adopting a systems approach to the electricity growth nexus research, this
study filled a critical gap in the electricity and economic growth literature and provided a
more robust perspective regarding the role electricity plays in the economic development
of nations, especially in developing countries like Nigeria. Implementing a systems
49
approach can enable the analysis of the interface relationships and interdependencies
between the subsectors that compose the electricity delivery system. This type of analysis
may shed light on the state of infrastructure at the different interface points in the electric
power delivery system. Using Laszlo’s (1996) hypothetico-deductive method will support
the understanding of gaps in the coordinated joint effort required to sustain Nigeria’s
electricity delivery system.
Methodological differences may be responsible for the diversity in the electricity
growth literature. Issues of the extent of sector analysis regarding the number of variables
or factors under study may also account for the differences in the results. In this regard,
the differences in bivariate, multivariate, linear, and nonlinear analysis of variables
become prominent.
Chapter 3 contains an explanation of the data analysis plan for the study,
including threats to validity (internal and external). Chapter 3 also contains a description
of the dependent and independent variables for the study and a detailed explanation of the
research design and methodology. In Chapter 4, there is a detailed explanation of the
study implementation and data analysis to explore the relationship between power
generation capacity, power transmission capacity, distribution capability, and industrial
output in Nigeria.
50
Chapter 3: Research Method
The purpose of this quantitative correlational study was to examine the
relationship between industrial output in Nigeria and available monthly power generation
capacity, available monthly transmission capacity, and monthly distribution capability.
Research on the impact of electricity on economic development is vast with varied
outcomes, but none considered electricity as a system. I implemented a systems approach
anchored on von Bertalanffy’s (1968) and Laszlo’s (1996) works to examine the
interrelationships and interdependencies between the elements that make up the
electricity delivery system and industrial output in Nigeria, which provided a better
understanding of how electricity affects economic growth, especially in third world
countries like Nigeria. This chapter contains a discussion of the research methodology
and design for this study, including research questions, data collection, procedures for
analysis, and threats to validity.
Research Design and Rationale
In quantitative methods, researchers use the quantitative properties of variables
systematically and scientifically to investigate the relationships between specific
variables (Edmonds & Kennedy, 2017). Measurement is a critical component of
quantitative research as a means of understanding the relationship(s) between variables as
numerical systems (Edmonds & Kennedy, 2017). I analyzed the relationship between
independent and dependent variables to determine the relationship between electricity
supply variables and industrial output, especially in third world countries like Nigeria.
The independent variables in this study are the available monthly generation capacity,
51
available monthly transmission capacity, and monthly distribution capability. The
dependent variable is the monthly industrial output in Nigeria. Quantitative analysis goes
beyond just a set of techniques applied to data; it represents a process of systematically
analyzing research questions, research methodology, and observed patterns in data
(Scherbaum & Schockley, 2015). Implementing quantitative analysis enables conceptual
models that serve as invaluable resources in choosing appropriate analysis, research
methodology, and research questions. This study involved numeric data and the analysis
of relationships between variables, hence the use of quantitative methods in examining
the research questions.
I implemented a quantitative method in this study with a correlational design. The
use of multiple regression models enabled the evaluation of the relationships between the
independent variables and the dependent variable (Gujarati & Porter, 2009; Wooldridge,
2013). A multiple regression model was utilized to answer the research questions. From
systems thinking perspective, I implemented correlational analysis (pairwise correlation
matrix) to study the relationship between available monthly generation capacity,
available monthly transmission capacity, and monthly distribution capability as
subsystems of the electricity supply system. This type of analysis deepened the
understanding of the interrelatedness and interdependency between elements of the power
supply delivery subsystems. The pairwise correlation coefficients enabled the analysis of
the strength of the interface relationship between the subsystems.
Multiple regression analysis appears consistently in the energy (electricity) and
economic growth literature. Multiple regression is important in analyzing
52
nonexperimental data (Berry & Feldman, 1985). Multiple regression analysis enables the
understanding of the correlation between independent variables and a single dependent
variable (Frankfort-Nachmias & Leon-Guerrero, 2015; Reinard, 2011). Some of the
advantages of multiple regression over simple correlation include testing the interaction
effects between variables and estimating the extent to which a set of independent
variables explain the variations in the dependent variable (Reinard, 2011). The use of
multiple regression also enables the identification of the relative importance of each
variable (Reinard, 2011) and was utilized in this study to test the relative relationship
between each independent variable and the dependent variable in this study.
Multiple correlation analysis cannot infer causality; therefore, the result of a
correlational study can only be used to gauge the extent of the association between
variables and not causation (Asamoah, 2014). I utilized secondary data for this study, and
it presents the issues of applicability and fit for purpose. In this regard, the problems of
availability, relevance, accuracy, and sufficiency were veritable concerns.
Methodology
This section contains the work plan for this study by explaining the procedures
and processes undertaken during this study. Research methodology comprises
assumptions, postulates, and methods that enable a researcher to choose methods and
render the study open to critique, analysis, replication, repetition, and adaptation (Given,
2008). In this regard, research methods are the tools that enable researchers to collect the
data they need and to derive the broad assumptions and procedures that comprise the
overall research methodology of the study.
53
Population
The population for this study consisted of 72 months during the period from 2015
to 2020. The privatization of the Nigerian electricity industry became effective in 2015
through Order 136 of the Nigeria Electricity Regulatory Commission for the
commencement of the transitional electricity market. In this regard, the population size is
72.
Sampling and Sampling Procedures
There was no sampling. I worked with the whole population of 72 months. The
focus of the study was the effects of some preselected operational power systems data on
industrial output in Nigeria. The emphasis was on the sequential impact of variables
based on time, which involved monthly operational data for the power delivery variables
that represent the independent variables, and monthly data on industrial output in Nigeria
that represents the dependent variable. The research question and the associated
hypotheses testing constituted the inclusion criteria. The data sets selected from the
power systems operational data were available monthly power generation capacity,
available monthly transmission capacity, and monthly distribution capacity. Monthly
industrial output was chosen as the national economic data to test the hypothesis and
answer the research questions.
I used G*Power software to calculate the minimum sample size necessary to
answer the research question using multiple linear regression with three predictor
variables (Appendix). For medium effect size, an alpha level of 5%, and 80% power (1-β
error probability), the minimum sample size was 77. A total population size of 72 for
54
each group, three independent variables, and one dependent variable was used to address
the research questions. The population size of 72 for each group was because monthly
operational data from 2015 to 2020 was used to address the research questions. The
effective privatization of the Nigerian electricity industry occurred in 2015 through Order
136 of the Nigeria Electricity Regulatory Commission for commencement of the
transitional electricity market, and this represented a significant structural shift in the
country; therefore, data before 2015 would not be used. In this regard, any data before
2015 could be unreliable for this study as it may introduce issues of structural breaks into
the data with the potential to render the test results unreliable.
Secondary Data
This study consists of an examination of operational data of the electric utilities in
Nigeria and the Nigerian industrial output as a component of the national GDP within the
same period. The data for this study were mainly from the operational database of the
NESO and NBS. The monthly data on generation capacity, reported as declared available
generation capacity, was obtained from the NESO website (https://nsong.org/). The data
for monthly transmission capacity and distribution capability reported as load delivered
to the distribution companies was obtained from the database of NESO. The data on the
industrial output, reported as part of the GDP statistics, was obtained from the NBS
website (https://www.nigerianstat.gov.ng/). Because the NESO and the NBS compiled
this data, I considered these data sets reliable.
Walden University requires that the Walden Institutional Review Board (IRB)
approve study procedures, including data collection processes and procedures for any
55
research. Most of the data for this study are electric utility operational data in third-party
institutional databases. The Walden IRB approval ensures compliance regarding the
ethical usage of third-party data in the approval process for a data request. Most of the
data on industrial output is available on the NBS website, so permission was not required.
Data Analysis Plan
I used the IBM Statistical Package for Social Sciences (SPSS) and STATA
statistical software packages to analyze data in this study. SPSS and STATA were used to
implement multiple regression analyses to examine the relationship between the
dependable variable and three independent variables. Multiple regression analysis enables
the establishment of multiple regression coefficients that measure changes in the
dependent variable per unit change in each of the independent variables while holding the
other independent variables constant (Gujarati & Porter, 2009). The multiple regression
coefficients were used to test the partial effects of generation capacity, transmission
capacity, and distribution capability on industrial output.
There are some data requirements or assumptions connected with the use of
multiple regression analysis. There are eight data assumptions associated with the
implementation of multiple linear regression: (a) the measurement of the dependent
variable should be on a continuous scale (interval or ratio); (b) the study should include
two or more independent variables that should be measured on a continuous scale
(interval or ratio); (c) there must be an independent observation of the variables; (d) there
must be a linear relationship between the dependent variable and each of the independent
variables, and between the dependent and the independent variables collectively; (e) there
56
must be data homoscedasticity; (f) there should not be multicollinearity between the
independent variables; (g) there should not be outliers in the data; and (h) the error term
or residuals should be approximately normally distributed (Laerd Statistics, 2018). A
violation of any of the assumptions could affect the validity of the results; nonetheless,
SPSS and STATA provide mechanisms for normalizing the data when available data do
not meet some of these assumptions (Laerd Statistics, 2018). Operational data of electric
utilities and industrial output, a component of the national GDP data, were used for this
study as independent and dependent variables, respectively.
The research question and study hypotheses formed the basis for exclusion and
inclusion criteria to narrow down the data set. After narrowing down the data set, data
screening was performed for missing data and likely data entry errors. Descriptive
statistics were used to provide summary information about the independent and
dependent variables in the study; the variables are continuous interval ratio.
The following are the research question and hypotheses for this study:
Research question: Is monthly industrial output in Nigeria related to available
monthly power generation capacity, available monthly power transmission capacity, and
monthly power distribution capability?
H0: Monthly industrial output in Nigeria is not related to available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability.
57
Ha: Monthly industrial output in Nigeria is related to available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability.
As systems theory was the theoretical foundation for this study, there was a focus
on the interface relationship between the power sector delivery systems and industrial
output in Nigeria. Systems within a complex system serve as interfaces in coordination
with other parts of the system in a joint effort to sustain the whole system (Laszlo, 1996).
From a systems perspective, I used Pearson’s correlation analysis to examine the
correlation between power generation capacity, transmission capacity, and distribution
capability to analyze the efficacy of the interface relationships between the variables.
This type of analysis is necessary because a poor correlation between power generation
capacity and transmission capacity may mean an inability to adequately or efficiently
transmit the available generation. The same logic holds for the interface relationship
between transmission and distribution sectors. Multiple regression correlation
coefficients were used to examine the effects of each independent variable on the
dependent variable in the study to answer the research questions by controlling for the
effects of two independent variables each time. Multiple regression analysis enabled the
examination of the combined correlation between all the independent variables and the
dependent variable in the study to test the research hypothesis. The predetermined alpha
value for this study was 0.05% (Type I error/false positive); that is, the premise for
significance in this study is a p-value of less than or equal to 0.05 at a confidence interval
of 95%.
58
Threats to Validity
External Validity
Research validity is the various ways factors a researcher has not accounted for in
a study could result in an alternative outcome with the potential to impact the study in
unanticipated ways (Greenstein, 2006). The danger in these scenarios is that the
alternative outcomes can prevent the perception of the true or actual effects of the
independent variables on the dependent variables (Greenstein, 2006). Drew et al. (2008)
presented research validity from a technical viability perspective in terms of rigor and
methodical value. Rigor and methodical value of a study are dependent on the
researcher’s ability to find the most valid information about a hypothesis regarding a
phenomenon (Drew et al., (2008). Drew et al. described external validity in terms of
ability or constraints to the generalizability of the findings from a study. Threats to
external validity occur when researchers draw erroneous inferences from sample data
(Creswell, 2009). In this regard, external validity is about methodological rigor and the
researcher’s ability to account for the effects of variables in a study.
Since this is nonexperimental design research, the effects of external validity are
limited. In this study, the threat to external validity hinges on small sample size issues
and the ability to draw generalizable inferences from such a sample. Because secondary
data was used for the study, there is little room for remedial actions; however, the
population size for this study is adequate, going by the sample size calculation using
G*Power software.
59
Internal Validity
Since this is a nonexperimental study, internal validity issues would be limited to
the problems of omitted variables. Drew et al. (2008) described internal validity in terms
of the effects of extraneous factors (variables) other than the dependent and independent
variables in the study. These issues become poignant when a study involves prediction
and causal analysis as in experimental studies. Since this is primarily a correlational
study, the problem of omitted variables is of limited effect.
Statistical Conclusion Validity
There are no constructs or instruments designed for this study’s specific purposes;
thus, construct validity is of little concern. Since secondary data was used in this study,
applicability, fit for purpose, availability, relevance, accuracy, and sufficiency are more
important validity issues. Literature on research methods indicates that a researcher’s
ability to make a valid conclusion from a study is a critical validity issue. The focus of
SCV is on adequate analysis, the use of appropriate statistical methods, and the ability to
draw correct inferences from the sample used to answer the research question(s) (García-
Pérez, 2012). The need for SCV arose because conclusions drawn from insufficient data
analysis sometimes could be at variance with those from appropriate data analysis
(García-Pérez, 2012). In this regard, SCV is about how well researchers analyze data and
the extent to which correct inferences are made from such research.
Multiple issues could lead to SCV violations. Low statistical power is one of the
threats to SCV and manifests as type II error, which results from a small sample size
(Busk, 2010). I used G*Power software with 80% statistical power to determine 77 as the
60
minimum sample size to avoid SVC issues associated with low statistical power. The
available population size for this study is 72; this difference was not considered
significant enough to cause SCV issues. Further, The violation of an assumption or
assumptions of test statistics might result in wrong inferences and lead to SCV issues
(Busk, 2010). Another SCV issue is the unreliability of measurement instruments that
could result in type II error. Secondary data was used for this study and was obtained
from national institutions that keep both the electricity utility data (NESO) and national
economic data (NBS). In this regard, there is a high tendency that the data will be of high
integrity, and as such, the unreliability of measures will not affect SCV for this study.
Ethical Procedures
Consideration for research ethics is one of the protocols or processes for
completing research at Walden University. The main task of the Walden Institutional
Review Board (IRB) is to make sure that research performed in the university satisfies
Walden university standards and complies with requirements of U.S. federal regulations.
Intending researchers are expected to fill out Form A as a first step in obtaining the
Walden IRB approval (Walden University, n.d.). Core to the Walden IRB approval is
ethical concerns regarding the treatment of human participants, institutional permissions,
recruitment of material and processes, data collection or intervention activities, and how
the researcher seeks to mitigate or manage these issues and treat data. Other ethical issues
of concern include how confidential data will be stored and managed to maintain
confidentiality.
61
This research involves using secondary data and does not include the recruitment
of human participants or intervention activities. The first step in ethical procedures for
this study was completing Form A and all the associated processes as laid out in the
Walden IRB process web page (https://academicguides.waldenu.edu/research-center).
Since this study involves using secondary data in the form of operational data of electric
utilities in Nigeria and the industrial output data, which is public domain data published
by the NBS, most of the ethical issues involving humans were avoided. Notably, there
were no interaction with humans, recruitment materials, and data concerning humans.
Since this is not experimental design research, there was no intervention during the study,
thus precluding ethical concerns.
Because of the nature of the data source for this study, institutional permits were
not needed when such data were already in the public domain. The electric utilities’
operational data was password protected and stored in a secure location, including a
personal laptop and a secure cloud storage facility. This research is not within my work
environment; thus, issues of conflict of interest and power differentials were avoided.
Since humans are not involved, ethical concerns regarding the use of incentives did not
arise.
Summary
The purpose of this multiple regression correlational study was to use a systems
approach to study the effect of electricity on industrial output in Nigeria. In this regard,
electricity is treated as a natural system to investigate the relationship between the power
delivery system elements and industrial output. In this chapter, I examined and described
62
the research design, methodology, and data analysis plan for the study. The issues of
threats to validity and ethical considerations regarding the study were also investigated.
Chapter 4 includes a description of the study execution and the data analysis approach
used to explore the relationship between industrial output in Nigeria and available
monthly power generation capacity, available monthly power transmission capacity,
monthly distribution capability. Chapter 5 includes an interpretation of the results of data
analysis from Chapter 4 and recommendations arising from the data analysis.
63
Chapter 4: Results
The purpose of this study was to establish whether monthly industrial output in
Nigeria was related to available monthly power generation capacity, available monthly
transmission capacity, and monthly distribution capability. Researchers have analyzed the
relationship between electricity and economic development, mainly through the lens of
electricity consumption. In this study, I used a system thinking approach to disaggregate
electricity into subsystems that constitute electricity delivery system, other than
electricity consumption, as predictor variables to broaden the understanding of the
relationship between electricity and economic development. The research question and
hypotheses that guided this research were:
RQ: Is monthly industrial output in Nigeria related to available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability?
H0: Monthly industrial output in Nigeria is not related to available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability.
Ha: Monthly industrial output in Nigeria is related to available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability.
The independent variables are the available monthly power generation capacity,
available monthly transmission capacity, and monthly power distribution capability. In
64
this study, industrial output in Nigeria is the dependent variable and is measured in Naira,
the Nigerian currency.
In this chapter, I discuss the data collection processes for this study. This chapter
also includes the presentation of descriptive analysis of the study variables and the
pairwise correlation analysis to examine the strength and direction of correlation between
the independent variables and multiple regression analysis to reveal the effect of each
independent variable on the dependent variable. The results of the analyses facilitated the
answering of the research question.
Data Collection
Time Frame, Recruitment, and Response Rates
There were no issues of recruitment or response rates as secondary data were used
for this study. The NESO website (https://nsong.org/) was the source of daily power
generation, transmission, distribution data from 2015 to 2020. I aggregated these data
into average monthly values for the study. The data on monthly industrial output in
Nigeria were derived from the NBS website (https://www.nigerianstat.gov.ng/) for the
same period.
Data Collection Discrepancies, Baseline, and Demographic Characteristics of
Sample
There were no significant discrepancies in the data collection plan I presented in
Chapter 3 and the actual processes and data collected for the study. Data collection for
the study was only after the IRB approval (approval no. 07-21-21-0477044). The whole
65
population was used for the analysis; hence, there was no sampling during data collection
and collation.
The available monthly generation capacity represents the aggregated available
generation capacity from all power generating plants in Nigeria measured in MW. The
transmission capacity, measured in MW, represents the capacity of the entire power
transmission network in Nigeria and is reported on the NESO website daily. Distribution
capability represents the total daily energy the distribution companies in Nigeria
distribute, reported as energy utilized from the NESO website. The industrial output, the
dependent variable in the study, represents the GDP component of all the industrial
sectors in Nigeria as reported in the NBS website.
Representative Sampling
I used G*Power 3.1 software to determine that the minimum sample size for the
study was 77, but the available sample size was 72 and they were all used in the study
analysis. There was no sampling since I used all the available population as my sample.
The assumption was that the difference would not adversely affect the validity and
generalizability of the study.
Study Results
Descriptive Statistics
The data for this study are GDP and operational data for the electricity industry in
Nigeria and presents peculiar considerations regarding the nature of data and implications
for operational performance. Though the average GDP for the period under study was
521,191 million Naira (SD = 22,654), the maximum and minimum values show a
66
difference of about 100,000. Figure 1 shows that the difference in the minimum and
maximum values was not due to economic improvement trends; rather, it indicated peaks
and dips over the period. The mean generation capacity was 5,005 MW (SD = 863) with
minimum and maximum values of 2,893 MW and 6,718 MW, respectively.
Figure 1
Industrial Output
Figure 2 indicates that the disparity in generation capacity was not due to growth
over the period. The transmission capacity recorded over the period ranged from 5,890
MW to 8,060 MW, with a mean value of 6,945 MW (SD = 844). Figure 2 shows an
upward trend in transmission capacity and could be the reason for the difference between
the minimum and maximum transmission capacity over the period. The average
distribution capability during the period was 3,404 MW (SD = 397) with minimum and
maximum values of 1940 MW and 4,240 MW, respectively. From Figure 2, it is likely
that operational issues, rather than capacity growth, explain the high disparity between
the minimum and maximum distribution capacity values. The assumption was that the
data is of high quality.
400,000.00
450,000.00
500,000.00
550,000.00
600,000.00
2015 2016 2017 2018 2019 2020
Naira (Million)
Industrial Output
67
Figure 2
Trend Analysis for Power Delivery Subsystems
Examining the descriptive statistics shown in Table 4 from systems thinking
perspective reveals capacity mismatch across the power sector delivery value chain.
-
2,000.00
4,000.00
6,000.00
8,000.00
10,000.00
2015 2016 2017 2018 2019 2020
MW
Declared GX Capacity (MW) Average TRX Capacity (MW)
Average DX Capacity (MW)
68
Table 4
Descriptive Statistics
Industrial
Output
(Million
Naira)
Generation
Capacity
(MW)
Transmission
Capacity
(MW)
Distribution
Capability
(MW)
N
Valid
72
72
72
72
Missing
0
0
0
0
Mean
Statistic
521191
5005
6945
3404
Std. Error
2669
101
99
46
Median
525650
5090
6866
3486
Std.
Deviation
22654
863
844
397
Minimum
460578
2893
5890
1940
Maximum
563522
6718
8060
4240
Skewness
Statistics
-.447
-.227
.067
-1.01
Std. Error
.283
.283
.283
.283
Kurtosis
Statistics
.03
-.716
-1.83
2.10
Std. Error
.559
.559
.559
.559
Linear Correlation
I developed a correlation matrix using Pearson’s product-moment correlation to
examine the relationship between the subsystems that constitute the electricity supply
system in Nigeria, including available monthly generation capacity, available monthly
transmission capacity, and monthly distribution capability. The essence was to use the
strength and direction of correlation between the subsystems as a nexus for exploring the
interface relationship and interrelatedness between the elements that constitute the power
delivery value chain. The Pearson’s correlation was also used to evaluate the relationship
between industrial output and subsystems that comprise the electricity supply system in
Nigeria. There are five assumptions for the implementation of Pearson’s product-
69
moment: (a) the variables should be continuous; (b) there should be an equal observation
of the variables; (c) there should be a linear relationship among the variables; (d) no
significant outliers; and (e) for inferential statistics, the variables should satisfy the test
for bivariate normality (Laerd Statistics, 2018). As has been established earlier, the
variables satisfy assumptions (a) and (b). A visual inspection of the SPSS scatterplot
shows a linear relationship between the independent variables (see Figure 3, Figure 4,
and Figure 5). The scatterplots also revealed no outliers in the data; hence four of the
assumptions for correlational analysis were met.
Though Figure 3, Figure 4, and Figure 5 show some form of linearity or
correlation between the elements of the power supply value chain (generation capacity,
transmission capacity, and distribution capability), Figure 2 indicated otherwise. Figure 2
shows little or no symmetry in the trends of the elements of the power delivery system.
For a value flow process like power generation, transmission, and distribution, the
expectation is that the trends ought to be tightly coupled. The logic behind this assertion
is that electricity generated needs to be transmitted, and electricity transmitted needs to be
distributed to end users for consumption or utilization for residential, commercial, or
industrial purposes. The lack of symmetry in trends portends little or weak or no
correlation between the power delivery value chain elements. Figure 2 indicates that the
data is time series in nature, and there is no independence of observations. Time series
data are values or observations of a variable taken at defined time intervals over a period
(Gujarati & Porter, 2009; Shrestha & Bhatta, 2017).
70
Overall, Figure 3 shows little or weak linearity between generation capacity and
transmission capacity. Figure 4 shows a moderate linear relationship between generation
capacity and distribution capability. This conclusion is also observable in Figure 2.
Figure 5 shows a weak linear relationship between transmission capacity and distribution
capability. Figure 2 indicates no relationship between transmission capacity and
distribution capability.
Figure 3
Plot of Generation Capacity and Transmission Capacity
71
Figure 4
Plot of Generation Capacity and Distribution Capability
Figure 5
Plot of Transmission Capacity and Distribution Capability
Due to some unique qualities of time series, like a trend, most common analytical
methods might not be applicable (Shrestha & Bhatta, 2018). Implementing the correct
72
methodology in analyzing time series data is critical as the wrong approach (model)
could lead to biased and unreliable estimates (Shrestha & Bhatta, 2018). A typical feature
of time series data is that it is either stationary or nonstationary. A time series data is
stationary if its trend values like mean and variance, remain constant or tend to revert to
its original value after every change in time; that is, the data properties are not affected by
time (Gujarati & Porter, 2009; Shrestha & Bhatta, 2017). Time series is nonstationary if
the mean, variance, and covariance of the data change over time and thus contain a unit
root (Gujarati & Porter, 2009; Shrestha & Bhatta, 2017). When time series data is
nonstationary, it is impossible to generalize such data over different periods. In this
regard, it is usual to transform nonstationary time series data to its stationary form to
make it more amenable to most common data analytical methods (Gujarati & Porter,
2009).
I conducted unit root tests to determine the stationarity of the independent and
dependent variables using the Augmented Dickey-Fuller (ADF) test. ADF and Phillips-
Perrone unit root tests are the most widely used to test for stationarity of time series data
(Gujarati & Porter, 2009; Shrestha & Bhatta, 2017). The ADF tests indicated that all the
independent variables were nonstationary (see Table 5, Table 6, and Table 7. The ADF
test indicated that industrial output, the dependent variable, was stationary (see Table 8).
For all the independent variables, the absolute values of the test statistic were smaller
than the critical values at 5%, indicating that the variables were nonstationary, but for the
dependent variable, the absolute value of the test statistic was higher than the critical
value at 5% indicating that it was stationary (Green, 2012).
73
Table 5
Generation Capacity Augmented Dickey-Fuller Test for Unit Root
Test Statistic
Critical Value
1%
5%
10%
Z(t)
-2.546
-3.552
-2.914
-2.592
MacKinnon approximate p-value for Z(t) = 0.0632.
Table 6
Transmission Capacity Augmented Dickey-Fuller Test for Unit Root
Test Statistic
Critical Value
1%
5%
10%
Z(t)
-2.088
-4.106
-3.480
-3.168
MacKinnon approximate p-value for Z(t) = 0.8112.
Table 7
Distribution Capability Augmented Dickey-Fuller Test for Unit Root
Test Statistic
Critical Value
1%
5%
10%
Z(t)
-2.760
-3.552
-2.914
-2.592
MacKinnon approximate p-value for Z(t) = 0.0693
Table 8
Industrial Output Augmented Dickey-Fuller Test for Unit Root
Test Statistic
Critical Value
1%
5%
10%
Z(t)
-3.570
-3.552
-2.914
-2.592
MacKinnon approximate p-value for Z(t) = 0.0064
I implemented the first difference method to transform the independent variables
to stationary time series data, including the dependent variable. The transformation of the
dependent variable was to ensure that all the variables were in the same level form during
the correlation analysis. The ADF test was used to determine that the dependent variable
74
remained stationary in the first difference form. First differencing is one of the data
transformation methods for nonstationary time series data (Gujarati & Porter, 2009;
Shrestha & Bhatta, 2018). First differencing will be discussed further in the next section.
After first differencing, further tests using the ADF method indicated that all the
independent variables became stationary. Table 9 shows the result of Pearson’s
correlation analysis conducted to assess the linear relationship between industrial output
in Nigeria and available monthly generation capacity, available monthly transmission
capacity, and monthly distribution capability. The population size reduced to 71 due to
data transformation arising from the first differencing. The result from Table 9 is
consistent with interpretations from Figure 2 for the independent variables. The Pearson’s
correlation analysis reveals that there is no statistically significant correlation between
generation capacity and transmission capacity r (67) = .04, p > .05, and there is a
negative but no statistically significant correlation between transmission capacity and
distribution capability r (67) = - .10, p > .05. The results also indicate a statistically
significant small correlation between industrial output and distribution capability r (67) =
.25, p < .05, with distribution capability explaining 6% of the variation in industrial
output. There was negative but not statically significant relationship between industrial
output in Nigeria and generation capacity r (67) = -.07, p > .05, and transmission capacity
r (67) = -.15, p > .05. The finding that only distribution capability has a statistically
significant relationship with industrial output has an important implication for electricity
delivery in Nigeria.
75
Table 9
Pearson’s Correlations for the Dependent and Independent Variables
Industrial
Output
Generation
Capacity
Transmission
Capacity
Distribution
Capability
Industrial
Output
Pearson’s
Correlation
1
-.071
-.154
.247*
Sig. (2-tailed)
.556
.199
.038
N
71
71
71
71
Generation
Capacity
Pearson’s
Correlation
-.071
1
-.035
.299*
Sig. (2-tailed)
.556
.770
.011
N
71
71
71
71
Transmission
Capacity
Pearson’s
Correlation
-.154
-.035
1
-.103
Sig. (2-tailed)
.199
.770
.394
N
71
71
71
71
Distribution
Capability
Pearson’s
Correlation
.247*
.299*
-.103
1
Sig. (2-tailed)
.038
.011
.394
N
71
71
71
71
*. Correlation is significant at the 0.05 level (2-tailed).
Multiple Regression Analysis
In this study, I used multiple regression analysis, α = .05 (two-tailed), to examine
the relationship between monthly industrial output, the dependent variable, and available
monthly generation capacity, available monthly transmission capacity, and monthly
distribution capability, the independent variables, and to estimate the extent to which
each independent variable explains variations in the dependent variable. Preliminary
analysis was conducted to test the assumptions for multiple linear regression, including
tests for linearity, level of measurement (scale or interval), number of variables (two or
more), presence of autocorrelation, and multicollinearity, homoscedasticity, and
normality.
76
Assumption Testing
From the correlation analysis, the assumptions for the number of variables and
level of measurement were satisfied. The test for autocorrelation using the Durbin-
Watson statistic (d) yielded a value of .710, see Table 10. Further, d ranges from 0 to 4
and a d = 2 indicates the absence of autocorrelation, values of d above 2 indicate negative
autocorrelation, and values below 2 show the presence of positive autocorrelation (Laerd
Statistics, 2018; Gujarati & Porter, 2009). In this regard, the model showed the presence
of positive autocorrelation. This result was not unexpected given the time series nature of
the data and unit roots in the independent variables as established during Pearson’s
correlation analysis. Autocorrelation indicates a correlation between data points or
observations ordered in time (Laerd Statistics, 2018; Gujarati & Porter, 2009).
Table 10
Model Summaryb
Model
R
R2
Adjusted
R
2
Std. Error of
the Estimate
Durbin-
Watson
1
.352a
.124
.085
21668.522
.710
a. Predictors: (Constant), Available Distribution Capacity, Available Transmission
Capacity, Available Generation Capacity
b. Dependent Variable: Industrial Output
First differencing is one of the methods of correcting autocorrelation in a linear
regression model (Gujarati & Porter, 2009). First differencing involves the subtraction of
the lagged value of a variable from the preceding value in time. In this regard, first
differencing will result in the loss of one observation as the first observation does not
have a precedent (Gujarati & Porter, 2009). I implemented the 1st order differencing
77
method to the data, and the d statistic improved to 2.01, see Table 11, indicating the near
absence of autocorrelation and the presence of 1st order autocorrelation (Gujarati &
Porter, 2009).
Table 11
Model Summaryb
Model
R
R2
Std. Error of
the Estimate
Durbin-
Watson
1
.318a
.101
.061
17873.31286
2.015
a. Predictors: (Constant), DIFF(DistCap,1), DIFF(TransCap,1), DIFF(GenCap,1)
b. Dependent Variable: DIFF(IndPut,1)
The population linear regression model for this study is represented by:
InduPuti = β1 + β2GenCapi + β3TrxCapi + β4DxCapi+ ui (4.0)
Where InduPut is the dependent variable (industrial output), GenCap, TrxCap, and
DxCap are generation capacity, transmission capacity, and distribution capability,
respectively, u is the stochastic disturbance term or the error term, and i is the ith
observation and can be replaced by t in a time series (Gujarati & Porter, 2009). In
equation 4.0, β1 is the intercept term, and β2, β3, and β4 are the partial regression
coefficients of the independent variables. The equation 4.0 can be rewritten as:
InduPuti-1 = β1 + β2GenCapi-1 + β3TrxCapi-1 + β4DxCapi-1 + ui-1 (4.1)
Where InduPuti-1, GenCapi-1, TrxCapi-1, DxCapi-1 are the lagged values of the dependent
and independent variables respectively, and ui-1 is the lagged value of the error term.
First differencing involves subtracting equation 4.1 from equation 4.0, which yields:
ΔInduputi = β2ΔGenCapi + β3ΔTrxCapi + β4ΔDxCapi + Δui (4.2)
78
In equation 4.2, Δ is known as the difference operator and indicates the application of
successive differences of the variables in the equation (Gujarati & Porter, 2009). That is,
ΔInduputi = (InduPuti - InduPuti-1), ΔGenCapi = (GenCapi - GenCapi-1), ΔTrxCapi =
(TrxCapi - TrxCapi-1), ΔDxCapi = (DxCapi - DxCapi-1), and Δui = (ui - ui-1).
Where vi = Δui = (ui - ui-1).
In this regard, equation 4.2 can be written as:
ΔInduputi = β2ΔGenCapi + β3ΔTrxCapi + β4ΔDxCapi + vi (4.3)
According to Gujarati and Porter 2009, equation 4.1 is the level form, and equation 4.2 is
the first difference form of the regression model.
The scatter plot of the studentized residuals (SRE_1) against the unstandardized
predicted values (PRE_1) was used to determine the existence of a collective linear
relationship between the dependent and independent variables (Laerd Statistics, 2018),
See Figure 6. Figure 6 shows some form of a collective linear relationship between the
dependent and independent variables. Partial regression plots between each independent
variable and the dependent variable were used to test the assumption for linearity. No
significant violations for linearity were identified, see Figure 7, Figure 8, and Figure 9.
To check for heteroscedasticity, the review of the scatter plot of the studentized residuals
(SRE_1) against the unstandardized predicted values (PRE_1) showed some form of
increasing funneling of the data points that may indicate heteroscedasticity (Laerd
Statistics, 2012). The collinearity statistics were measured using the variance inflation
factor (VIF) and showed that the highest VIF in the model was 2.37, indicating the
absence of multicollinearity, see Table 12. Laerd Statistics (2018) noted that VIF values
79
lower than 10 indicate the absence of multicollinearity. An inspection of the Cook’s
distance variable (COO_1) for the model showed that none of the cases had a value
greater than 1. A COO_1 value greater than 1 should be investigated as it may indicate
outlier(s) in the dataset (Laerd Statistics, 2018). In this regard, there was no violation of
the assumption for outliers in the model. A review of the histogram with superimposed
normal curve and the P-P plot for the regression model showed that there was no
significant violation for the test for normality, though the OLS regression is still reliable,
to some extent, in the presence of deviations from normality (Laerd Statistics, 2018), see
Figure 10 and Figure 11.
Figure 6
Plot of Collective Linearity of the Dependent and Independent Variables
80
Figure 7
Plot of Linearity Between Industrial Output and Generation Capacity
Figure 8
Plot of Linearity Between Industrial Output and Transmission Capacity
81
Figure 9
Plot of Linearity Between Industrial Output and Distribution Capability
Table 12
Multicollinearity Statistics
Tolerance
VIF
Generation
Capacity
.423
2.366
Transmission
Capacity
.552
1.810
Distribution
Capability
.545
1.834
82
Figure 10
Histogram with Superimposed Normal Curve to Interpret Normality
Figure 11
Normal P-P Plot to Check for Normality
Statistical Analysis Findings
Research question: Is monthly industrial output in Nigeria related to available
monthly power generation capacity, available monthly power transmission capacity, and
monthly power distribution capability?
83
H0: Monthly industrial output in Nigeria is not related to available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability.
Ha: Monthly industrial output in Nigeria is related to available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability.
The tests for the assumptions for multiple linear regression indicated the
likelihood of heteroscedasticity and autocorrelation in the dataset. All models are wrong
to a certain degree and the best any researcher can achieve is to approximate the process
of modeling and estimation of the outcome variable (Hayes and Cai, 2007). This
observation is regarding the difficulty in strictly satisfying all the OLS assumptions and
the subjective or qualitative nature of validating some assumptions, especially for
heteroscedasticity and linearity using graphical plots (Gujarati & Porter, 2009). Using
heteroscedasticity-consistent standard error estimators in estimating OLS parameters
could reduce the effects of heteroscedasticity in OLS models without the requirements for
the transformation of variables (Hayes and Cai, 2007). The Newey-West method of
correcting OLS standard errors, known as the heteroscedasticity and autocorrelation-
consistent standard errors or Newey-West standard errors, in addition to correcting for
heteroscedasticity, also corrects for autocorrelation of errors (Gujarati & Porter, 2009). In
this regard, the Newey-West standard errors method obviates the need for data
transformations in the presence of heteroscedasticity and autocorrelation.
84
With likely heteroscedasticity and autocorrelation in the model, I ran an OLS
regression with the Newey-West errors estimator using STATA to account for both
autocorrelation of errors and heteroscedasticity and avoid the dangers of type I error that
may arise due to either heteroscedasticity or autocorrelation, or both in the model. The
reason for using STATA is because SPSS does not support the function for
heteroscedasticity and autocorrelation-consistent standard errors analysis. The regression
coefficients and standard errors are shown in Table 13. Only distribution capability (t =
2.17, p < .05) statistically significantly predicted industrial output in the model and is
consistent with the findings from the Pearson’s correlation analysis. In this regard, I
rejected the null hypothesis that monthly industrial output in Nigeria is not related to
available monthly power generation capacity, available monthly power transmission
capacity, and monthly power distribution capability.
Table 13
Regression with Newey-West Standard Errors
Industrial
Output
Coefficient
Newey-
West std.
err.
t
P>|t|
[95% conf. interval]
Generation
Capacity
1.282343
4.40031
0.29
0.772
-7.49834
10.06302
Transmission
Capacity
-1.199524
3.639242
-0.33
0.743
-8.46152
6.06247
Distribution
Capability
19.39199
8.954204
2.17
0.034
1.52415
37.25982
cons_
457077.8
22308.92
20.49
0.000
412561.1
501594.6
Number of observations = 72, Maximum lag = 1
The resulting multiple regression equation for the model is represented by:
InduPut = 457077 + 1.28GenCap – 1.2TrxCap + 19.4DxCap (4.4)
85
Where InduPut is the dependent variable (industrial output), the independent variables
GenCap, TrxCap, and DxCap are available monthly generation capacity, available
monthly transmission capacity, and monthly distribution capability, respectively. Given
that distribution capability was the only independent variable that was statistically
significant, equation 4.4 would not accurately represent the relationship between the
dependent and independent variables in the study. In this regard, I refitted the regression
model using a simple linear regression with monthly industrial output as the dependent
variable and monthly distribution capability as the only independent variable. The results
of the simple linear regression are shown in Table 14.
Table 14
Refitted Regression with Newey-West Standard Errors
Industrial
Output
Coefficient
Newey-
West std.
err.
t
P>|t|
[95% conf. interval]
Distribution
Capability
19.91581
5.165661
3.86
0.000
9.613228
30.2184
cons_
453382.1
18065.99
25.10
0.000
417350.6
489413.6
Number of observations = 72, Maximum lag = 12
The multiple regression equation for the final model is represented by:
InduPut = 453382.1 + 19.2DxCap (4.5)
Where InduPut is the dependent variable (industrial output), the independent
variable DxCap is monthly distribution capability. In the final model, distribution
capability statistically significantly predicted industrial output (t = 3.86, p < .000), and
distribution capability accounted for 12% of the variability in industrial output in Nigeria.
The regression model indicates that a 1MW increase in distribution capability will result
86
in a 19.2 Naira increase in industrial output in Nigeria. This result indicates that an
increase in distribution capability will lead to a rise in industrial output, which could
positively affect the overall GDP of the country. The final model also shows that when
distribution capability is zero, industrial output will be 453382.1, where 453382.1 is the
constant or intercept term. This scenario is plausible given Osakwe’s (2017) finding that
71% of industries in Nigeria generate their power, suggesting that even without power
supply from the distribution segment of the power supply value chain, there would still
be a significant level of industrial production in Nigeria.
Summary
The purpose of this quantitative correlational study was to establish if monthly
industrial output in Nigeria was related to available monthly power generation capacity,
available monthly transmission capacity, and monthly distribution capability. Pearson’s
correlation analysis revealed no statistically significant correlation between generation
capacity and transmission capacity. The correlation analysis also revealed that there was
no statistically significant correlation between distribution capability and transmission
capacity, and only distribution capability was statistically significantly related to
industrial output in Nigeria. The results indicated a statistically significant small
correlation between industrial output and distribution capability r (67) = .25, p < .05, with
distribution capability explaining 6% of the variation in industrial output.
The multiple regression analysis enabled the examination of the extent to which
generation capacity, transmission capacity, and distribution capacity predict industrial
output in Nigeria. The OLS regression with the Newey-West errors estimator revealed
87
that only distribution capability statistically significantly predicted industrial output (t =
3.86, p < .000), and distribution capability accounted for 12% of the variability in
industrial output in Nigeria. The result is consistent with the outcome of the Pearson’s
correlation analysis. The interpretation of these findings is discussed in Chapter 5 with
associated literature, limitation of the study, and recommendations for future studies.
88
Chapter 5: Discussion, Conclusions, and Recommendations
The purpose of this quantitative correlational study was to examine the
relationship between monthly industrial output in Nigeria and available monthly power
generation capacity, available monthly power transmission capacity, and monthly power
distribution capability. Based on von Bertalanffy’s (1968) and Laszlo’s (1996) seminal
works, I used the systems theoretical perspective to identify and examine how the power
sector variables affect industrial output in Nigeria because little is known about how the
elements that make up the electric power supply value chain impact industrial output in
Nigeria.
In this study, I presented electricity as a natural open system consisting of
available monthly power generation capacity, available monthly transmission capacity,
and monthly power distribution capability as independent variables and sought to
examine the interrelatedness between these variables. Pearson’s correlation analysis was
used to examine the correlation between the elements that constitute Nigeria’s electricity
supply value chain and the relationship between the dependent and independent variables.
The results indicated no statistically significant correlation between available generation
capacity and available transmission capacity r (67) = .04, p > .05, and there is a negative
but no statistically significant correlation between transmission capacity and distribution
capability r (67) = - .10, p > .05. The results also showed a statistically significant
relationship between industrial output and distribution capability r (67) = .25, p < .05.
In addition to the Pearson’s correlation analysis, I conducted a multiple linear
regression analysis to examine the relationship between the industrial output, the
89
dependent variable, and available monthly generation capacity, available monthly
transmission capacity, and monthly distribution capability, the independent variables, and
to estimate the extent to which each independent variable explains variations in the
dependent variable. I ran an OLS regression with the Newey-West errors estimator to
account for both autocorrelation of errors and heteroscedasticity. The final model
indicated that only distribution capability (t = 3.86, p < .000) statistically significantly
predicted industrial output.
Interpretation of Findings
Contribution to Literature
Most researchers in the electricity and economic development literature use
electricity consumption as an independent variable in their analysis. The scholars who
studied the electricity economic development nexus in Nigeria used either electricity
supply or electricity consumption as an independent variable. The results of the study in
SSA countries indicate that the poor state of electric infrastructure and services
negatively impact economic development (2020; Azolibe & Okonkwo, 2020; Chakamera
& Alagidede, 2018; Kodongo & Ojah, 2016; Owusu-Manu et al., 2019). A veritable
interpretation of these findings is that the poor state of electricity infrastructure moderates
the possible effects of electricity on economic development, thus making electricity
consumption or supply a constrained variable.
The Pearson’s correlation analysis results from this study indicated little or weak
correlation between the elements of the power supply value chain (generation capacity,
transmission capacity, and distribution capability). There was no statistically significant
90
correlation between available generation capacity and available transmission capacity r
(67) = .04, p > .05, and there was negative but no statistically significant correlation
between transmission capacity and distribution capability r (67) = - .10, p > .05. The
significant difference in means between generation capacity and distribution capability
indicates the presence of possible constraints in distribution capability. These constraints
could be due to poor power distribution infrastructure and services or constraints at the
transmission and distribution interface points. This finding aligns with the study on the
effects of poor electricity infrastructure and economic development in SSA countries.
Examining the difference in means in conjunction with the visual interpretation of
Figure 2 and Pearson’s correlation analysis indicates the possibility of stranded
generation capacity during the period under study. This realization or possibility is a
contradiction given the unenviable position Nigeria occupies in the league of nations
regarding electricity access gap. Osakwe’s (2017) finding that 71% of industries generate
their power further reinforces the notion of stranded (idle) generation capacity amidst
established scarcity, likely caused by constraints in the distribution capability or at the
distribution and transmission interface points. The likelihood of unutilized generation
amidst scarcity is a significant outcome of this study.
The prospect of constraints in distribution capability in Nigeria, arising from
Pearson’s correlation analysis results, has significant implications in the electricity and
economic development literature, especially for energy-constrained nations like Nigeria.
Constraints at the transmission/distribution interface points or distribution capability
within the power supply value chain indicate that using electricity consumption as an
91
independent variable in the electricity and economic development literature could lead to
misleading or spurious conclusions.
The results of the multiple linear regression with the Newey-West errors estimator
showed that only distribution capability in the regression model statistically significantly
predicted industrial output. If distribution capability is constrained, either at the
transmission and distribution interface points or within the distribution system, the
conclusion that distribution capability significantly predicts industrial output could be
misleading. This assertion is an important outcome of this study and a significant
contribution to electricity and economic development literature. If there were no
constraints, like in energy-sufficient countries, the results might be different.
From the Africa energy report for 2014, accessed from the website of the
International Energy Agency (https://www.iea.org/), growth in household electricity
consumption is about 70% in Nigeria. The multi-year tariff order, accessed from the
website of the Nigerian electricity regulator (https://nerc.gov.ng/), indicated that
commercial and industrial use of electricity account for 27% of total consumption, with
the industrial sector accounting for only 7% of total consumption. Within the context of a
high prevalence of self-generation within the industrial sector, these findings indicate that
removing the perceived constraints could skew consumption more toward residential and
commercial use than for industrial purposes. In this regard, removal of the perceived
constraints in electricity distribution as indicated from this study could result in different
outcomes regarding the results of the Pearson’s correlation and multiple regression
92
analyses, especially regarding the correlation between distribution capability and
industrial output.
The capacity misalignment across the elements of the power delivery value chain
in Nigeria, as adduced from the difference in means and Pearson’s correlation analysis,
confirms literature on the negative effects of poor electricity infrastructure on economic
development in SSA countries. This study offers a deeper understanding of the
relationship between electricity and economic development in electricity-challenged
countries. In this regard, this study divides literature in electricity and economic
development into literature for energy-sufficient and literature for energy-challenged
countries, especially regarding support for policy development.
Theoretical Implication of Findings
Systems thinking was the theoretical framework for this study, anchored on the
seminal works of von Bertalanffy (1968) and Laszlo (1996). The systems approach
enabled understanding electricity supply or consumption as a system made up of
interdependent subsystems. Laszlo presented the hypothetico-deductive method to
examine the interface relationships between systems and the joint effort required to
maintain a dynamic equilibrium. In this study, I used a systems approach to identify and
investigate the interface relationships between power generation, transmission, and
distribution, as elements of the power delivery system and how these variables affect
industrial output in Nigeria.
Laszlo (1996) developed four propositions with which to investigate and verify
the existence of systems. One of Laszlo’s four propositions about natural systems is that
93
they serve as coordinating interfaces in nature’s holarchy. Holarchic duality represents
the joint efforts systems need to maintain dynamic equilibrium (Laszlo, 1996).
The reason for Pearson’s correlation analysis in this study was to test the
holarchic duality properties of the Nigerian power supply value chain as a natural system.
I used the correlation coefficient to test the extent or strength of relationship and
interdependence between the elements of the power supply system (generation capacity,
transmission, capacity, and distribution capability). As a natural system, the expectation
was that the strength of correlation between the Nigerian power supply system elements
would be strong or close to unity to confirm the holarchic duality property of a natural
system. The Pearson’s correlation analysis results indicate weak or no correlation
between the Nigeria power supply system elements. There was no statistically significant
correlation between generation capacity and transmission capacity, and there was a
negative but no statistically significant correlation between transmission capacity and
distribution capability. These findings indicate interface issues among the power delivery
value flow elements or infrastructure deficits within the subsystems that constitute the
power delivery system in Nigeria.
The value of the systems approach is in identifying misalignment of capacity in
the Nigerian power supply value chain. This misalignment is with an attendant poor
power supply amidst idle generation capacity. Using Laszlo’s (1996) concept of holarchic
duality, it is evident from Pearson’s correlation analysis that the coordinated joint effort
required for the attainment of dynamic equilibrium at the interface points for a natural
system is lacking in the Nigerian power supply value chain. These findings exemplify
94
how the use of theory could influence research and how outcomes of such research could
support policy development and practice in terms of system planning and development
Limitations of the Study
There were a few limitations to the study with the potential to impact the validity,
trustworthiness, generalizability, and reliability. The a priori power analysis at 80%
power (1-β error probability), using the G*Power 3.1 software, determined 77 as the
minimum sample size; however, the available population size was 72. Small sample size
issues introduce threats to SCV (Busk, 2010). This issue was made worse by
implementing the first difference data transformation methodology, which further
reduced the sample size to 71.
Natural gas, a fuel source for thermal generation in Nigeria, a possible covariate
in this study, was not included as a variable in the study. The NESO (https://nsong.org/)
frequently reported lack of gas as a constraint to available generation capacity. In 2015,
thermal generation constituted up to 82% of the power generated in Nigeria (Osakwe,
2017). Extraneous variables that could affect the study, other than the identified
independent and dependent variables, could threaten internal validity (Drew at al., 2008).
Excluding this variable could be a threat to internal validity as it could be interpreted as
an omitted variable.
Another limitation of the study is the use of secondary data. The use of secondary
data presents the issues of applicability and fit for purpose. In this regard, issues of
availability, relevance, accuracy, and sufficiency become veritable concerns.
Applicability and fit for purpose issues did not arise in this study. The Nigerian system’s
95
operator was the sole aggregator of operational data for the electricity industry in Nigeria,
and there was no means of cross-validating the data posted by this agency. In this regard,
the issue of data quality remains a veritable concern.
Recommendations
The Pearson’s correlation analysis showed a statistically significant small
correlation between industrial output and distribution capability r (67) = .25, p < .05, with
distribution capability explaining 6% of the variation in industrial output. The regression
analysis indicated that distribution capability statistically significantly predicted industrial
output (t = 3.86, p < .000), and distribution capability accounted for 12% of the
variability in industrial output in Nigeria. Both tests indicate that distribution capability is
statistically significantly related to industrial output and is an important contribution to
policy development for improved electricity delivery for enhanced industrial output in
Nigeria.
Using a systems approach in this study provided invaluable insights into
relationships and interdependencies between the elements of the power supply value
chain in Nigeria. The study results indicate capacity misalignment that can be due to
infrastructure deficits at the power transfer interface points within the power supply value
chain or constraints within the electricity distribution systems due to either infrastructure
deficits or poor services, or both. These capacity misalignments provide an opportunity
for future studies to identify underlying factors that lead to the deficits and the type of
relationship or correlation between these factors to support policy development.
96
The result of this study applies to Nigeria, but there is an opportunity for future
studies to extend this type of study to countries within the subregion (West Africa) and
the rest of SSA countries. The extension of this study to other countries like Nigeria will
test the generalizability of the findings to other electricity-challenged countries in the
world. Extending this study can potentially influence research approaches in the
electricity and economic development literature, especially in electricity-challenged
countries.
In the study, idle generation was identified as a fallout of capacity misalignment
in the Nigerian power supply value chain. This finding is revealing in the context of the
fact that Nigeria has the largest electricity gap in the world from the World Bank 2021
energy progress report. There is an opportunity arising from this study to examine the
relationship between the rate of electricity access in Nigeria and available distribution
capability and other sources of energy.
Seventy-one percent of Nigerian industries generate their power, further
acknowledging the poor state electricity supply in Nigeria (Osakwe, 2017). In this regard,
there is an opportunity for future research to examine the correlation or effect of self-
generation on industrial output in Nigeria. Such a study will support policy development,
especially regarding improving services and removing power distribution constraints to
increase industrial output in Nigeria.
Gas was identified as a possible covariate in this study but was not one of the
variables examined. There was no single aggregator for all sources of gas for power
(private and government) in Nigeria. Given the high proportion of thermal power plants
97
that require gas as fuel and the reported constraints to available power generation due to
gas shortages, further studies may examine likely covariates that constrain gas supply for
power in Nigeria. When gas supply shortages are factored in, this type of research can
facilitate the determination of the extent of idle power generation capacity. A study of
this nature could support policy development, especially policy for infrastructure
development for power, gas supply, and improved services.
Implications for Social Change
The objective of this study was to present electricity supply as a system and
provide a better understanding of the interrelatedness between the elements that make up
the electricity supply value chain in Nigeria and how they impact industrial output. As an
energy source, electricity is an important input in the industrial sector because of its ease
of conversion (Stern et al., 2019). Poor-quality power supply adversely impacts
Nigeria’s industrial output, leading to high reliance on self-electricity generation and
increased cost of production, which renders local industries less competitive with foreign
products (Chete et al., 2014). Poor electricity access, low-quality power supply, and high
electricity cost were the three main reasons for low capacity utilization, lack of
competitiveness, and lack of growth in Nigeria’s industrial sector (Osakwe, 2017).
The Pearson’s correlation and regression analysis results indicate that distribution
capability is statistically significantly related to industrial output and may provide support
for a coordinated electricity power development policy in Nigeria. Improved electricity
supply will positively impact the industrial sector by reducing or eliminating the high
cost of self-generation that will boost the competitiveness of local industries. Such
98
improvements can catalyze enhanced production output, which would lead to economic
growth and eradication of poverty through increased economic activities that lead to
gainful employment and positive social change. Improvements in electricity supply
would facilitate electricity access growth in Nigeria, positively impacting the education
and health sectors, including other public services. In this regard, it is conceivable that
electricity affects people’s lives in multiple ways. This study could provide the impetus
for improvements in electricity supply in countries like Nigeria, leading to positive social
change in many ways.
This study presented electricity supply as a system, which allowed the
identification of the elements that make up the electricity supply value chain. The result
of this study enabled the identification of capacity mismatch along the electricity supply
value chain. This type of mismatch could be either due to infrastructure deficits or poor
services or both and may be responsible for the poor state of electricity supply in Nigeria.
This study enabled the identification of electricity consumption, a common independent
variable in the electricity and economic development literature, as a constrained variable
and consequently could invalidate or put to the test the validity of such studies, especially
in electricity-challenged countries like Nigeria. By adopting systems theoretical
perspective, future studies in the electricity and economic development literature would
deepen understanding regarding the relationship between electricity and economic
development and support coherent policy development.
99
Conclusions
The purpose of this correlational study, anchored on a systems approach, was to
examine the relationship between the power delivery subsystems and their effect on
industrial output in Nigeria. The literature on electricity and economic development is
diverse. Multiple studies in the same country yielded different results that make
developing a coherent policy for improved power supply and economic growth difficult.
This study indicated that electricity consumption is a constrained variable and may
invalidate studies where it has been used as an independent variable in electricity-
constrained countries like Nigeria. This study also indicated the possibility of idle
generation capacity amidst reports on the poor state of electricity supply in Nigeria.
Using the systems theoretical approach in this study contributes to the electricity
and economic development literature by identifying and utilizing elements that make up
the electricity delivery value chain as predictor variables to study the effects of electricity
on industrial output in Nigeria. The result indicated capacity incongruencies across the
power supply value chain. Pearson’s correlation analysis showed no statistically
significant correlation between generation capacity and transmission capacity, and there
was a negative but no statistically significant correlation between transmission capacity
and distribution capability. The capacity mismatch could be due to infrastructure deficits
at the interface points across the supply value chain or due to poor services, or both.
The results of the multiple regression analysis indicated that only distribution
capability statistically significantly predicted industrial output in the model. This study
supports coherent policy development by bringing clarity to the workings of the
100
electricity supply systems and how electricity impacts industrial growth in countries like
Nigeria. Both Pearson’s correlation and regression analysis indicate that distribution
capability is statistically significantly related to industrial output and provides policy
direction for improved electricity delivery. The three main outputs from this study are (a)
the likelihood of the existence of idle generation even amidst scarcity, (b) that electricity
consumption is a constrained variable and may invalidate studies where electricity has
been used as an independent variable in electricity challenged countries like Nigeria, and
(c) the existence of capacity incongruencies across the power supply value chain that
indicate poor planning. All these findings have implications for literature and policy
development.
101
References
Abokyi, E., Appiah-Konadu, P., Sikayena, I., & Oteng-Abayie, E. (2018). Consumption
of electricity and industrial growth in the case of Ghana. Journal of Energy, 2018,
1–11. https://doi.org/10.1155/2018/8924835
Acaravci, A. (2010). The causal relationship between electricity consumption and GDP in
Turkey: Evidence from ARDL bounds testing approach. Economic Research-
Ekonomska Istraživanja, 23(2), 34–43.
https://doi.org/10.1080/1331677X.2010.11517410
Adams, K. M., Hester, P. T., Bradley, J. M., Meyers, T. J., & Keating, C. B. (2014).
Systems theory as the foundation for understanding systems. Systems
Engineering, 17(1), 112–123. https://doi.org/10.1002/sys.21255
Akinwale, y. A., & Muzindutsi, P.-F. (2019). Electricity consumption, trade openness and
economic growth in South Africa: An ARDL approach. Journal of Economic
Cooperation & Development, 40(1), 135–155.
https://www.sesric.org/publications-jecd.php
Ali, S., Zhang, J., Aamir, A., & Mahmood, A. (2020). Impact of electricity consumption
on economic growth: An application of vector error correction model and
Artificial Neural Networks. Journal of Developing Areas, 54(4), 89–104.
https://muse.jhu.edu/article/738663
Alley, I., Egbetunde, T., & Blessing, O. (2016). Electricity supply, industrialization and
economic growth; evidence from Nigeria. International Journal of Energy Sector
Management, 10(4), 511–525. https://doi.org/10.1108/IJESM-10-2015-0005
102
Amin, S. B., & Murshed, M. (2017). An empirical analysis of multivariate causality
between electricity consumption, economic growth and foreign aid: Evidence
from Bangladesh. Journal of Developing Areas, 51(2), 369–380.
https://muse.jhu.edu/article/657948
Amissah, M., Gannon, T., & Monat, J. (2020). What is systems thinking? [Conference
session]. Expert perspectives from the WPI systems thinking colloquium of 2
October 2019. Systems, 8(1), 1–26. https://doi.org/10.3390/systems8010006
Arnold, R. D., & Wade, J. P. (2015). A definition of systems thinking: A systems
approach. Procedia Computer Science, 44, 669–678.
https://doi.org/10.1016/j.procs.2015.03.050
Asamoah, M. K. (2014). Re-examination of the limitations associated with correlational
research. Journal of Educational Research and Reviews, 2(4), 45–52.
http://www.sciencewebpublishing.net/jerr/archive/2014/July/Abstract/Asamoah.ht
m
Azolibe, C. B., & Okonkwo, J. J. (2020). Infrastructure development and industrial sector
productivity in Sub-Saharan Africa. Journal of Economics and Development,
22(1), 91–109. https://doi.org/10.1108/JED-11-2019-0062
Baloch, M., Danish, P., & Meng, F. (2019). Modeling the non-linear relationship between
financial development and energy consumption: statistical experience from
OECD countries. Environmental science and pollution research international,
26(9), 8838–8846. https://doi.org/10.1007/s11356-019-04317-9
Berry, W., & Feldman, S. (1985). Multiple regression in practice. Sage.
103
Boardman, j., Sauser, B., John, L., & Edson, R. (2009). The conceptagon: A framework
for systems thinking and systems practice. [Conference session]. IEEE
International Conference on Systems, Man, and Cybernetics. San Antonio, TX.
https://doi.org/10.1109/ICSMC.2009.5346211
Bonnema, G. M., & Broenink, J. F. (2016). Thinking tracks for multidisciplinary system
design. Systems, 4(4), 1–17. https://doi.org/10.3390/systems4040036
Busk, P. L. (Ed). (2010). International Encyclopedia of Education (3rd ed.).
https://doi.org/10.1016/B978-0-08-044894-7.01690-0
Cabrera, D., Cabrera, L., & Powers, E. (2015). A unifying theory of systems thinking
with psychosocial applications. Systems Research and Behavioral Science, 32(5),
534–545. https://doi.org/10.1002/sres.2351
Castelle, K. M., & Jaradat, R. M. (2016). Development of an instrument to assess
capacity for systems thinking. Procedia Computer Science, 95, 80–86.
https://doi.org/10.1016/j.procs.2016.09.296
Caws, P. (2015). General systems theory: Its past and potential. Systems Research and
Behavioral Science, 32(5), 524–521. https://doi.org/10.1002/sres.2353
Chakamera, C., & Alagidede, P. (2018). The nexus between infrastructure (quantity and
quality) and economic growth in Sub Saharan Africa. International Review of
Applied Economics, 32(5), 641–672.
https://doi.org/10.1080/02692171.2017.1355356
104
Chan, W. T. (2015). The role of systems thinking in systems engineering, design and
management. Civil Engineering Dimension, 17(3), 126–132.
https://doi.org/10.9744/ced.17.3.126-132
Chete, L., Adeoti, J, Adeyinka, F., & Ogundele, O. (2014). Industrial development and
growth in Nigeria: Lessons and challenges [Paper presentation]. WIDER
Working Paper 2014/019, UNU-WIDER, Helsinki, Finland.
Chol, K. J. (2020). The relationship between economic growth, energy consumption and
environmental pollution based on ARDL model. [Conference session].
Proceedings of the 5th International Conference on Social Sciences and Economic
Development, Xian, China. https://doi.org/10.2991/assehr.k.200331.011
Clancy, T. (2018). Systems thinking: Three system archetypes every manager should
know. IEEE Engineering Management Review, 46(2,) 32–41.
https://doi.org/10.1109/EMR.2018.2844377
Creswell, J. w. (2009). Research design: Qualitative, quantitative, and mixed methods
approaches. Sage.
Demetis, D. S., & Lee, A. S. (2017). Taking the first step with systems theorizing in
information systems: A response. Information and Organization, 27(3), 163–170.
https://doi.org/10.1016/j.infoandorg.2017.06.003
Drew, C., Hardman, M., & Hosp, J. (2008). Designing and conducting research in
education. Sage. https://doi.org/10.4135/9781483385648
Edmonds, W. A., & Kennedy, T. D. (2017). An applied guide to research designs:
Quantitative, qualitative, and mixed methods. Sage.
105
Frankfort-Nachmias, C., & Leon-Guerrero, A. (2015). Social statistics for a diverse
society. Sage
García-Pérez, M. A. (2012). Statistical conclusion validity: Some common threats and
simple remedies. Frontiers in Psychology, 3(325), 1–11.
https://doi.org/10.3389/fpsyg.2012.00325
Given, L. (Ed). (2008). The Sage encyclopedia of qualitative research methods. Sage.
Green, W. (2012). Econometric analysis. Prentice Hall.
Greenstein, T. (2006). Methods of family research. Sage.
https://doi.org/10.4135/9781412990233
Grohs, J. R., Kirk, G. R., Soledad, M. M., & Knight, D. B. (2018). Assessing systems
thinking: A tool to measure complex reasoning through ill-structured problems.
Thinking Skills and Creativity, 28, 110–130.
https://doi.org/10.1016/j.tsc.2018.03.003
Guan, X., Zhou, M., & Zhang, M. (2015). Using the ARDL-ECM approach to explore
the nexus among urbanization, energy consumption, and economic growth in
Jiangsu Province, China. Emerging Markets Finance and Trade, 51(2), 391–399.
https://doi.org/10.1080/1540496X.2015.1016840
Gujarati, D. N., & Porter, D. C. (2009). Basic econometrics. McGraw-Hill/Irwin.
Ha, J., Tan, P.-P., & Goh, K.-L. (2018). Linear and nonlinear causal relationship between
energy consumption and economic growth in China: New evidence based on
wavelet analysis. PLOS ONE Open Access Journal, 13(5), 1–21.
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0197785
106
Harris, G., & Caudle, L. (2019). A systems approach to establishing an advanced
manufacturing innovation institute. System, 7(3), 41.
https://doi.org/10.3390/systems7030041
Hasan, A., Zaman, A., Skider, Z. I., & Wadud, A. (2018). The dynamics of electricity
consumption, energy use and GDP in Bangladesh. Romanian Economic Journal,
20(65), 34–49. http://www.rejournal.eu/node/3512
Hayes, A., & Cai, L. (2007). Using heteroskedasticity-consistent standard error
estimators in OLS regression: An introduction and software
implementation. Behavior Research Methods 39 (4), 709–722.
https://doi.org/10.3758/BF03192961
Hirsh, R. F., & Koomey, J. G. (2015). Electricity consumption and economic growth: A
new relationship with significant consequences? The Electricity Journal, 28(9),
72–84. https://doi.org/10.1016/j.tej.2015.10.002
Istaiteyeh, R. (2016). Electricity consumption and real GDP: Are they causal for Jordan.
International Journal of Economic Perspectives, 10(4), 526–540.
Jiranyakul, K. (2016). Causal linkages between electricity consumption and GDP in
Thailand: evidence from the bound test. Economics Bulletin, 36(2), 921–930.
https://ideas.repec.org/a/ebl/ecbull/eb-14-01010.html
Kodongo, O., & Ojah, K. (2016). Does infrastructure really explain economic growth in
Sub-Saharan Africa? Review of Development Finance, 6(2), 105–125.
https://doi.org/10.1016/j.rdf.2016.12.001
107
Kordova, S., Frank, M., & Miller, A. (2018). Systems thinking education—seeing the
forest through the trees. Systems, 6(3), 1–14.
https://doi.org/10.3390/systems6030029
Kunda, D., & Chisimba, M. (2017). Assessment of the effects of electricity consumption
on the economy using Granger causality: Zambian case. Database Systems
Journal, 8(2), 44–56. https://www.dbjournal.ro/archive/28/28_5.pdf
Kwakwa, P. A. (2017). Electricity consumption in Egypt: A long-run analysis of its
determinants. OPEC Energy Review, 41(3), 3-22.
https://doi.org/10.1111/opec.12091
Laerd Statistics. (2018). Linear regression. https://statistics.laerd.com/spss-
tutorials/linear-regression-using-spss-statistics.php
Laszlo, E. (1996). The systems view of the world: A holistic vision for our time. Hampton
Press.
Lu, W. (2017). Electricity consumption and economic growth: Evidence from 17
Taiwanese industries. Sustainability, 9(1), 1–15.
https://doi.org/10.3390/su9010050
Mawejje, J., & Mawejje, D. (2016). Electricity consumption and sectoral output in
Uganda: an empirical investigation. Journal of Economic Structures, 5(21), 1–16.
https://doi.org/10.1186/s40008-016-0053-8
Moldavska, A., & Welo, T. (2016). Development of manufacturing sustainability
assessment using systems thinking. Sustainability, 8(5), 1–26.
https://doi.org/10.3390/su8010005
108
Monat, J., Amissah, M., & Gannon, T. (2020). Practical applications of systems thinking
to business. Systems, 8(2), 14. https://doi.org/10.3390/systems8020014
Monat, J. P., & Gannon, T. F. (2015). What is systems thinking? A review of selected
literature plus recommendations. American Journal of Systems Science, 4(1), 11–
26. http://article.sapub.org/10.5923.j.ajss.20150401.02.html
Monat, J. P., & Gannon, T. F. (2018). Applying systems thinking to engineering and
design. Systems, 6(3), 1–20. http://dx.doi.org/10.3390/systems6030034
Mononen, L. (2017). Systems thinking and its contribution to understanding future
designer thinking. The Design Journal, 20(1), S4529–S4538.
https://doi.org/10.1080/14606925.2017.1352949
Mutingi, M., Mbohwa, C., & Dube, P. (2017). System dynamics archetypes for capacity
management of energy systems. Energy Procedia, 141, 199–205.
https://doi.org/10.1016/j.egypro.2017.11.038
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
Nwankwo, O. C., & Njogo, B. O. (2013). The effect of electricity supply on industrial
production within the Nigerian economy (1970-2010). Journal of Energy
Technologies and Policy, 3(4), 34–42.
https://www.iiste.org/Journals/index.php/JETP/article/view/5407
109
Odeleye, A., & Olunkwa, N. (2019). Industrialization: Panacea for economic growth.
Academic Journal of Economic Studies, 5(2), 45–51.
https://ideas.repec.org/a/khe/scajes/v5y2019i2p45-51.html
Ogundipe, A., Akinyemi, O., & Ogundipe, O. (2016). Electricity consumption and
economic development in Nigeria. International Journal of Energy Economics
and Policy, 6(1), 134–143. https://econjournals.com/index.php
Ogundipe, A., & Apata, A. (2013). Electricity consumption and economic growth in
Nigeria. Journal of Business Management and Applied Economics, 2(4), 1–14.
http://eprints.covenantuniversity.edu.ng/id/eprint/1777
Ologundudu, M. (2015). The impact of electricity supply on industrial and economic
performance in Nigeria. International Global Journal of Management and
Business research, 2(4), 9–28.
http://mtu.edu.ng/mtu/oer/JounalAcrticle/CHMS/THE%20IMPACT%20OF%20E
LETRICITY.pdf
Önday, Ö. (2018). The relationship between concepts of rational, natural and open
systems: Managing organizations today. International Journal of Information,
Business and Management, 10(1), 247–258.
https://ijibm.elitehall.com/index4.htm
Osakwe, P. (2017). Unlocking the potential of the power sector for industrialization and
poverty alleviation in Nigeria. Paper presented at the 2017 United Nation
conference on trade and development.
https://unctad.org/en/pages/PublicationWebflyer.aspx?publicationid=1854
110
Oshry, B. (2007). Seeing systems: Unlocking the mysteries of organizational life. Berrett-
Koehler Publishers
Owusu-Manu, D., Jehuri, A. B., Edwards, D. J., Boateng, F., & Asumadu, G. (2019). The
impact of infrastructure development on economic growth in sub-Saharan Africa
with special focus on Ghana. Journal of Financial Management of Property and
Construction, 24(7), 253–273. https://doi.org/10.1108/JFMPC-09-2018-0050
Passmore, W. A. (1988). Designing effective organizations: The sociotechnical systems
perspective. Wiley.
Rashid, A., & Yousaf, N. (2016). Linkage of financial development with electricity-
growth nexus of India and Pakistan. EuroEconomica, 34(2), 151–160.
http://journals.univ-danubius.ro/index.php/euroeconomica/article/view/3037
Reinard, J. (2011). Communication research statistics. Sage.
Ridzuan, A., Kamaluddin, M., Ismail, N., Razak, M., & Haron, N. (2020).
Macroeconomic indicators for electrical consumption demand model in Malaysia.
International Journal of Energy Economics and Policy, 10(1), 16–22.
https://doi.org/10.32479/ijeep.8139
Rousseau, D. (2015). General systems theory: Its present and potential. Systems Research
and Behavioral Science, 32(5), 522–533. https://doi.org/10.1002/sres.2354
Rousseau, D. (2017). Systems research and the quest for scientific systems principles.
Systems, 5(2), 1–16. https://doi.org/10.3390/systems5020025
111
Saghir, J., & Santoro, J. (2018). Urbanization in sub-Saharan Africa: Meeting challenges
by bridging stakeholders. Center for Strategic and International Studies.
https://www.csis.org/analysis/urbanization-sub-saharan-africa
Salahudin, M., & Alam, K. (2016). Information and communication technology,
electricity consumption and economic growth in OECD countries: A panel data
analysis. International Journal of Electrical Power & Energy Systems, 76, 185–
193. https://doi.org/10.1016/j.ijepes.2015.11.005
Samu, R., Bekun, F., & Fahrioglu, M. (2019). Electricity consumption and economic
growth nexus in Zimbabwe revisited: Fresh evidence from Maki cointegration.
International Journal of Green Energy, 16(7), 540–550.
https://doi.org/10.1080/15435075.2019.1598417
Sankaran, A., Sanjay, K., Arjun, K., & Mousumi, D. (2019). Estimating the causal
relationship between electricity consumption and industrial output: ARDL bounds
and Toda-Yamamoto approaches for ten late industrialized countries. Heliyon,
5(6), 1–9. https://doi.org/10.1016/j.heliyon.2019.e01904
Scherbaum, C. A., & Shockley, K. M. (2015). Analyzing quantitative data: for business
and management students. Sage.
Sekantis, L., & Motlokoa, M. (2015). Evidence on the nexus between electricity
consumption and economic growth through empirical investigation of Uganda.
Review of Economic and Business Studies, 8(1), 149–165.
https://doi.org/10.1515/rebs-2016-0020
112
Sekantis, L. P., & Timuno, S. (2017). Electricity consumption in Botswana: The role of
financial development, industrialization and urbanization. Review of Economic
and Business Studies, 10(1), 75–102. https://doi.org/10.1515/rebs-2017-0049
Serrano, R., Rodrigues, L. H., Lacerda, D. P., & Paraboni, P. B. (2018). Systems thinking
and scenario planning: Application in the clothing sector. Systemic Practice &
Action Research, 5, 509–537. https://doi.org/10.1007/s11213-017-9438-3
Shrestha, M. B., & Bhatta, G. R. (2018). Selecting appropriate methodological
framework for time series data analysis. The Journal of Finance and Data
Science, 4(2), 71–89. https://doi.org/10.1016/j.jfds.2017.11.001
Sibo-Ingrid, H., Gomez, C.-D. A., & Liou, S. (2018). Systems thinking: a paradigm for
advancing technology to enhance humanity. [Conference session]. 2018
International Conference on Orange Technologies (ICOT), Nusa Dua, Bali,
Indonesia. https://doi.org/10.1109/ICOT.2018.8705910
Stern, D., Burke, P., & Bruns, S. (2019). The impact of electricity on economic
development: A macroeconomic perspective. Energy and Economic Growth.
https://energyeconomicgrowth.org/content/publications
Stroh, D. P. (2015). Systems thinking for social change: A practical guide to solving
complex problems, avoiding unintended consequences, and achieving lasting
results. Chelsea Green Publishing.
Ubi, S., & Effiom, L. (2013). The dynamic analysis of electricity supply and economic
development: Lessons from Nigeria. Journal of Sustainable Society, 2(1), 1–11.
https://doi.org/10.11634/216825851302163
113
Ugwoke, T., Dike, C., & Elekwa, P. (2016). Electricity consumption and industrial
production in Nigeria. Journal of Policy and Development Studies, 10(2), 8–19.
http://www.arabianjbmr.com/JPDS_index.php
Vemuri, P., & Bellinger, G. (2017). Examining the use of systemic approach for adoption
of systems thinking in organizations. Systems, 5(3), 43–52.
https://doi.org/10.3390/systems5030043
Verhoeff, R. P., Knippels, M.-C. P., Gilissen, M. G., & Boersma, K. T. (2018). The
theoretical nature of systems thinking: Perspectives on systems thinking in
biology education. Frontiers in Education, 3(40), 1–11.
https://doi.org/10.3389/feduc.2018.00040
von Bertalanffy, L. (1968). General systems theory: Foundations, development,
applications. Braziller.
Walden University. (n.d.). Research ethics review process by IRB.
https://academicguides.waldenu.edu/research-center/research-ethics/review-
process
Weissenberger-Eibl, M. A., Almeida, A., & Seus, F. (2019). A systems thinking approach
to corporate strategy development. Systems, 7(1), 1–16.
https://doi.org/10.3390/systems7010016
White, S. M. (2015). Systems theory, systems thinking. [Conference session]. 2015
Annual IEEE Systems Conference (SysCon) Proceedings, Vancouver, BC,
Canada. https://doi.org/10.1109/SYSCON.2015.7116787
114
Whitehead, N. P., Scherer, W. T., & Smith, M. C. (2015). Systems thinking about
systems thinking: A proposal for a common language. IEEE Systems Journal,
9(4), 1117–1128. https://doi.org/10.1109/JSYST.2014.2332494
Wooldridge, J. M. (2013). Introductory econometrics: A modern approach. Cengage
Learning.
Zare, F., Bagheri, A., & Elsawah, S. (2017). Using system archetypes for problem
framing and a qualitative analysis: A case study in Iranian water resource
management. [Conference session]. 22nd International Congress on Modelling
and Simulation. Hobart, Tasmania, Australia. https://mssanz.org.au/modsim2017/
Zhang, J., & Broadstock, D. (2016). The causality between energy consumption and
economic growth for China in a time-varying framework. The Energy Journal,
37, 29–53. https://doi.org/10.5547/01956574.37.SI1.jzha
Zhang, T., Shi, X., Zhang, D., & Xiao, J. (2019). Socio-economic development and
electricity access in developing economies: A long-run model averaging
approach. Energy Policy, 132, 223–231.
https://doi.org:10.1016/j.enpol.2019.05.031
