Content uploaded by Mwoya Byaro
Author content
All content in this area was uploaded by Mwoya Byaro on Feb 20, 2024
Content may be subject to copyright.
Journal Pre-proof
Tackling energy poverty: Do clean fuels for cooking and access to
electricity improve or worsen health outcomes in sub-Saharan Africa?
Mwoya Byaro , Nanzia Florent Mmbaga , Gemma Mafwolo
PII: S2772-655X(24)00003-X
DOI: https://doi.org/10.1016/j.wds.2024.100125
Reference: WDS 100125
To appear in: World Development Sustainability
Received date: 5 July 2023
Revised date: 14 November 2023
Accepted date: 18 February 2024
Please cite this article as: Mwoya Byaro , Nanzia Florent Mmbaga , Gemma Mafwolo , Tack-
ling energy poverty: Do clean fuels for cooking and access to electricity improve or
worsen health outcomes in sub-Saharan Africa?, World Development Sustainability (2024), doi:
https://doi.org/10.1016/j.wds.2024.100125
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©2024 Published by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
1
Tackling energy poverty: Do clean fuels for cooking and access to
electricity improve or worsen health outcomes in sub-Saharan Africa?
Mwoya Byaro1, Nanzia Florent Mmbaga2, Gemma Mafwolo1
1, Institute of Rural Development Planning, P.O BOX 11957, Mwanza, Tanzania
2Local Government Training Institute, P.O BOX 1125, Dodoma, Tanzania
Corresponding author email: mwoyabyaro2018@gmail.com
Abstract
Sub-Saharan Africa (SSA) countries continue to suffer from energy poverty, with 35% and
19% of the average population, having access to electricity and clean fuels for cooking
technologies, respectively.
This study examines whether access to electricity and clean
fuels for cooking and technology improves or worsens health outcomes (i.e. infant,
child and maternal mortality) in 48 sub-Saharan African countries from 2000 to 2020.
We applied the panel quantile regression to estimate the impact of access to electricity, and
clean fuels for cooking on health outcomes while controlling for health care expenditure
and income, using lagged explanatory variables as instruments to eliminate endogeneity.
We also applied the Kernel-based Regularized Least Squares (KRLS), a machine learning
techniques for robustness of the results.
Our results show that access to electricity reduce
infant, child, and maternal mortality to all quantiles (i.e., the 25th, 50th, 60t h, 75th
and 90th).
Similarly, clean fuels for cooking and technologies reduce maternal, infant and
child mortality to most quantiles.
This means that increased access to electricity, clean
fuels for cooking and technologies will have a significant impact on reducing child, infant
and maternal mortality in SSA
. The findings also reveal that clean fuels for cooking
and technologies increase both infant and under-five mortality in some quantiles.
This is likely because cooking is also the leading cause of house fires, killing both
infants and children under the age of five. Therefore, home cooking safety is also
important to prevent unnecessary deaths of infants and children.
Our study suggests
short-and long-term energy policies to end energy poverty and ultimately improve
population health in the SSA.
2
Keywords: Energy poverty, electricity access, clean fuels for cooking, health outcomes
1. Introduction
Energy poverty is one of the most serious energy problems in Africa [1]. It takes many
forms and is usually defined as a situation where there is little or no access to
electricity [1]. It also refers to households being unable to meet their basic domestic
energy needs (i.e. cooking, lighting), a problem that affects both developed and
developing countries to varying degrees [2]. Access to a reliable electricity supply is
an important prerequisite for social and economic growth, especially in developing
countries like Africa, which is important to achieve the Sustainable Development
Goals (SDG 7) by 2030. SDG 7 calls for universal access to modern, sustainable,
accessible and economical energy for all by 2030 [3]. Using household appliances
such as radios, televisions, refrigerators and cell phones, and turning on lights at night
requires access to electricity. This suggests that improving access to electricity helps
alleviate energy poverty, which is a critical component of international policy
interventions [1].
Despite the fact that electricity is used profitably for a variety of activities in African
households, the main challenge remains that electricity prices are high and cannot
be afforded by low-income people [3]. As a result, low-income people still use non-
renewable energy (e.g., wood, charcoal) for cooking [4], [5]. Fossil fuel pollution,
primarily caused by cooking energy (i.e. charcoal, wood and other local fuels), causes
chronic respiratory and eye diseases that are responsible for more than 1.5 million
deaths per year in both mothers and children [6], [7].
This is because many women in
developing countries like Africa continue to use kerosene, wood, animal manure and
charcoal, which has led to numerous infant deaths due to the lack of clean fuels and
cooking technologies. This implies that inaccessibility to electricity and continued use of
wood, charcoal and other conventional fuels make it difficult for SSA to meet the 2030
3
Sustainable Development Goals.
In 2020, 69% of the world’s population had access to clean fuels and technologies for
cooking [8]. Globally, access to clean fuels and cooking technology has increased by
just 12% over the past decade, and if this trend continues, people in low-and middle-
income countries like SSA, in particular, will not have access to clean fuels by 2030
[8]. Furthermore, the energy report showed that among the 20 countries with a high
deficit in access to clean fuels and cooking technologies between 2016 and 2020, 10
countries are represented in Africa, including Tanzania (4%), Democratic Congo (4%)
and Uganda (1%), Madagascar (1%), Mozambique (5%), Ghana (11%),
Niger (3%),
Nigeria (11%), Kenya (16%) and Ethiopia (7%). On the other
hand, evidence has
shown that SSA countries lack access to clean fuels and technologies for cooking,
indicating fuel poverty [5]. While
trying to alleviate energy poverty in SSA by
improving access to electricity, clean fuels and cooking technology, it is also linked
to improving people’s well-being and ensuring their good health [27]. Meanwhile,
lack of clean fuels and cooking technologies and access to electricity, when combined
with inadequate welfare and health systems, can pose challenges
by putting people at
greater health risk [27]. In this situation, understanding the health impact of access
to electricity, clean fuels for cooking and technolo
gies are important to help
policymakers propose appropriate health promotion
interventions in the African
context.
Numerous studies have shown that access to electricity can be detrimental to health.
For example, the production of industrial goods can lead to air pollution, which can
affect the health of mothers and their children (see, for example, [9] [10] [11] [12] [13]
[14]. It is important to note that three main pillars are needed to alleviate energy poverty
in low-income households such as in sub-Saharan Africa (SSA), including improved
access to electricity, the use of safer and cleaner cooking fuels, and the use of renewable
energy sources [15].
Basically, the expansion and accessibility of electricity leads to the availability of
food, medicine, clean and safe drinking water that can be stored in refrigerators, which
4
could reduce the risk of maternal death, infant death, and other health complications
[6] [16] [10] [17] [18]. Similarly, access to electricity improves the ability of
healthcare organizations to provide quality services through the use of modern medical
equipment such as diagnostic equipment, and improve hygiene, safety, and
refrigeration of medicines and vaccines, which has a positive impact on mortality rates
in healthcare and promotes positive health outcomes [19]. Despite all these health
benefits of access to electricity and clean fuels for cooking, Africa still suffers from
severe energy poverty.
The motivation for this study was driven by five key factors: First, the majority of
people in sub-Saharan Africa (SSA) do not have access to electricity (65%),
contributing to the region’s energy poverty. Only 35% of the population in the region
has access to electricity on average [3]. Second, compared to developed regions, SSA
has higher maternal, infant, and child mortality rates, that may have been caused by
energy poverty [20], [21]. Third, more than half of the world’s poorest people live in
sub-Saharan Africa, where high poverty rates are closely linked to poor health and
access to electricity helps reduce poverty [22]. Fourth, sub-Saharan Africa is no
exception to the fact that the majority of its population relies heavily on traditional
cooking techniques using wood, charcoal and cow dung. Fifth, SSA has the least
access to clean fuels and cooking technologies [5]. Finally, previous studies
attempting to link energy poverty and health outcomes have focused primarily on
developed countries [23], [24], [25], [26]
and have failed to include both access to
electricity and clean fuels for cooking and technology when examining health outcomes.
In sub-Saharan Africa, few studies, such as Shobande [27] have used all-cause
mortality to explain the relationship between access to electricity and health outcomes.
Our study focuses on broader health indicators such as infant, child and maternal
mortality. Although previous literature has shown an association between under-five
and maternal mortality rates and electricity consumption [28], this study focuses on
tackling energy poverty, particularly through the use of clean fuels for cooking and
technologies, and access to electricity in relation to infant, child and maternal mortality
5
in 48 selected sub-Saharan African countries from 2000 to 2020.
Our study contributes to the literature in many ways. First, understanding the health impacts
of clean fuels for cooking and access to electricity will help policymakers in SSA promote
and prioritize the deployment of modern, clean fuels for cooking and technologies in Africa
and ultimately reduce energy poverty. Second, it fills a gap in the existing literature by
employing Powell's [29] generalized panel quantile regression and a novel machine
learning technique known as Kernel-based Regularized Least Squares (KRLS) for the
robustness of the results. Third, this is the first study in the African context to use both
quantile regression and kernel-based regularized least squares (krls) to analyze the
relationship between access to electricity, clean cooking fuels, and health outcomes (i.e.
maternal, infant and under-five mortality) for updated data from 2000 to 2020.
In other
words, no previous studies in SSA have used generalized quantile regression and
kernel
-based regularized least squares
to examine the relationship between mortality,
clean cooking fuels, and access to electricity. Several studies used mean regression
estimates to analyze data for specific countries or Africa in general [27] [12] [9] [2]
[10] [14].
Fourth, we contribute methodologically to the current study compared to the
existing literature on access to electricity, clean fuels for cooking and health outcomes
nexus.
The main benefit of using generalized quantile regression is that it is robust to outliers
and examines non-linearity between a set of dependent and independent variables
[30]. Although countries in SSA have different levels of indicators of energy poverty
and health outcomes (e.g., infant, child, and maternal mortality), applying quantile
regression is useful.
This means that quantile regression explore the heterogeneous effect
of independent variables on the distribution of dependent variables across different
quantiles
. Thus, using the generalized panel-quantile approach is helpful in overcoming
the shortcomings of previous mean-regression-based studies. Furthermore, we used the
lagged explanatory variables in generalized quantile regression as instruments to
address endogeneity issues in the studied variables [29], [30].
Overall, our study aims to answer two questions:
does access to electricity and clean
6
fuels for cooking improve or worsen health outcomes in sub-Saharan Africa (i.e., infant,
child, and maternal mortality)? Are there asymmetric (i.e. positive and negative) impacts of
access to electricity and clean fuels for cooking on health outcomes (i.e., infant, child, and
maternal mortality) in SSA from 2000 to 2020?
Our preliminary findings show that access to electricity and clean fuels for cooking and
technology reduced maternal, infant, and under-five mortality in sub-Saharan Africa from
2000 to 2020. Moreover, there are asymmetric (i.e., positive and negative) impacts of clean
fuels for cooking and technology on both infant and under-five mortality.
The paper is divided into four parts. Section 2 reviews the literature on the link between
energy poverty and mortality. The third section shows data sources and methodology.
Section 4 presents the findings and discussion, while Section 5 provides the conclusion,
policy recommendations, limitations and future research directions.
2. Literature review
2.1 Theoretical literature
There is a paucity of theoretical literature on the relationship between energy poverty
and health [27]. The definition of fuel poverty is still debated among scholars, and no
consensus has been reached [31]. The inability of households to meet their basic energy
needs (i.e., cooking, heating, and lighting) and to use modern energy services is what
economists refer to as "energy poverty"[32], [25]. Furthermore, energy poverty is
largely caused by lack of access to electricity and clean fuels for cooking and is
associated with economic underdevelopment [33]. Therefore, the theoretical literatures
on energy poverty and health outcomes are ad-hoc and rely more on empirical
literature.
On the other hand, the theoretical link between energy consumption and health is well
explained by the Gary production theory. In this theory, health is considered as an output
that enters the utility function and, on the other hand, as one of the inputs into the
production functions [58]. Likewise, the theoretical model used in our study follows
7
Grossman's [59] demand for health care, which viewed health as an initial capital asset that
loses value over time and can lead to death. Therefore, health can be improved over time
and human life can be extended for future well-being, which also depends on the
consumption of goods such as energy consumption and medicines.
2.2 Empirical review
The empirical literature on energy poverty and its health consequences world- wide is
extensive. Table 1 summarizes the existing literature that has been done on the
relationship between energy poverty and health outcomes in both developed and
developing countries. The extensive literature indicates a significant connection
between energy poverty and under-five mortality. These include
contributions by [34],
[10], [35], [36], [37], [38], and [39] to examine energy poverty in the light of variables such
as mortality, health capital, under-five mortality and crude death.
Some literature shows
that rural areas have poor electrification, which is reported to be among the main
causes of the higher mortality rate in rural areas. For example, rural areas reported
using unclean fuels as the main energy source for cooking, lighting, and heating ([40]
[41]). Likewise, other literature such as [42], [43], [44], [17], [18], [45], has shown that
limited energy access causes diseases and high mortality.
Overall, Table 1 shows a paucity of empirical literature on the relationship between energy
poverty (i.e., a proxy for lack of access to electricity, clean fuels use, and modern cooking
technologies) and health outcomes in sub-Saharan Africa. The impact of energy poverty on
health outcomes in sub-Saharan Africa has not been examined using generalized quantile
regression techniques. To our knowledge, this is the first study to use generalized quantile
regression techniques to analyze the impact of clean fuels for cooking and technologies on
infant, under-five and maternal mortality in sub-Saharan Africa. The use of Kernel-based
Regularized Least Squares (KRLS) as a robustness check for results in an African context
also fills the gap in the previous studies on clean fuels, access to electricity and health
outcomes and contributes to the existing literature.
Table 1: A comparative review of this study and previous literature
Author/s
Study area and time
Variables
Methodology
Key findings
Shobande (2023)
29 Sub-Saharan
Africa,1990-2015
Mortality (crude death
rate per 1000 population),
electricity use, solid fuel,
indoor air pollution
Fixed effects
regressions, Mundlak
methodology
Access to electricity reduces the health
risks associated with burning solid fuels in
Africa.
Zhu et al. (2023)
China, 2000-2010
Under-five mortality,
solid fuel, GDP per
capita, rural population,
education
Fixed effects,
instrumental IV
variable Approach
A decrease in households solid fuels
reduces under-five mortality in China
Mohammed and
Akuoko (2022)
Ghana, 1993-2014
Electricity access, infant
mortality, income
distribution, birth
intervals
Pooled cross-
sectional estimates
Having access to electricity lowers the
risk of infant mortality.
Sule et al.(2022)
33 African countries,
2010-2017
Under-five mortality,
energy poverty
Panel cointegration
test, panel causality
test
Improving household access to energy
reduces under-five mortality.
Chen et al.
(2021)
Asia and Africa
Indoor air pollution,
childhood pneumonia.
Systematic review,
Meta-analysis
Switching to other processed solid fuels
that reduce indoor air pollution and
prevent pneumonia and morbidity and
mortality.
Li et al. (2022)
Rural China, 2014
and 2016
Clean energy for cooking,
age, education level,
women health
Propensity Score
Matching (PSM) and
difference in
difference method
Clean energy application improves the
health of rural women in China
Acheampong et
al. (2021)
South Asia, Sub-
Sahara Africa and
Caribbean-Latin
America-
1990-2018
Human development,
Human capital, Life
expectancy, Maternal
mortality, Under-five
mortality
two-stage least
squares approach
Access to energy is crucial for human
development but does not equally benefit
all the components of human
development.
Author/s
Study area and time
Variables
Methodology
Key findings
Banerjee et al.
(2021)
50 developing
countries, 1990-2017
Access to electricity,
infant mortality, life
expectancy, poverty head
count ratio, real GDP per
capita
Fixed effect, two
stage fixed effect
In developing countries, higher energy
consumption is associated with improved
health and education outcomes.
Nie et al. (2021)
China, 2012-2018
Life satisfaction,
depression index, solid
fuel measure,
Fixed effect, two
stage least square
Energy poverty raises the risk of
depression.
Ouedraogo &
Jiya (2021)
24 SSA
countries,1990-2018
Access to electricity,
health capital, incidence
of malaria, under-five
mortality rates, crude
death rates,
The ordinary least
squares and
instrumental
variables-two stage
square methods
Improving access to electricity tends to
reduce the under-five mortality rate as
well as the number of crude deaths in SSA
Rahman &Alam
(2021b)
SAARAC-ASEAN
countries (2002-
2018)
Electricity, female
education, public health
expenditure, health
outcomes
PCSE model, FGLS
model, and the pair-
wise Granger
causality test
Access to electricity has a positive effect
on women's life expectancy at birth and a
negative effect on adult mortality
Ayub et al.
(2021)b
Kenya
Cooking energy sources,
cleanliness cooking
system
efficiency cooking energy
Systematic review
Electric cooking has been found to be
clean and release no emissions during
cooking, significantly reducing both
indoor and environmental pollution
Anser et al.
(2020)
Asian economies
from 1995-2018
Mortality rate,
Greenhouse gas emission,
Respiratory diseases, Per
capita GDP.
System Generalized
Method of Moment
(GMM)
The emission of greenhouse gases is
responsible for the high mortality rate and
the occurrence of respiratory diseases in a
short period of time.
Irwin et al.
(2020)
Low and middle
income countries
Electricity access,
morbidity, mortality rate.
Systematic review
Electrification is associated with positive
health effects, including reduced
mortality, lower morbidity rates and
better access to and quality of care
Author/s
Study area and time
Variables
Methodology
Key findings
Shobande
(2020a)
Africa
Mortality rate, infant rate,
health expenditure,
GDP per capita, CO2
emissions
Grossman models
along with pooled
OLS and system
GMM estimators
Both infant and child mortality rates under
the age of 5 are negatively and
significantly related to energy use and
CO2 emissions in the region.
Shobande
(2020b).
Africa
Mortality, Electricity
consumption, GDP,
Improved water source
Vector
Autoregressive
(VAR), Granger-
causality model
Energy predictors have a negative and
significant impact on infant mortality
rates.
Wright et
al.(2020)
Global
Clean cooking.
Household air pollution.
Energy access Gender
equity, Improved cook
stoves, Renewable energy
Review, descriptive
statistics
Greenhouse gas emissions from burning
fossil fuels, pollution, and limited access
to clean energy affect respiratory diseases
and cause high mortality rates
Hystad et al.
(2019)
11 low- to middle-
income countries
2002-2015
Health outcome, all-cause
mortality, cause-specific
mortality
Cox proportional
hazards models with
SAS.
The use of solid fuels for cooking is a risk
factor for mortality and cardiovascular
disease. The mortality rate is higher in
households that use solid fuel for cooking
than in households that use clean fuel.
Njoh et al.
(2019)
Africa
Basic utilities, health
expenditure per capita,
women's literacy,
mortality
Log linear regression
analysis
Inverse relationship between access to
basic services including electricity and
infant mortality rates.
Wang et al
(2019)
China, 2000-2014
Clean energy, maternal
mortality, income
Fixed effects
Clean energy reduces maternal mortality
in China
Apenteng et al.
(2018)
25 SSA
countries,2007-2011
Energy crisis, power
outages, health outcomes
Binomial regression
models
Positive association between power
outages frequency and mortality in
healthcare facilities, along with zero-
inflated negative binomial regression
models.
Lewis (2018)
United States, 1930 -
1960
electricity access, infant
mortality, fertility
Differences in the
timing of electricity
Rural electrification has increased access
to electricity and tends to reduce child
Author/s
Study area and time
Variables
Methodology
Key findings
access across rural
counties
mortality by 15% to 19%.
Suhlrie et al.
(2018)
Malawi, 2013-2014
Energy, health facilities,
health outcome
Descriptive statistics
and logistic and
multinomial
regressions.
The provision of energy improves service
delivery and overall health outcomes for
patients
This study
48 Sub-Saharan
African countries,
2000-2020
Health expenditure,
income, maternal
mortality, infant and
under-five mortality,
access to electricity, clean
fuel and cooking
technology, internet
access
Generalized quantile
regression developed
by Powell (2020)
Clean fuels for cooking and technologies,
as well as access to electricity reduce
infant, under-five, and maternal mortality
in sub-Saharan African countries.
8
3. Data sources and methodology
Between 2000 and 2020, annual unbalanced panel data were collected from a sample
of 48 sub-Saharan African countries (SSA). The dataset contains observations of
various cross-sections collected over time. Countries were selected based on data
availability and all data were extracted from World Bank Development Indicators [21].
The selected variables, their units and sources are listed in Table 2. Lists of 48 SSA
countries are attached in the appendix.
Table 2: Variables and unit of measurement
Variables
Unit of measurement
Maternal mortality
U Deaths per 100,000 live births
GDP per capita (income)
Constant 2015 US dollars
Infant mortality
Deaths per 1000 live births
Under-five mortality rate
Death per 1000 live births
Access to electricity
(%) of population with access
Clean fuels and cooking technologies
% of population with clean fuels & cooking tech.
Health expenditure
% of GDP
Source: World Bank's Development Indicators (2021).
3.1 Model estimation
Based on other empirical research such as Shobande [10] and Grossman's [59] theoretical
demand for health care, health can be improved over time and human life can be extended
for future well-being, which also depends on the consumption of goods such as energy
consumption and income.
Thus, we estimate the impact of access to electricity, clean
fuels and cooking technologies on health outcomes (maternal, infant and under-five
mortality) as follows:-
=.......................................... (1)
Where health represents (maternal, infant and under-five mortality), for i=1,2....48 countries
and 1, 2......T (time).
Energy access includes (access to electricity, clean fuels for
cooking and technologies).
Countries are shown by and time . is a vector of control
explanatory variables including GDP per capita and health expenditure The selection of
control variables are justified by its use in other previous empirical studies linking to health
outcomes [20], [60], [67].
The term and are coefficient of explanatory variables.
9
= idiosyncratic error (error term for country and time given by
= time specific fixed effects,
= is the country specific fixed effects constant in time.
Powell (2020)'s generalized quantile approach is given from equation (1) as follows:
( ) = ....2
In equation 2, the quantile is given by symbol ( ).
Using generalized quantile regression (GQR), the explanatory variables can be
separated into treatment variables (energy access) and control variables. GQR permits
the conditional dependence of the dependent variable (treatment effect) on the
treatment variable but not on the control variables [46]. According to [29], the
generalized quantile regression method is used in the context of lagged explanatory
variable as instruments to estimate the impacts of unconditional quantile treatment
and to address both endogenous and exogenous policy variables. Finally, it uses
adaptive Markov chain Monte Carlo sampling (MCMC) and numerical optimization in
the estimation process to estimate the results
and control endogeneity of variables
[29],
[60],[61].
4. Results and findings
Table 3 shows the descriptive statistics for unbalanced panel data for 48 selected sub-
Saharan African countries from 2000 to 2020.
Table 3: Descriptive statistics
Access to
electricity
(%)
Clean fuel
& cooking
technology
(%)
Health
expenditure
(% of GDP)
GDP per
capita
(US
dollar)
maternal
mortality ( per
100,000 live
births)
infant
mortality
(per 1000
live
births)
Under-
five
mortality(
per 1000
live births)
(N)
910
960
913
972
165
1008
1008
Mean
37.32
19.48
5.34
2120.45
450.32
59.83
92.5
Median
33.82
4.3
4.81
1081.93
405
58.35
90.7
Minimum
1.27
0
1.26
258.63
5
11.8
13.7
Maximum
100
100
20.41
16438.64
2241
138.1
224.9
Note: N=Number of observations
10
Table 3 shows that the average access to electricity in SSA is 37.32%. However,
access to electricity varies in many countries. For example, In 2019/2020, only 39.9%
of all households in Tanzania were connected to electricity. Rwanda had 46.6%,
Uganda 42.1%, Kenya 71.4%, Ethiopia 51.1%, Malawi 14.9%,
Congo 19.1%,
Burundi 11.7%, Somalia 49.7%, Togo 54.0%, Zimbabwe 52.7%, Mauritius 99.7%, and
Seychelles 100.0% [47]. Only Mauritius
and the Seychelles were able to connect all
households to the electricity grid. Kenya, Zimbabwe, Ethiopia and Togo had a
connectivity level of over 50%.
Although SSA countries have great potential for clean energy generation such as
hydro, wind and solar, both production and consumption are still below average.
Installation and running the power grid are also more expensive in Africa. For
example, a domestic electricity unit in Tanzania costs between $0.098 and $0.101 per
kWh, while in Kenya it costs between $0.177 and $0.145. Similarly, it costs for
Congo ($0.08-10.095), Malawi ($0.110-0.152), Rwanda ($0.241-0.090), and Mauritius
($0.137-0.123). These few examples show that running electricity in African countries
is still the most expensive and contributes to energy poverty [47].
The average use of clean fuels and cooking and technology in SSA is 19.48%. Clean fuels
for cooking and technology, as well as access to electricity in the region reflect persistently
high levels of energy poverty. The average maternal mortality rate is 450 per 100,000 live
births. The Sustainable Development Goals (SDGs 3) call for reducing the maternal
mortality rate to less than 70 per 100,000 live births. Similarly, the infant and under-five
mortality rates are 59 and 93 per 1,000 live births, respectively. With these improvements in
health outcomes, SSA is still a long way from achieving Sustainable Development Goal 3,
which aims to reduce infant mortality to at least 12 per 1,000 live births and child mortality
to up to 25 per 1,000 live births.
Table 4 shows the results of the quantile estimation. Our results show that access to
electricity negatively impacts under-five, infant and maternal mortality at almost all
quantiles (i.e., 25th, 50th, 60th, 75th, and 90th). This means that increased access to
electricity reduces infant, child and maternal mortality in almost all quantiles (i.e. 25th,
50th, 60th, 75th and 90th). This is evidenced by a statistically significant result and a
11
negative sign of the coefficients of the variable. Overall, the positive role of access to
electricity in reducing infant, maternal and child mortality is evident in the lower, middle
and higher quantiles (25th, 50th, 60th, 75th and 90th).
The results also show that clean fuels and cooking technology reduce maternal mortality in
the 25th, 50th, 60th and 75th quantiles. This means that an increase in clean fuels and
cooking technologies has a significant impact on reducing maternal mortality in SSA. The
results also show that clean fuels and cooking technology reduce infant mortality in SSA at
the 50th, 60th, 75th and 90th quantiles. Furthermore, clean fuels and technology reduce
under-five mortality in the 50th, 75th and 90th quantiles. This means that an increase in
clean fuels and cooking technology has a significant impact on reducing under-five
mortality in the region. However, at the 25th and 60th quantiles, clean fuels and cooking
technology have a positive impact on increasing under-five mortality in SSA. This means
an increase in clean fuels and cooking technology could lead to the deaths of children. The
same result applies to infant mortality in the 25th quantile. This is likely because cooking is
also the leading cause of house fires, killing both infants and children under five. Overall,
clean fuels and cooking technology have heterogeneous impacts on infant and under-five
mortality in SSA.
On the other hand, the results show that in most quantiles, increases in income (GDP per
capita) and health expenditure lead to reductions in maternal, infant and under-five
mortality in sub-Saharan Africa. This means that income and public health expenditure
positively influence health and are an important factor in population survival in sub-
Saharan Africa [62], [63]. Lastly, our result also shows that there is no multicollinearity
among the independent variables since the variance inflation factor (VIF) is less than 10.
12
Table 4
: Generalized Quantile results estimates
Note: Standard errors in parentheses, ***, ** and * show significant at 1%, 5% and 10% level respectively.
Acceptance rate is set at 0.5 for the algorithm. The algorithm performs 1000 draws and burn in of 100 through
MCMC diagnostic. Year dummies (time -fixed effects) are included in the regression. All independent
variables are lagged by one as instrumental variables.
4.1 Robustness test
To determine the robustness of our results, we performed a Kernel based Regularized Least
Square (KRLS) method, which is used as an alternative to the quantile regression method.
Kernel Regularized Least Square (KRLS) is a machine learning technique with econometric
characteristics developed by Hainmueller and Hazlett [64]. It is designed to engage with
econometric models without imposing arbitrary functional forms on the relationship
between the dependent and independent variables. It outperforms traditional econometric
and other learning techniques in different ways. First, there is no need to pretest the
variables using unit roots [65]. Second, average marginal effects and pointwise derivatives
are estimated. Third, heterogeneity, additivity and nonlinearity are taken into account [66].
Fourth, it addresses regression or classification problems of cross-sectional dependence and
A: Impact of energy access on maternal mortality in SSA
Lower quantile
Middle quantile
Higher quantile
Variables
0.25
0.50
60
0.75
90
Access to electricity
-1.034***(.164)
-.352***(.088)
-.252 (.164)
-.119***(.019 )
-.823***( .053)
Clean fuels & tech. cooking
-.242***(.038)
-.101*(.056)
-.097**(.038)
-.261*** (.012)
-.005( .068)
Health expenditure
-.350(.439)
-.389***(.104)
-.113 (.345)
-.262***(.030)
-2.33***(.265)
GDP per capita (income)
.252 (.186)
-.277**(.112)
-.321(.221)
-.188***(.017)
-.717***( .040)
Observations
159
159
159
159
159
Instruments used
4
4
4
4
4
B:Impact of energy acc