Energy 239 (2022) 122365

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Energy efficiency and sustainable development goals (SDGs)

Abdulrasheed Zakari a, b, Irfan Khan a, Duojiao Tan c, *, Rafael Alvarado d, Vishal Dagar e
a
School of Management and Economics, Beijing Institute of Technology, Beijing, 100081, China
b
Alma Mater Europaea- ECM Maribor, Slovenia
c
Accounting School, Hubei University of Economics, Wuhan, PR China
d
Esai Business School, Universidad Espíritu Santo, Samborondon, 091650, Ecuador
e
Amity School of Economics, Amity University, Noida, 201 301, India

a r t i c l e i n f o a b s t r a c t

Article history: This study attempts to connect Sustainable Development Goals (SDGs) with energy efficiency for 20
Received 10 August 2021 Asian and Pacific (AP) countries using Data Envelopment Analysis (DEA) from 2000 to 2018. The Panel
Received in revised form Correction Standard Error (PCSE) estimates found that sustainable economic development and energy
3 October 2021
efficiency are positively related, suggesting that sustainable economic development is associated with
Accepted 17 October 2021
Available online 25 October 2021
increased energy efficiency. Similarly, we found that sustainable financial development is also positive,
indicating that sustainable financial development increases energy efficiency. In further analyses, we
confirmed a positive impact of green innovation on energy efficiency. The S-GMM estimator confirmed
Keywords:
Data envelope analysis
similar findings, showing robust to alternative econometric model. Based on these findings, we
Energy efficiency recommend policies that tend to promote sustainable economic and financial development.
PCSE © 2021 Elsevier Ltd. All rights reserved.
S-GMM
Sustainable development

1. Introduction driven by sustainable development and growth [6], which means

an economy's growth increases energy demand. If energy is con-
The sudden surge in energy use and carbon dioxide emission strained, economic growth pulls back in turn. This shows the
(CO2) in recent times is alarming. Many believe that the rising bidirectional relationship that exists between sustainable devel-
population level in urban areas is the brain behind the surge [1,2], opment and growth in energy efficiency. Furthermore, financial
while others argued that goods and services production might be development could affect energy demand by protecting customers
the cause [3,4]. Whatever may be the case, energy has come to stay from borrowing money to buy items such as automobiles, houses,
as a factor that determines sustainable development. Hence, it has refrigerators, air conditioners, and washing machines. In other
been projected that global industrial sector energy use will rise to words, financial development provides an easier path for cus-
about 30% by 2050 [5], while consumption of end-products is tomers to satisfy their needs and wants, which can affect a coun-
projected to grow and exceed the approximate figure of 310 qua- try's overall demand for energy [7].
drillions British Thermal Units (BTU) [5]. Given the projection, en- Following the Sustainable Development Goals (SDGs)- 7, there is
ergy use rise is likely to affect economic activity, society, and the the need for cheaper, affordable, and clean energy. Putting it
environment. Therefore, governments and policymakers should differently, the provision of energy efficiency and reduction in
pay attention to cleaner energy policies to reduce carbon dioxide carbon dioxide emissions becomes necessary by 2030 [8]. Hussaini
emissions. [9] argued that energy efficiency, as often used as the term for
There are two possible ways to address these issues, that is technology, appears behavioral in practice but refers to the eco-
sustainable financing and economic development. Energy has been nomic value of fuel consumption. On the other hand, energy effi-
ciency provides an effective and efficient energy cost for producing
output and services [10,11]. More importantly, it reduces carbon
dioxide emissions, reduces demand for energy import, and lowers
* Corresponding author.
E-mail addresses: [email protected], [email protected] the cost on the individual household and economy as a whole [12].
(A. Zakari), [email protected], [email protected] (I. Khan), Beyond this, buildings, transportations, industries, and energy
[email protected] (D. Tan), [email protected] (R. Alvarado), generation sectors benefit from energy efficiency.
[email protected] (V. Dagar).

https://doi.org/10.1016/j.energy.2021.122365
0360-5442/© 2021 Elsevier Ltd. All rights reserved.
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

Against this background, we focussed on energy efficiency and globally becoming acceptable to source affordable and accessible
sustainable development for economic and financial factors vital clean energy. The global climate and environmental issues are
for energy efficiency. Countries with sustainable development major reasons behind the sudden surge in energy efficiency
foresight must embrace energy efficiency and carbon dioxide [23e25]. However, others believe that companies and firms' quest
reduction policies [13]. However, previous studies [13e15]) focused for profits is also the one-side reason for energy efficiency [20,21].
on economic growth and financial development as a determinant of Be that as it may, be; it is the reality that energy efficiency holds
energy efficiency. We argued that economic growth focuses on more benefits to tap from. Therefore, the promotion of energy ef-
producing goods and services without considering the environ- ficiency became very important in the mind of the progressive.
mental consequences that expose the lapses of using economic Hence, we review the existing literature on their suggested ways to
growth as the determinant of energy efficiency. promote energy efficiency.
Secondly, prior literature [15e18]; considered financial devel- One factor that can improve energy efficiency is by sustaining
opment to determine energy efficiency. This indicator reflects economic development. For instance, implementing renewable
financial base growth, while the future generation is exposed to energy has a significant cost burden; however, selling carbon credit
uncertainty. Thirdly, the Asian region top the continents as the can offset this cost burden, leading to indirect health benefits
most energy secured region, with 7 of its member countries ranked [26e28]. Another way to achieve energy efficiency in a country is to
among the world's most attractive renewable energy source mar- over-hold its economic activities, considering the health of our
kets [19]. Despite this growth of the energy market in Asia, little has environment. That means promoting economic growth and energy
been done to research how to promote energy efficiency using consumption is not out of place while considering energy efficiency
SDGs. It will be unfair to side-line these nations' tenacity in up- [14,29]. Rosen [30] argued that countries with sustainable eco-
holding clean energy system policies in our study. nomic development could draw from their success only by focusing
Given this fact, we considered 20 Asian and Pacific countries and on discovering sustained energy resources.
investigated the role of sustainable development goals (i.e., sus- Financial development is another factor that significantly drives
tainable economic and financial development) in promoting energy energy efficiency [31]. Through energy intensity, financial devel-
efficiency. Secondly, we estimate the impact of green innovation in opment spurs energy efficiency. Specifically, the development of
achieving energy efficiency. We used DEA analysis to construct the the finance system tends to bring relief to energy users by giving
energy efficiency index, while the Panel Correction Standard Error confidence in the possibility of outsourcing finance for energy,
(PCSE) model and robustness of the S-GMM estimator to achieve which consequently increases energy efficiency [16]. In support,
our target objectives. Chang (2015) argued that financial development indicators could
Our study makes incremental contributions to literature and promote energy efficiency through energy consumption. However,
practice in several ways. First, this study enriches the literature on the magnitude of this impact depends on the income level and the
energy efficiency by focusing on the country-level outcome. Most of aspect of financial development. For example, private and domestic
the existing literature [20,21] largely dominates by firm-level credit primarily increase energy consumption and energy-saving or
studies. We argue that the government provides and regulates efficiency in low-income countries.
the energy market in most Asian and Pacific countries, enabling However, Li et al. [17] argued that energy efficiency could be
firms’ access and affordable energy. The government has the power negatively affected, which means it can witness a downturn. The
and motivation to influence energy saving or efficiency improve- possible factor in the downturn of energy efficiency is urbanization.
ment; hence, the country-level study on the energy efficiency of the Urbanization prompts a larger migration of rural population to an
environment is essential. Thirdly, we contribute to the literature on urban area, leading to an increase in energy usage and consump-
the sustainable economy by focusing on the effect of nexus be- tion. Because of the divergence in the living conditions of urban and
tween development in terms of income and efficient consumption rural residents, the surge in energy consumption will reduce energy
of energy as compared with previous research [14] and Chang efficiency and unnecessary energy waste. More so, urbanization
2015) that provided analysis of economic development and finan- leads to unemployment in the rural area, as such government is
cial development such as economic growth. We argued that sus- forced to seek for solution to the rising unemployment and
tainable economic and financial development is emerging as a sometimes demand to relax some environmental regulation on
critical area nation find better ways to manage the economy selected enterprises known for high energy consumption and look
without jeopardizing the future generation. up to them to provide more job opportunities, which also likely to
Although our study shared similarities with Ziolo et al. [22]; we reduce energy saving or efficiency [18]. Similarly, financial devel-
argued that our study provides more robust findings of energy ef- opment reduces energy efficiency in high-income countries, spe-
ficiency nexus because we focussed on top renewable energy cifically the value of traded stocks and stock market turnover
countries in the world. As done in Ziolo et al. [22]; lumping more (Chang, 2015).
countries into renewable energy driving countries may produce There are quite a few studies concerning energy saving or effi-
biased results. Arguably, usage or consumption of renewable en- ciency, sustainable financing, and economies worldwide. However,
ergy and regulations matter in many countries in Asia and the Pa- the previous studies, such as Ziolo et al. [22]; emphasized the OECD
cific (Consultancy.asia, 2020). Fourthly, we extended the energy economies while neglecting the countries that are most attracted to
efficiency nexus by including the green innovation nexus. clean energy. According to a news report by EY, Asia is growing to
This paper is prepared by keeping in mind the following. Section the top of the globe as the continent with the most attractive
2, present the literature review, while the research methodology renewable energy use (Consultancy.asia, 2020). As we know,
employed herein are reported in section 3. Section 4 shows the renewable energy forms part of energy efficiency. It would not be
results and discussion based on empirical data sets collected for the wise to exclude countries that have embraced renewable energy
study. We concluded and made a policy recommendation in Section use in their system, unlike the previous literature, which ignored
5. the countries with higher renewable energy demand. Given this
fact, the findings from previous studies may not reflect the reality of
2. Literature review the nexus between energy efficiency and sustained economy. Sec-
ondly, previous studies failed to account for the innovation related
It is clear from the United Nations eUN -SDG goal e 7 that it is to environmental performance, despite the global calls for a green
2
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

economy and environment. We are motivated by this shortfall to for adopting OLS-PCSE is for all variables to be stationary at
examine the role of nexus for development in terms of income and either level or first difference or a mixture of them.
efficient energy consumption in the AP countries while accounting 2. Baseline estimation, we test for OLS-PCSE estimation on the
for green innovation. variables under consideration.
3. Alternative estimation, we estimate another alternative variable
of interest using the OLS-PCSE estimator.
3. Data and econometric method
4. Robustness estimation is a post-analysis estimate. We check
3.1. Sample data the possibility of changes in the main finding following the inclu-
sion of lags of dependent and independent variables and check for
This study employs a panel dataset consisting of 20 Asia and endogeneity using OLS-PCSE and S-GMM models, respectively.
Pacific countries between 2000 and 2018. Data selection was sub-
jected to data availability, and details of our variables and source 3.3. Estimation method
are reported in Table 1, while the list of sampled countries is pre-
sented in Appendix A. Furthermore, the choice of Asia and Pacific 3.3.1. Unit-root test
countries is due to the position it holds being among the top energy In this section, we estimate for stationarity of the variables. To
users and the most stringent in environmental policies region in achieve this, we perform LLC and HTT unit-root tests to show if
the world (see Table 2). there is stationary or not in the series [42]. LLC and HTT unit-root
tests are fantastic tests because they offer more reliability. The
general form of the model considered in this paper can be written
3.2. Theoretical background and model specification as follows:

Following prior studies [22,32], we constructed Total Factor yit ¼ ai þ dt þ mit t ¼ 1; …:T : i ¼ 1; …:N
Energy Efficiency (TFEE) using Data Envelopment Analysis (DEA)
enveloped in stata software. Coelli [33] argued that linear mathe- n o
uit ¼ rui;t1 þ εit εit  0; s2t s2ε ðiÞ
matical techniques could measure the efficiency of decision-
making units (DMU). Going by that assertion, Total Factor Energy
. .
Efficiency (TFEE) is calculated by taking the division of the actual yit ¼ ð1  rÞai þ ð1  rÞdt þ pyi;t1 þ εit r <1
energy input over energy input as demonstrated by Hu and Wang
[34] in the following equation: n o
or yit ¼ yi; t1 þ d þ εit εit  0; s2t s2ε ðiÞ r¼1
TEI ði; tÞ AEI ði; tÞ  TAði; tÞ
TFEE ði; tÞ ¼ ¼ (1)
AEI ði; tÞ AEI ði; tÞ
As eq (1) the index TFEE represents the efficiency level of energy
consumption in a country, TEI is the target energy inputs in a 3.3.2. Ordinary least square with panel corrected standard error
country, and AEI represents the actual energy inputs of a country. (OLS-PCSE)
The target energy input is the minimum energy input in a country; Beck and Katz [43] developed the PCSE model to overcome the
therefore, the actual energy input is larger than or equal to target shortcomings posed by Time Series Cross-Section (TCSC) model
energy inputs, making TFEE equal to zero or unity. In essence, a using OLS procedures. Meanwhile, OLS has been criticized due to
unity TFEE score occurs when actual energy inputs of DMU are the presence of non-spherical error and the inability of the model
equal to targeted energy inputs, indicating higher energy efficiency. to provide consistent estimates. Therefore, the PCSE model is an
However, the TFEE score is zero when the actual energy input is far alternative for the OLS model because it provides accurate esti-
away from the targeted energy inputs, suggesting low energy mation in contemporaneously correlated or heteroskedastic panel
efficiency. errors. Given the following equation:
Therefore, we follow the previous studies [35,36] and
[22,37e41]), assess the role of sustainable economic and financial yit ¼ qyit1 þ xit b þ εit (4)
development on the energy efficiency of 20 Asian and Pacific Accepting that the blunders in condition (4) are time-related
countries. We specify our baseline model as follows: autonomous, the change covariance framework of the mistakes
can be built as follows:
Total factor energy efficiencyit ¼ b0
þ b1 sustained economic developmentit þ εit (2) U ¼ S5IT (5)

Where S is the N x N lattice of mistake differences and contem-

Total factor energy efficiencyit ¼ b0 poraneous covariances and address the Kronecker item.
þ b1 sustained financial developmentit þ εit (3) Given S refer to the T x N grid of the OLS residuals. E'E/T gives a
predictable gauge of S. PCSEs are accordingly assessed by the
As eq (2) assesses the role of sustainable economic development square base of the askew of:
on energy efficiency, while eq (3) shows the relationship between
 0 
sustainable financial development and energy efficiency. Where “i” EE
refer to country, “t” is the year and “ε” is the error term. All other ðX 0 XÞ1 X 0 5 IT XðX 0 XÞ1 (6)
T
variables are defined in Table 1, while Fig. 1 demonstrates the
sequence of methodology as follows:

1. The pre-estimation analysis is the criteria for adopting the 3.3.3. System generalized moment of method (S-GMM)
estimation model by conducting a unit root test. The condition We employed the system, Generalized Method of Moments
3
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

Table 1
Summary of empirical Studies.

Authors Countries Variables Methodology Findings

The role of sustainable economic development on energy efficiency

[21] Jiangsu province, China Capital, Labour, energy, output value, waste DEA and Tobit There is a positive relationship
water, waste gas, soot, age, tax, expense, size, models between output efficiency and
watertreat, and Gastreat energy efficiency
Türkoglu and Kardogan 100 countries Energy efficiency and sustainable development Conceptual There is a positive relationship
[14] approach between sustainable
development and energy
efficiency
Rosen [30] USA energy efficiency, sustainable development, Regression Sustainable economic
industrialized society, developing society, model development promote energy
sustainable energy resources, electricity efficiency
generation, heat generation, energy utilization,
Ganda and Ngwakwe energy efficiency, sustainable development, Conceptual Energy efficiency plays multiple
[13] carbon reduction, climate change, and poverty approach roles towards sustainable
reduction economic development
Role of sustainable financial development on energy efficiency
[31] 58 Belt & Road (B&R) economies Energy efficiency, financial development, trade, Panel EGLS There is a positive relationship
urban population, Land area, and technological with random between financial development
progress effects and and energy efficiency
Tobit
regression
Ahmed [16] BRICS countries Energy consumption, real GDP, financial Panel VECM Financial development and
development, and trade openness model capital accumulation contribute
to energy efficiency
Li et al. [17] 30 Chinese provinces Energy efficiency, urbanization rate, foreign Spatial panel A negative relationship
trade, income level, energy price, and energy data models between urbanization and
consumption energy efficiency
Sadorsky [7] 22 emerging countries Energy consumption, financial development, S-GMM A positive relationship between
financial development and
energy consumption

(GMM) estimators, to capture the dynamic models of panel data as identification.

first use by (Arellano and Bond [44] and Arellano and Bover [45]. Simultaneously, lagged values of regressors are included as in-
The standard system, GMM, in levels (1) and in first difference (2) struments, which remove fixed effects that are seen influencing the
are summarised as follows: assessed relationships. In this case, Helmet transformations are
advised to be performed for the regressors, which transformations
X
5 embody the mean differencing of the indicators: the mean of future
Yit ¼ s0 þ s1 Yitt þ s2 X1it þ s3 X2it þ s4 X3it þ dj Wh;i;tt observations is subtracted from the variables instead of deducting
h¼1 the previous observations for the contemporaneous one [46]. These
þ hi þ xt þ εit transformations provided same-level conditions between for-
(7) warding differenced variables and lagged values. This condition
does not care about the number of lags because the set-aside target
minimizes data loss, with the exemption of the last observation for
Yit Yitt ¼ s1 ðYitt Yit2t Þþ s2 ðX1itt X1it2t Þ each country.
þ s3 ðX2itt X2it2t Þþ s4 ðX3itt X3it2t Þs1 Yitt Lastly, the exclusion restriction is the process by which the
X
5   dependent variable is influenced by the strictly exogenous variables
þ s2 X1it þ s3 X2it þ s4 X3it þ dj Wh;i;tt Wh;i;t2t exclusively through the endogenous explaining variables. The dif-
h¼1 ference in Hansen Test is used to assess the validity of the exclusion
þðxt  xt2t Þþðεit εit2t Þ restriction for instrument exogeneity. Notably, the standard
(8) instrumental variable approach rejects the null hypothesis of the
Sargan Over identifying Restrictions test. This means that in-
Note that Yit is the dependent variable of country i at period t ; struments do not outline the outcome variable via the suspected
s0 is a constant, while t represents the coefficient of autoregression endogenous variables. At the same time, in the GMM estimation
which presents specifications that take into cognizance the issues technique, the Difference in Hansen Test is the required informa-
in degrees of freedom. The coefficient X1;X2;X3 are independent tion criterion used to investigate if years are strictly exogenous.
variables; W show the vector of control variables, hi is the country-
specific effect, xt is the time-specific constant and εit the error term.
The S-GMM estimation following three sets of procedures, 4. Results and discussions
which are identification, simultaneity, and exclusion restrictions
[46,47]; Asongu and Odhiambo, (2019a & b) and Tchamyou [46]. 4.1. Univariate analysis
First, the S-GMM assumed that all independent variables are sus-
pected to be endogenous variables. Therefore, adopted “gmmstyle” Table 3 presents the descriptive statistics of the variables. The
for predetermined variables, in which years are treated as strictly statistical information is presented in the following order: mean,
exogenous, and the method for treating the ivstyle (years) is “iv 1st, 90th, and standard deviation. The mean of Total factors energy
(years, eq (diff)).” This is because the years cannot become efficiency is 3906.501, suggesting more than 3000 units of energy
endogenous in first-difference [47]. This process is known as efficiency were used among the sampled countries. However, the
4
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

Table 2
Variables description.

Variable name Definition Sources

Total factor energy efficiency Analyzed by input factors: gross capital from (constant 2010 USD) as capital, labor, and primary DEA
energy consumption
Adjusted net saving Modified net savings involves damage from emission (% of GNI) WDI (2021)
GDP per capita GDP per capita (constant 2010 US$) WDI (2021)
Industry value-added Industry (including construction), value added (% of GDP) WDI (2021)
Urbanization Urban population (% of total Population) WDI (2021)
Renewable energy consumption Renewable energy consumption (% of total final energy consumption) WDI (2021)
Financial development Domestic credit to the private sector by banks (% of GDP) WDI (2021)
Financial market Financial Markets Index IMF database (2021)
Financial institution Financial Institutions Index IMF database (2021)
Research and development Expenditure on R&D (% of GDP) WDI (2021)
Health expenditure Expenditure on Health Infrastructure (% of GDP) WDI (2021)
Foreign direct investment Foreign direct investment, net inflows (% of GDP) WDI (2021)

standard deviation is higher, indicating high variation among the that urbanization causes a reduction in energy efficiency. Overall,
sampled countries; specifically, Hong Kong has the highest energy sustainable economic development promotes energy efficiency,
efficiency (see Fig. 2). This may be due to their stringent policy on suggesting saving, industrialization, and renewable energy con-
cleaner energy consumption. Among the sustainable economic sumption to meet UN sustainable development goal 7 (affordable
development variables (modified/adjusted net saving, GDP of coun- and clean energy).
tries per capita, industry-specific value-added, urbanization, and In the case of sustainable financial development, we found that
renewable energy consumption), we observed a similar trend except domestic credit to the private sector, capital markets, financial in-
for urbanization with the mean value of 53.746 and a very small stitutions, R&D spending, health spending, and foreign direct in-
standard deviation, suggesting small variation among the sampled vestment (0.538***, 2.286***, 2.234***, 0.125***, 1.427***, and
countries. Furthermore, most of the variables of sustainable 0.833***) are positive and significant at 1%, an indication that do-
financial development (i.e., domestic or institutional credit facilities mestic facilities for credit facilitation for the private sector, financial
for the private sector, financial market, financial institution, and health market, financial institution, expenditure on innovation (R&D),
expenditure) show a low standard deviation value, indicating small expenditure on health infrastructure and foreign direct investment has
variation among the sample countries. In contrast, research and an increasing influence on the energy efficiency in the Asian and
development and foreign direct investment show high standard de- Pacific countries. Put it differently, it means the sustainable finan-
viation, suggesting high variation among sampled countries. cial development variables are associated with increased energy
efficiency. One possible reason for the positive influence is that
financial institutions offer funds to individuals and firms through
4.2. Multivariate analysis loans and grants to enable them to expand their production base. In
essence, firms can purchase more innovative energy systems to
4.2.1. Main results achieve energy efficiency. This is partly consistent with Mukhtarov
Table 4 show the regression results, and we presented the re- et al. [52] and Sadorsky [7]; who found that financial development
sults base on the sustainable development for the variables which promotes energy consumption.
are connected in terms of economic benefits (i.e., adjusted net
saving, GDP per capita, industry value-added, urbanization, and
renewable energy consumption) in column 1, while (in) column 2, 4.2.2. The effect of the green innovation
we reported the sustainable development for the variables (i.e., The previous analysis on the relationship between sustainable
domestic credit to the private sector, financial market, financial economic and financial development and energy efficiency has
institution research and development, expenditure on health infra- been represented in Table 4, but this estimation largely ignores any
structure and foreign direct investment). external pressure that could change the direction of energy effi-
The coefficient of adjusted net saving, industry value-added, and ciency. In practice, technological innovations are cheaper and
renewable energy consumption (1.405***, 1.473***, 1.668***, and environmental friendly; hence, they are likely to achieve energy
0.411***) are positive and significant at 1%, suggesting that efficiency and optimization [53]. Many countries encourage inno-
adjusted net saving, industry value-added, and renewable energy vation activities to achieve specific objectives such as productivity
consumption are attached with the increase of energy efficiency for and environmental performance. Therefore, we carried out the test
Asia and Pacific. This follows the assertion that individual and or- on whether other policies such as green innovation affects energy
ganization per capita income increases in an economic develop- efficiency. For this reason, we introduced a new equation; Green
ment environment while the energy sector remains competitive. innovation is a variable that measures the total number of patents in
Hence, people have more income to spend on competitive energy environmental-related technologies.
and achieve optimization. This is partly consistent with Razzaq The results are presented in Table 5, and the coefficient of green
et al. [48] and Liimatainen and Po € lla
€nen [49]; who found that innovation (1.8907***) is positive and significant at 1%, suggesting
sectoral economic development promotes energy efficiency. that green innovation promotes energy efficiency. This is possible
However, we found that the coefficient of urbanization because green innovations are smarter, cheaper, and environmen-
(1.668***) is negative and significant at 1%, indicating that ur- tally friendly. Hence, they are cost-effective, and therefore, energy
banization reduces energy efficiency. This is possible because the efficiency can be achieved. To put it differently, green innovation
migration of people to urban areas from their native or rural areas can promote environmental performance through energy effi-
increases energy consumption and energy wastage due to the rural ciency without harming the economy. This is consistent with Sun
habit of carelessness for energy usage. This finding is consistent et al. [54]; who found that green innovation and institutional
with Lv et al. [50]; Oryani et al. [51]; and Li et al. [17]; who found quality promote energy efficiency. Column 2 presents the
5
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

Fig. 1. Conceptual framework.

Table 3
Descriptive statistics.

Variables N Mean 1st Std. Dev. 99th

Total factor energy efficiency 380 3906.501 0.08 12907.102 67170.795

Adjusted net saving 380 15.724 0.633 11.329 40.033
GDP per capita 380 15253.399 455.277 17727.486 56058.466
Industry value added 380 30.792 0 13.646 72.657
Urbanization 380 53.746 13 28.468 100
Renewable energy consumption 380 26.466 0 29.995 91.999
Domestic credit to private sectors 380 72.002 3.424 49.537 214.128
Financial market 380 0.391 0 0.295 0.9
Financial institution 380 0.492 0.2 0.226 0.94
Research and development 380 0.718 0 1.087 4.227
Health expenditure 380 4.398 0 2.543 10.835
Foreign direct investment 380 4.312 2.757 7.823 41.065

regression analysis of green innovation with alternative control significant at 5%, indicating that green innovation promotes energy
variables (sustainable financial development). The results show efficiency. The finding is a robust alternative to our finding in col-
that the coefficient of green innovation (1.3507**) is positive and umn 1.

6
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

Fig. 2. Descriptive figure by countries.

4.3. Robustness and endogeneity (lagged and S-GMM) significant; this is consistent with the main results in Table 4.
Secondly, we considered the effect of the lag of the energy effi-
In this section, we test the reliability of our model in light of our ciency, adjusted net saving, and domestic credit to the private sector.
main results in Table 4. This becomes necessary to have reliable To do this, we took the lag of energy efficiency, adjusted net saving,
findings and a stable model. Secondly, reverse causality is also and domestic credit to the private sector separately, and the results
possible; hence, we employed different econometric identification are presented in Table 7. The results show that the effects on energy
to address these concerns. Firstly, we included the country and year efficiency remain positive under the lag of energy efficiency,
effect to capture the time-varying country-effect on energy effi- adjusted net saving, and domestic credit to the private sector. This
ciency. As presented in Table 6, the results indicate that sustainable is not quantitatively different from the main results in Table 4.
economic and financial development variables remain positive and Thirdly, following Wang et al. [55], we employed the System-

Table 4
Main results.

Variables (1) (2)

Total factor energy efficiency Total factor energy efficiency

Domestic credit to the private sector 0.538*** (0.146)

Financial market 2.286*** (0.289)
Financial institution 2.234*** (0.305)
Research & development expenditure 0.125*** (0.0358)
Health expenditure 1.427*** (0.183)
Foreign direct investment 0.833*** (0.118)
Adjusted net saving 1.405*** (0.117)
GDP per capita 0.169** (0.133)
Industry value-added 1.473*** (0.178)
Urbanization 1.668*** (0.247)
Renewable energy consumption 0.411*** (0.0938)
Constant 7.227*** (0.494) 2.479*** (0.226)
Observations 380 380
R-squared 0.693 0.247
Number of countries 20 20

Standard errors in parentheses.

***p < 0.01, **p < 0.05, *p < 0.1.

7
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

Table 5
The effect of total green innovation.

Variables (1) (2)

Total factor energy efficiency Total factor energy efficiency

Total green innovation 1.8907*** (5.3608) 1.3507** (1.6907)

Domestic credit to the private sector 0.00656* (0.00776)
Financial market 0.0265** (0.0369)
Financial institutions 0.127*** (0.0327)
Research and development expenditure 0.00605*** (0.00597)
Health expenditure 0.0687*** (0.0217)
Foreign direct investment 0.0279*** (0.00976)
Adjusted net savings 0.000963*** (0.000258)
GDP per capita 0.0573* (0.0302)
Industry value-added 0.0386* (0.0233)
Urbanization 0.349*** (0.108)
Renewable energy consumption 0.0124*** (0.00481)
Constant 0.341*** (0.0738)- 0.594*** (0.0163)
Observations 380 380
R-squared 0.671 0.726
Number of countries 20 20

Standard errors in parentheses.

***p < 0.01, **p < 0.05, *p < 0.1.

Table 6
Robustness e Country and year effect.

Variables (1) (2)

Total factor energy efficiency Total factor energy efficiency

Domestic credit to the private sector 0.338*** (0.108)

Financial market 0.411** (0.321)
Financial institution 1.958*** (0.630)
Research and development expenditure 0.0520** (0.0245)
Health expenditure 1.672*** (0.496)
Foreign direct investment 0.00796 (0.0542)
Adjusted net saving 0.283*** (0.0786)
GDP per capita 1.496*** (0.350)
Industry value-added 0.265** (0.107)
Urbanization 2.568*** (0.784)
Renewable energy consumption 0.0685* (0.0463)
Year-effects Yes Yes
Country-effects Yes Yes
Constant 3.343*** (0.944) 1.574** (0.652)
Observations 380 380
R-squared 0.933 0.932
Number of countries 20 20

Standard errors in parentheses.

***p < 0.01, **p < 0.05, *p < 0.1.

Table 7
Robustness: lagged of total factor energy efficiency, adjusted net saving, and financial development.

Variables (1) (2) (4)

Total factor energy efficiency lag Total factor energy efficiency Total factor energy efficiency

Adjusted net saving 1.392*** (0.123)

GDP per capita 0.150** (0.149) 0.182** (0.127)
Industry value added 1.528*** (0.182) 1.560*** (0.179)
Urbanization 1.642*** (0.282) 1.648*** (0.225)
Renewable energy consumption 0.407*** (0.0987) 0.412*** (0.0946)
Adjusted net saving _lag 1.433*** (0.114)
Financial market 2.258*** (0.290)
Financial institution 2.142*** (0.344)
Research and development expenditure 0.122*** (0.0374)
Health expenditure 1.443*** (0.192)
Foreign direct investment 0.806*** (0.123)
Domestic credit to the private sector _lag 0.436*** (0.153)
Constant 7.332*** (0.515) 7.292*** (0.476) 2.348*** (0.236)
Observations 360 360 360
R-squared 0.687 0.712 0.237
Number of countries 20 20 20

Standard errors in parentheses.

***p < 0.01, **p < 0.05, *p < 0.1.

8
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

Table 8
Robustness- endogeneity.

Variables (1) (2)

Total factor energy efficiency Total factor energy efficiency

Domestic credit to the private sector 0.00672** (0.101)

Lag Total factor energy efficiency 0.852*** (0.0350) 0.962*** (0.0210)
Financial market 0.0171** (0.199)
Financial institution 0.0849** (0.236)
Research and development expenditure 0.00573* (0.0411)
Health expenditure 0.0381** (0.121)
Foreign direct investment 0.0374*** (0.0535)
Adjusted net saving 0.284*** (0.0728)
GDP per capital 0.0153** (0.0719)
Industry value-added 0.122*** (0.114)
Urbanization 0.169* (0.171)
Renewable energy consumption 0.0633*** (0.0402)
Constant 0.906*** (0.350) 0.0288 (0.148)
Observations 360 360
Number of countries 20 20

Standard errors in parentheses.

***p < 0.01, **p < 0.05, *p < 0.1.

Generalized Method of Moments (S-GMM) by Arellano and Bond Finally, using robust econometric identification strategies such as
[44] to address omitted variable bias. As presented in Table 8, the the lag of variables and S-GMM, the result of this paper demon-
results show that our main finding that sustainable economic strates that empirical analysis has not faced any potential problems
development and financial development variables are significantly from endogeneity.
related to energy efficiency is robust. Following our findings, our study first recommends that coun-
tries in the Asian and Pacific regions should maintain higher energy
5. Conclusion and policy implications efficiency rates by promoting sustainable economic development.
The possible ways to improve sustainable economic development
We have examined the relationship between sustainable eco- are:
nomic developments, financial development energy efficiency.
Unlike prior studies on energy efficiency [22], we extended the 1) The tax and benefit system: Government should intervene by
scope by focusing on the countries with the highest renewable promoting equity to reduce inequality as well as poverty.
energy use and considering a specific type of innovation where Through the tax and benefit system, the government can apply a
government policies are relevant. Our data covered 20 countries of progressive tax that takes more from the higher income earners
Asia and the Pacific between 2000 and 2018, and the region has and redistributes for the welfare of lower-income earners. The
been recognized among the top energy-intensive economies. Given establishment of National Insurance Funds will help map out
this fact, we introduced green innovation to ascertain the effort of and distribute contributory benefits such as pensions and job-
these countries in maintaining a green environment. Green inno- seekers allowance.
vation is measured as the number of patents in environmental- 2) Industrialization: Government should initiate a process to-
related technologies and is sourced from OECD Statistics (2021). wards industrialization. Such a process includes public-private
Data on energy efficiency was calculated using DEA methods, while partnerships, technology and innovation, and creating a
sustainable economic and financial development variables were comparative advantage, where both parties invest in the
collected from the World Bank and the IMF Database. industry's setup and operation. More so, technologies and
In the light of this, we employed robust econometric modeling, innovation funds initiatives for industrialization should be
which shows that sustainable economic development significantly established to flourish innovation and foster speedy industri-
drives energy efficiency in Asia and the Pacific regions. More spe- alization. While creating a comparative advantage will help
cifically, we found that investment for sustainable economic resource utilization and distribution.
development is an option to increase energy efficiency. The analogy 3) Promotion of public utilities: Countries can also improve
here is that every individual and organization is well better off economic development by increasing spending on public goods
under a well-developed economic system. Given this, the energy such as education, public transport, and healthcare. Therefore, a
market becomes competitive, where individuals or organizations free public utility scheme should be put in place to improve the
buy energy at a very competitive price, leading to energy efficiency. welfare of lower-income earners and would be fund by the
Similarly, we found that sustainable financial development pro- progressive tax.
motes energy efficiency; this relies upon the assertion that financial
institutions such as banks can fund businesses given sustainable Secondly, free-market economy (supply side) policies are also
financial development. Therefore, stakeholders in the energy sec- instruments for sustainable financial development. The possible
tors can purchase energy and sell at a cheaper rate, leading to en- way to improve sustained financial systems are:
ergy efficiency.
We also tested the green innovation view theory and the rela- 1) Privatization of state-owned financial institutions: The
tionship between energy efficiency. According to this theory, green public-owned financial institutions could be offered for privat-
innovation may pressure the government and companies to deviate ization by initiating a public enterprise where the higher bidder
from using non-environmental friendly energy. Our results indicate and the most expertise are offered. This is critical in improving
that green innovation is positively related to energy efficiency, financial system performance because it is associated with
suggesting that our findings are sensitive to green innovation. lower costs, lower prices, and higher operating efficiency.

9
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

2) Promotion of foreign direct investment: Set up an Investment emissions from two sources in the ECOWAS region: new insights from
quantile regression and decomposition analysis. Sci Total Environ 2020;748:
Promotion Agency (IPA) to target suitable foreign investors and
141329.
become the link between the investor and the financial insti- [5] Shrestha P. Global energy use projected to nearly double by 2050. https://
tution. On the one side, it should act as a one-stop shop for re- www.energylivenews.com/2020/01/08/global-energy-use-projected-to-
quirements investors demand from the host country. On the nearly-double-:&#x223C;:text¼Energy%20use%20in%20the%20global,British
%20thermal%20units%20(Btu); 2020.
other side, it should serve as a catalyst in the host's financial [6] Fankhauser S, Jotzo F. Economic growth and development with low-carbon
institution, prompting it to provide top-notch infrastructure and energy. Wiley Interdisciplinary Reviews: Climate Change 2018;9(1):e495.
ready access to skilled workers and managers that may be [7] Sadorsky P. The impact of financial development on energy consumption in
emerging economies. Energy Pol 2010;38(5):2528e35.
required to attract such investors. [8] Horoszowski M. The full list of the 17 united nations sustainable development
goals (with pictures) e #SDGs. https://blog.movingworlds.org/the-full-list-of-
Despite the robustness of our findings, this study is not without the-17-united-nations-sustainable-development-goals-with-picturessdgs/?
gclid¼Cj0KCQjwmIuDBhD_XARIsAFITC_4xAGvU-jKiJ20yiKa90H9c8u-
limitations. The major setback to this study is the non-availability of Qa7YIbnxQk27vNQvb55tK825SjL4aAlHjE_ALw_wcB; 2015.
data to cover a wide range of Asian and Pacific countries. [9] Hussaini IU. Households’ energy efficiency practices in a bereft power supply
economy of Nigeria. In: Energy-Efficient Approaches in Industrial Applica-
tions. IntechOpen; 2018.
Credit author statement [10] Lan Z, Zhao Y, Zhang J, Jiao R, Khan MN, Sial TA, Si B. Long-term vegetation
restoration increases deep soil carbon storage in the Northern Loess Plateau.
Sci Rep 2021;11(1):1e11.
Abdulrasheed Zakari: Writing - original draft, Review & Edit- [11] Davidson PJ, Henderson G. Improving energy efficiency in housing. 1989.
ing, Data curation, Irfan Khan: Review and editing Duojiao Tan: https://www.osti.gov/etdeweb/biblio/6953018.
Supervision, Validation, Rafael Alvarado: Conceptualization; [12] Environmental and Energy Study Institute. The Role of International Climate.
Finance 2021. https://www.eesi.org/briefings/view/102021cop.
Writing-review & editing, Vishal Dagar: Data curation, [13] Ganda F, Ngwakwe CC. Role of energy efficiency on sustainable development.
Investigation. Environ Econ 2014;5(1):86e99.
[14] Türkog lu SP, Kardog € The role and importance of energy efficiency for
an PSO.
sustainable development of the countries. March. In: International sustainable
Declaration of competing interest buildings Symposium. Cham: Springer; 2017. p. 53e60.
[15] Chang S. Effects of financial developments and income on energy consump-
tion. Int Rev Econ Finance 2015;35:28e44.
The authors declare that they have no known competing [16] Ahmed K. Revisiting the role of financial development for energy-growth-
financial interests or personal relationships that could have trade nexus in BRICS economies. Energy 2017;128:487e95.
[17] Li K, Fang L, He L. How urbanization affects China's energy efficiency: a spatial
appeared to influence the work reported in this paper.
econometric analysis. J Clean Prod 2018;200:1130e41.
[18] Sheng P, He Y, Guo X. The impact of urbanization on energy consumption and
efficiency. Energy Environ 2017;28(7):673e86.
Appendix. List of countries [19] Consultancyasia. Asia's top renewable energy markets: China, India, and
Japan. https://www.consultancy.asia/news/3325/asias-top-renewable-
energy-markets-china-india-and-japan; 2020.
[20] Nie PY, Yang YC, Chen YH, Wang ZH. How to subsidize energy efficiency under
duopoly efficiently. Appl Energy 2016;175:31e9.
[21] Jiang L, Zhou H, He S. Does energy efficiency increase at the expense of output
performance: evidence from manufacturing firms in Jiangsu province, China.
Energy 2021;220:119704.
Country Mean
[22] Ziolo M, Jednak S, Savi c G, Kragulj D. Link between energy efficiency and
Australia 5.040571 sustainable economic and financial development in OECD countries. Energies
Bangladesh 4.957529 2020;13(22):5898.
Bhutan 2.97368 [23] Jiang Q, Yang S, Tang P, Bao L. Promoting the polluters? The Competing ob-
Brunei Darussalam 5.19479 jectives of energy efficiency, pollutant emissions, and economic performance
in Chinese municipalities. Energy Res. Soc. Sci. 2020;61:101365.
China 3.54797
[24] Ma X, Cai L, Chen L, Fei B, Lu J, Xia C, Lam SS. Bamboo grid versus polyvinyl
Hong Kong SAR, China 55510.62
chloride as packing material in cooling tower: energy efficiency and envi-
India 4.090005 ronmental impact assessment. J Environ Manag 2021;286:112190.
Indonesia 2.523287 [25] Wang H, Han Y, Fidrmuc J, Wei D. Confucius Institute, belt and road initiative,
Japan 5.151099 and internationalization. Int Rev Econ Finance 2021;71:237e56.
Korea, Rep. 4.0198 [26] Oyedepo SO. Energy and sustainable development in Nigeria: the way for-
Malaysia 5.200626 ward. Energy Sustain. Soc. 2012;2(1):1e17.
Maldives 16027.55 [27] Spalding-Fecher R, Winkler H, Mwakasonda S. Energy and the world summit
Myanmar 646.4364 on sustainable development: what next? Energy Pol 2005;33(1):99e112.
Nepal 51.56566 [28] Oryani B, Koo Y, Rezania S, Shafiee A. Investigating the asymmetric impact of
New Zealand 5.54394 energy consumption on reshaping future energy policy and economic growth
Papua New Guinea 5825.082 in Iran using extended Cobb-Douglas production function. Energy 2021a;216:
Philippines 6.146061 119187.
[29] Han Y, Zhang F, Huang L, Peng K, Wang X. Does industrial upgrading promote
Singapore 4.902039
eco-efficiency?e A panel space estimation based on Chinese evidence. Energy
Sri Lanka 4.396148
Pol 2021;154:112286.
Thailand 5.085368
[30] Rosen MA. Energy efficiency and sustainable development. Int J Global Energy
Issues 2002;17(1e2):23e34.
[31] Mills EFEA, Dong J, Yiling L, Baafi MA, Li B, Zeng K. Towards sustainable
competitiveness: how does financial development affect dynamic energy ef-
References ficiency in Belt & Road economies? Sustain. Prod. Consum. 2021;27:587e601.
[32] Zhang XP, Cheng XM, Yuan JH, Gao XJ. Total-factor energy efficiency in
[1] Balezentis T. Shrinking ageing population and other drivers of energy con- developing countries. Energy Pol 2011;39(2):644e50.
sumption and CO2 emission in the residential sector: a case from Eastern [33] Coelli TJ. A guide to FRONTIER version 4.1: a computer program for stochastic
Europe. Energy Pol 2020;140:111433. frontier production and cost function estimation. In: CEPA Working papers, 7;
[2] De Souza Mendonça AK, Barni GDAC, Moro MF, Bornia AC, Kupek E, 1996. p. 1e33.
Fernandes L. Hierarchical modeling of the 50 largest economies to verify the [34] Wu D, Wang R. Combined cooling, heating and power: A review progress in
impact of GDP, population, and renewable energy generation in CO2 emis- energy and combustion. Science 2006;32(5e6):459e95.
sions. Sustain. Prod. Consum. 2020;22:58e67. [35] Li G, Zakari A, Tawiah V. Does environmental diplomacy reduce CO2 emis-
[3] Bullard III CW, Herendeen RA. The energy cost of goods and services. Energy sions? A panel group means analysis. Sci Total Environ 2020a;722:137790.
Pol 1975;3(4):268e78. [36] Li G, Zakari A, Tawiah V. Energy resource melioration and CO2 emissions in
[4] Nwaka ID, Nwogu MU, Uma KE, Ike GN. Agricultural production and CO2 China and Nigeria: efficiency and trade perspectives. Resour Pol 2020b;68:

10
A. Zakari, I. Khan, D. Tan et al. Energy 239 (2022) 122365

101769. Africa. Technol Forecast Soc Change 2019;139:169e84.

[37] Wu B, Fang H, Jacoby G, Li G, Wu Z. Environmental regulations and innovation [47] Tchamyou VS, Asongu SA. Information sharing and financial sector develop-
for sustainability? Moderating effect of political connections. Emerg Mark Rev ment in Africa. J Afr Bus 2017;18(1):24e49.
2021:100835. [48] Razzaq A, Sharif A, Najmi A, Tseng ML, Lim MK. Dynamic and causality in-
[38] Wang F, Lu Y, Li J, Ni J. Evaluating environmentally sustainable development terrelationships from municipal solid waste recycling to economic growth,
based on the PSR framework and variable weigh analytic hierarchy process. carbon emissions and energy efficiency using a novel bootstrapping autore-
Int J Environ Res Publ Health 2021b;18(6):2836. gressive distributed lag. Resour Conserv Recycl 2021;166:105372.
[39] Fan P, Deng R, Qiu J, Zhao Z, Wu S. Well logging curve reconstruction based on [49] Liimatainen H, Po€ lla
€nen M. The impact of sectoral economic development on
kernel ridge regression. Arab. J. Geosci. 2021;14(16):1e10. the energy efficiency and CO2 emissions of road freight transport. Transport
[40] Zhang K, Jia C, Song Y, Jiang S, Jiang Z, Wen M, Liu T. Analysis of Lower Pol 2013;27:150e7.
Cambrian shale gas composition, source and accumulation pattern in different [50] Lv Y, Chen W, Cheng J. Effects of urbanization on energy efficiency in China:
tectonic backgrounds: a case study of Weiyuan Block in the Upper Yangtze new evidence from short run and long run efficiency models. Energy Pol
region and Xiuwu Basin in the Lower Yangtze region. Fuel 2020b;263:115978. 2020;147:111858.
[41] Han X, Wei Z, Zhang B, Li Y, Du T, Chen H. Crop evapotranspiration prediction [51] Oryani B, Koo Y, Rezania S, Shafiee A, Khan MK, Mahdavian SM. The role of
by considering dynamic change of crop coefficient and the precipitation effect electricity mix and transportation sector in designing a green-growth strategy
in back-propagation neural network model. J Hydrol 2021b;596:126104. in Iran. Energy 2021:121178.
[42] Harris RD, Tzavalis E. Inference for unit roots in dynamic panels where the [52] Mikayilov JI, Mukhtarov S, Dinçer H, Yüksel S, Aydın R. Elasticity analysis of
time dimension is fixed. J Econom 1999;91(2):201e26. fossil energy sources for sustainable economies: A case of gasoline con-
[43] Beck N, Katz JN. What to do (and not to do) with time-series cross-section sumption in Turkey. Energies 2020;13(3):731.
data. Am Polit Sci Rev 1995;89(3):634e47. [53] Wang H, Wang M. Effects of technological innovation on energy efficiency in
[44] Arellano M, Bond S. Some tests of specification for panel data: Monte Carlo China: evidence from dynamic panel of 284 cities. Sci Total Environ 2020;709:
evidence and an application to employment equations. Rev Econ Stud 136172.
1991;58(2):277e97. [54] Sun H, Edziah BK, Sun C, Kporsu AK. Institutional quality, green innovation,
[45] Arellano M, Bover O. Another look at the instrumental variable estimation of and energy efficiency. Energy Pol 2019;135:111002.
error-components models. J Econom 1995;68(1):29e51. [55] Wang N, Shen R, Wen Z, De Clercq D. Life cycle energy efficiency evaluation
[46] Tchamyou VS, Erreygers G, Cassimon D. Inequality, ICT and financial access in for coal development and utilization. Energy 2019;179:1e11.

11