How do high-tech industry,energy efficiency (gas and oil),and social globalization affect load capacity factor in Saudi Arabia? A Time-varying quantile regression

Abstract

The significant allure of widely available fossil energy sources at a reasonable cost engenders formidable impediments to the transition toward renewable energy sources in Saudi Arabia, one of the world's foremost energy-producing nations. Mitigating the escalating levels of emissions and addressing the substantial ecological deficits requires a comprehensive investigation of the prospective contributions of energy efficiency and high-tech industry as integral components in the strategic response to environmental challenges. In light of this, our current study employs quarterly data spanning from 1990 to 2021 and introduces an innovative approach known as time-varying quantile regression (TVQR), which has not been previously utilized in the existing literature. We complement this with time-varying nonparametric quantile causality to assess the impacts of disaggregated energy efficiency, the high-tech industry, and social globalization on ecological quality (proxied by load capacity factor (LF)) across different quantiles and time intervals in Saudi Arabia. Our TVQR analysis reveals that both the high-tech industry and oil and gas efficiency have a positive impact on environmental quality, as evidenced by their ability to enhance the LF indicator across a significant portion of quantiles and time intervals. Conversely, economic growth and social globalization have a negative impact on environmental quality. Additionally, the selected explanatory variables exhibit significant predictive power over LF across various time frames and quantiles. Lastly, we have formulated a comprehensive policy framework aimed at enhancing ecological quality.

Keywords

Energy efficiency high-tech industry load capacity factor social globalisation quantile regression

Introduction

In the contemporary era, nations grapple with a myriad of economic and environmental challenges, necessitating innovative approaches to sustain economic growth (EG), enhance public health, and promote ecological sustainability. Foremost among these challenges is the enduring escalation of global warming and its associated ramifications on climate patterns, precipitating an array of severe meteorological phenomena and calamities, resulting in substantial human casualties and economic losses.¹ The sustained escalation in economic endeavors has engendered heightened levels of resource consumption and production, thereby exacerbating energy utilization and waste generation.^2,3 Consequently, carbon emissions, recognized as the primary catalysts of global warming, have exhibited a twofold surge since 1970.⁴ Notably, approximately 75% of worldwide carbon emissions emanate from the energy sector, underscoring the pivotal role that energy-related policies can play in mitigating climate change and ameliorating environmental conditions.

In light of the fact that many nations predominantly depend on the utilization of petroleum products to foster and maintain economic development, the associated increase in fossil energy consumption, coupled with development efforts, has led to a concurrent escalation in pollutant emissions, resulting in a noticeable deterioration of ecological quality.⁵ While scholars have posited the adoption of green energy as a viable solution to mitigate dependence on fossil fuel use, the persistently alluring attributes of fossil fuels in numerous countries, including their comparatively lower initial costs, the absence of mature green energy production technologies, the limited availability of green transportation alternatives, and the better accessibility of fossil energy, continue to present formidable barriers to the widespread adoption of renewables. Notably, such appealing characteristics of fossil energy sources are pronounced in countries such as Saudi Arabia, a prominent member of the largest global oil-producing nations. Hence, the importance of energy efficiency underscores its pivotal role in mitigating resource depletion, as it fundamentally rests upon the principle of achieving equivalent output while curbing energy input. Augmentations in energy efficiency not only curtail reliance on fossil fuels but also confer an array of ancillary advantages, encompassing the alleviation of electricity deficits, diminution of operational expenditures, bolstering industrial competitiveness, promoting the underpinning of energy security, and boosting ecological quality.^6,7 The augmentation of energy efficiency mandates the utilization of advanced technologies and the implementation of sustainable practices, as highlighted by Jin et al.^8,9 These include various measures including the deployment of energy-efficient household appliances, the establishment of intelligent grid infrastructure, the adoption of industrially efficient machinery and equipment, and the construction of environmentally friendly infrastructures. Collectively, such endeavors hold the potential to catalyze economic development, curtail energy consumption and emissions, and advance ecological sustainability.

This research evaluates the impacts of oil efficiency (OEF) and coal efficiency, high-tech industry (HTI), and social globalization (SGLO) on environmental sustainability. In the endeavor to diminish energy consumption, a paramount imperative lies in directing attention toward industrial restructuring, with an emphasis on the promotion of HTIs. This strategic orientation is underpinned by the capability of high-tech sectors to reduce energy utilization, thereby conferring a beneficial impact upon ecological quality.^10,11 HTIs predominantly leverage contemporary technological advancements, innovation, rigorous research, and automation to fabricate complex products, implement novel organizational methodologies, embrace advanced production techniques, and optimize supply chain operations. Consequently, this multifaceted approach can yield reductions in energy consumption, waste generation, and the generation of emissions, thereby contributing to environmental sustainability.^12,13

In the examination of the ramifications of energy efficiency and the HTI on environmental quality, due consideration is accorded to the pivotal role of SGLO. SGLO constitutes a significant channel for the dissemination of information, knowledge, and awareness pertaining to both the advantageous and detrimental facets of consumption and production. This function is realized through the facilitation of cross-border connections among individuals from diverse countries and regions, as well as the encouragement of international media and internet penetration and heightened international mobility.¹⁴ According to Ahmed et al.,¹⁵ SGLO assumes a pivotal role in the advancement of environmental consciousness, thereby inciting practices conducive to energy conservation, recycling, water preservation, and the adoption of cleaner product choices. Consequently, these affirmative consequences of SGLO have the potential to foster ecological sustainability, enabling societies to assimilate environmental awareness, and embrace eco-friendly practices that are prevalent in other nations.

In the context of comprehending the influence of energy efficiency, the HTI, and SGLO on ecological quality, it becomes imperative to employ a comprehensive metric of environmental sustainability capable of holistically encapsulating environmental degradation while duly considering the availability of resources, as indicated by biocapacity. In this context, it is noteworthy that conventional metrics, such as CO₂ emissions and ecological footprint (EF), have certain critical limitations. For instance, CO₂ emissions primarily gauge energy utilization, while EF fails to adequately account for the essential dimension of biocapacity^16,17 posit that the omission of biocapacity, which reflects the Earth's capacity to produce resources, results in a diminishment of both the comprehensiveness and precision of ecological quality assessments. Thus, in alignment with contemporary research studies,^18,19 this investigation uses the load capacity factor (LF) to measure ecological quality, as this indicator (i.e. biocapacity/EF) utilizes the data of both biocapacity and EF and represents a holistic proxy of environmental sustainability.

This empirical study opted to include Saudi Arabia as a research sample for several compelling rationales. Saudi Arabia stands as a notable case due to its pronounced reliance on the energy sector, positioning itself among the foremost global energy producers.²⁰ Saudi Arabia possesses significant energy assets, including substantial reserves of oil and natural gas. Approximately 90% of its export revenue is derived from petroleum products, and the oil sector alone constitutes a substantial 45% share of the nation's gross domestic product.²¹ As petroleum products are affordable and easily available in the country, the Saudi economy predominantly relies on oil and gas. Saudi Arabia was using 19615.88 (per capita kWh) of oil and 37569.61 (per capita kWh) of gas in 1990. However, escalated EG bolstered the consumption of oil and gas in Saudi Arabia, recording an increase of 69% and 38% in oil and gas, respectively, from 1990 to 2021.²² As a result, carbon emissions have experienced an escalation, warranting the inclusion of the nation within the ranks of the top 10 global carbon emitters.²³ Saudi Arabia had a biocapacity of 1.32 (gha per capita) and EF of 2.40 (gha per capita) in 1990, with an LF of 0.55, implying that the biocapacity of Saudi Arabia was enough to support only 45% of its EF. Nevertheless, persistent environmental degradation coupled with high consumption of energy and other resources expanded the EF (5.58) and reduced biocapacity (0.71), and the LF reached the level of only 0.13 in 2021.^24,25

As the LF values of 1 or above depict ecological sustainability (see Figure 1), the persistent reduction in LF in Saudi Arabia shows high ecological degradation and unsustainability in the production and consumption of resources. In the particular context of Saudi Arabia, the imperative of augmenting the energy efficiency within the oil and gas sectors assumes paramount significance, as it serves as a pivotal strategy for mitigating the substantial dependence on fossil energy sources, ameliorating environmental degradation, and steering toward the attainment of a sustainable environment. Under the Saudi Vision 2030, the country is looking to diversify its economy and increase its competitiveness to become a powerhouse of global investment. Furthermore, this endeavor encompasses a multifaceted array of initiatives aimed at curtailing environmental degradation. Consequently, an economic restructuring approach, coupled with a transition towards HTIs, might be advantageous for diminishing reliance on oil-based resources and enhancing energy efficiency, ultimately facilitating the realization of environmental sustainability objectives. Moreover, SGLO may elevate the level of awareness and sensitiveness within Saudi Arabia concerning the imperative of curbing energy consumption in the pursuit of fostering a healthy and sustainable environment.

Figure 1.

Ecological sustainability.

Against this background, this research investigates the effects of OEF and gas efficiency (GASEF), HTI, and SGLO on the LF. This is the first empirical investigation that assesses the impacts of these important variables on LF. Therefore, this study makes a considerable contribution to previous scholarly works. Secondly, this research introduces a novel concept of time-varying quantile regression (TVQR), which has never been used before in the available literature. TVQR will enable us to analyze the impacts of disaggregated energy efficiency, HTI, and SGLO on different quantiles of the distribution of the response variable across various subperiods. Therefore, this methodology offers supplementary insights that can be instrumental in formulating apposite environmental policies. Moreover, the adoption of the LF as an environmental quality metric serves to differentiate this empirical inquiry from prior research endeavors. This distinction arises from LF's encompassing nature, which takes into account both demand and supply facets, thus providing an inclusive evaluation of ecological quality.

Theoretical framework and literature

Theoretical framework

Theoretical frameworks concerning the impact of the HTI, GASEF, OEF, and SGLO on various aspects of society are crucial for understanding their complex interactions and implications.

The HTI, known for its swift innovation and technological progress, exerts profound transformations on economies, societies, and the environment. From a macroeconomic viewpoint, this sector fuels economic expansion by boosting productivity, fostering job opportunities, and enhancing competitiveness.^12,26 Nevertheless, its effects are not evenly distributed across society, giving rise to apprehensions regarding digital disparities and societal inequality. Furthermore, the high-tech sector's reliance on finite resources and its environmental impact present sustainability hurdles, underscoring the need for policies aimed at advancing green technologies and embracing principles of the circular economy.²⁷

İncreased GASEF and OEF plays a pivotal role in mitigating ecological harm, curbing greenhouse gas emissions, and bolstering energy stability. Economically, advancements in GASEF and OEF can yield cost decreases, heighten competitiveness, and diminish reliance on fossil fuel imports.²⁸ Nevertheless, realizing efficiency improvements demands investments in R&D, infrastructure enhancements, and the establishment of regulatory frameworks. Additionally, the shift toward RE alternatives and the electrification of transportation present hurdles for existing industries, mandating workforce retraining initiatives and transition strategies.²⁹

SGLO, marked by the interconnectivity of cultures, societies, and individuals, holds significant ramifications for values, identity, and social unity. It fosters the flow of knowledge, ideas, and cultural norms, fostering diversity and enriching societies.^30,31 However, SGLO also sparks tensions and obstacles, such as cultural assimilation, identity societal fragmentation and dilemmas. Addressing these issues necessitates policy initiatives that endorse cultural plurality, hearten intercultural discourse, and enact inclusive social strategies.³²

Summary of past studies

In pursuit of achieving their individual carbon neutrality objectives, numerous nations have formulated diverse strategies aimed at reaching this compelling goal. Additionally, the adverse impacts of climate change and global warming have compelled several nations to transition from pro-growth approaches to more environmentally friendly ones. Consequently, numerous research initiatives have been launched to investigate the influence of economic variables on ecological quality. These endeavors are intended to assist governments and various policymakers in crafting precise and effective sustainable development policies.

When it comes to examining the connection between the HTI and ecological quality or degradation, there is a limited body of empirical literature available. For instance, Xu and Lin^10,11 conducted a panel analysis using regional data from China to explore the impact of the HTI on reducing ecological degradation over the period from 1999 to 2015. Their study revealed that the decline in CO₂ in the selected regions could be attributed to the growth of the HTI. With the motif of devising SDGs measures for China, Shahzad et al.^12,13 utilized a dynamic estimator to research the implication of the HTI on ecological integrity, with the results indicating that growth in ecological excellence is accredited to the intensification of the HTI. A similar angle is viewed by the investigation of Xu and Lin^10,11 and Du et al.,³³ who, in their respective studies, highlighted that ecological deterioration is attributed to the initiatives put in place to promote HTI; though, these initiatives damage the ecosystem. On the contrary, the investigation put forward by Rafique et al.²⁶ showcased that the policies regarding HTI are sustainable since they lessen ecological damage by intensifying ecological integrity.

SGLO showcases the growing integration and linkage of various societies as well as cultures that expound beyond their domestic borders. This occurrence is triggered by factors such as global travel, trade, technological progress, and communication. Nonetheless, these drivers can have both detrimental and enhancing impacts on the ecosystem. For illustration, Shi et al.,³⁴ with the goal of proposing SDGs measures, used the N-11 nations in analyzing the nexus between SGLO and ecological integrity. The results, with the aid of the Method Of Moment Quantile Regressions (MMQR) estimator, showcased that the green environment in the selected nations is accredited to the growth of SGLO. Conversely, Wang et al.,³⁵ study contradicts the findings put forward by Shi et al.,³⁴ by showcasing the effect of SGLO in promoting the ecological excellence of China. Moreover, Kirikkaleli and Adebayo's³⁶ analysis using the Brazilian case affirmed that ecological excellence bolsters the role of SGLO. Similarly, Sharif et al.^31,37 investigated the role of SGLO in promoting ecological quality, using the G7 countries as a case study. Their study highlighted the significant positive impact of SGLO on ecological quality. These studies collectively demonstrate the complexity of the interrelationship between SGLO and the environment, and they pave the way for further research as concrete results regarding the SGLO-environment nexus remain elusive.

Recent empirical literature has increasingly emphasized the role of energy efficiency (EF) in bolstering ecological integrity. The interrelationship between ecological degradation and energy efficiency is of paramount importance in the context of ecological sustainability and the global effort to combat climate change. These two factors often exhibit an inverse correlation, meaning that as energy efficiency improves, ecological degradation tends to decrease. Conversely, a decrease in energy efficiency is allied with an increase in CO₂. For instance, Liu et al.,³⁸ in their study focusing on the role of energy efficiency, specifically coal efficiency, on CO₂ in the USA, utilized Wavelet tools to underscore the positive impact of coal efficiency in enhancing ecological integrity. Similarly, Alola et al.,³⁹ in their research examining the Indian context, and Jin et al.,^8,9 in their study of Germany, reported the positive influence of EF on ecological excellence. Furthermore, studies conducted by Akram et al.,^6,7 for the BRICS and Naimoglu and Akal⁴⁰ for Turkey highlighted the emissions-reducing effect of energy efficiency. Additionally, Zhang et al.,⁴¹ using data from China's 30 provinces spanning from 2012 to 2019 and employing panel threshold regression, emphasized the constructive role of energy efficiency in strengthening ecological integrity. Table 1 offers a comprehensive summary of the reviewed studies.

Table 1.

Glimpse of the literature.

Author(s)	Study period	Nation (s)	Technique (s)	Findings
Xu and Lin^10,11	1999–2015	Regions in China	Panel Analysis	HTI→CO₂ (−)
Shahzad et al.^12,13	1996−2019	China	Dynamic Connectedness	HTI→CO₂ (+)
Rafique et al.²⁶	1996–2019	76 BRI	SGMM and 2-SGMM	HTI→CO₂ (−)
Du et al.,³³	1990–2018	MINTS	MMQR	HTI→EF (+)
Shi et al.³⁴	2000−2019	N-11 nations	CS-ARDL	SGLO→CO₂ (+)
Wang et al.³⁵	1970−2021	China	ARDL	SGLO→CO₂ (−)
Kirikkaleli and Adebayo³⁶	2000−2020	Brazil	Dynamic ARDL	SGLO→CO₂ (−)
Sharif et al.^31,37	1995−2019	G7 countries	CS-ARDL	SGLO→CO₂ (−)
Adebayo and Ullah⁴²	1990−2020	Sweden	Wavelet Tools	OEF→CO₂ (−)
Liu et al.³⁸	1990−2020	USA	Wavelet Tools	COEF→CO₂ (−)
Alola et al.³⁹	1990−2020	India	Dynamic ARDL	EF→CO₂ (−)
Jin et al.^8,9	1974−2018	Germany	Augmented ARDL	EF→CO₂ (−)
Akram et al.^6,7	1990−2019	BRICS	Panel ARDL	EF→CO₂ (−)
Naimoglu & Akal⁴⁰	1974−2018	Turkiye	Augmented ARDL	EF→CO₂ (−)
Zhang et al.⁴¹	2012−2019	China	Panel threshold regression	EF→CO₂ (−)
Olanrewaju et al.⁴³	1990−2020	G7 nations	MMQR	EG→CO₂ (+)
Çitil et al.²	1990−2018	G20 nations	MMQR	EG→CO₂ (+)
Ojekemi et al.^{44, 45}	1990−2018	BRICS nations	AMG, CCEMG	EG→CO₂ (+)
	1990−2018	Transition economies	CS-ARDL and MMQR	EG→CO₂ (+)
Oyebanji et al.⁴⁶	1990−2018	Finland	ARDL	EG→LF(−)
Adebayo⁴⁷	1990−2019	Spain	Wavelet	EG→LF (−)

AMG: augmented mean group; ARDL: Auto Regressive Distributed Lag; CO₂: carbon emissions; CCEMG: common correlated mean group; COEF: coal efficiency; EFC: energy efficiency; EF: ecological footprint; EG: economic growth; HTI: high-tech industry; MINTS: Mexico, Indonesia, Nigeria, Turkey South Africa; MMQR: method of moment quantile regressions; OEF: oil efficiency; SGLO: social globalisation; SGMM: system generalized method of moments.

Gap in the literature

The literature review reveals a growing but still limited body of research focused on assessing the influence of globalization, EG, HTI, and energy efficiency on CO₂ emissions and ecological footprint. These studies primarily target individual countries (e.g. the USA, Germany, France, and China) or groups of nations (e.g. BRICS, Mexico, Indonesia, Nigeria, Turkey South Africa, G20, ASEAN, G7). Various econometric techniques, including CCR, Granger causality, DYNARDL, NARDL, FMOLS, MMQR, DOLS, Dumitrescu-Hurlin panel causality test, GMM, and DOLS, are employed for empirical analysis. This approach represents a significant advancement beyond the traditional quantile regression proposed by Koenker and Bassett.⁴⁸ Notably, our study introduces the innovative concept of TVQR for the first time in empirical analysis. TVQR allows us to examine the impact of independent variables on different quantiles of the distribution of the dependent variable across various subperiods. To the best of our knowledge, no existing literature examines the effects of disaggregated energy efficiency (specifically, OEF and GASEF) and HTI on LF, which serves as a detailed proxy for ecological quality encompassing both the demand and supply sides of the environment. In addition to addressing this research gap, our study pioneers the use of TVQR to explore the effects of disaggregated energy efficiency (OEF and GASEF), HTI, and SGLO on LF in the case of Saudi Arabia. This innovative method enriches our comprehension of the complex interplay between these variables and their impact on ecological quality across various quantiles and time intervals.

Data and methods

Data

In this article, Saudi Arabian data for the period from 1990 to 2021, which are the longest periods available for the variables, are employed. As a proxy for environmental quality, which is our dependent variable, we use the LF. The LF represents a precise ecological threshold by combining biocapacity and ecological footprint. This combination makes the LF a more comprehensive environmental quality measure than traditional carbon dioxide and ecological footprint measures, which do not take into consideration the environmental issues’ supply side.⁴⁹ Due to these important features, the LF is preferred as an indicator of environmental quality in this study. On the other hand, our explanatory variables are HTI, GASEF, OEF, SGLO, and EG. Table 2 presents the sources from which the variables are obtained and the measures used to measure them. Furthermore, Figure 2 illustrates the time plots of the annual raw values of the study variables. From the figure, it is seen that LF, GASEF, and OEF followed a decreasing trend during the sample period, while SGLO showed an increasing trend. Moreover, EG decreased until 2002 but has started to increase since this year. Lastly, although HTI increased significantly in 2001, it has been stagnating recently.

Figure 2.

Time plots of the variables’ annual values from 1990 to 2021.

Table 2.

Data sources and measurement.

Sign	Variables	Measurement	Source
LF	Load capacity factor	Biocapacity/ecological footprint	GFN [24,25]
HTI	High tech industry	Medium and high-tech manufacturing value added (% manufacturing value added)	WB [51]
GASEF	Gas efficiency	GDP, PPP (constant 2017 international $)/gas energy in tonnes	Authors Calculation
EG	Economic growth	GDP per capita (constant 2015 US$)	WB [51]
OEF	Oil efficiency	GDP, PPP (constant 2017 international $)/oil energy in tonnes	Authors Calculation
SGLO	Social globalisation	Index	KOF [52]

GDP: gross domestic product; PPP: purchasing power parity.

In order to reach a sufficient observation size, we first take the natural logarithm of the annual data in order to encounter any potential heteroscedasticity and then convert logarithmic values into quarterly series by applying the quadratic match-sum approach, as in the studies of Alola et al.⁵⁰ and Ozkan et al.⁵³

Methods

Time-varying quantile regression

Quantile regression (QR), introduced by Koenker and Bassett⁴⁸ and developed by Koenker,⁵⁴ is a statistical technique that extends the concept of linear regression by focusing on estimating conditional quantiles of a dependent variable rather than just the mean or expected value. The QR method particularly allows practitioners to see how the dynamics of the relationship between variables across distinct quantiles of the conditional distribution of the dependent variable change. In other words, the QR method primarily focuses on estimating the impact of an explanatory variable on the conditional quantiles of a dependent variable.⁵⁵

The QR model can be described as follows:

θ_{q} (Y | X) = β_{0 q} + X β_{1 q}

(1)where

θ_{q} (Y | X)

is the

q

-th conditional quantile of the dependent variable Y given the explanatory variable X. Moreover,

β_{0 q}

and

β_{1 q}

represent the intercept (constant) and the slope coefficient for the

q

-th quantile, respectively.

In the model presented in equation (1), $β_{0 q}$ and $β_{1 q}$ are the parameters to be estimated for each quantile $τ$ , and these parameters describe how the conditional quantile $θ_{q} (Y | X)$ changes with respect to changes in the explanatory variable X. By doing this, the QR method captures how the relationship between the variables varies across distinct quantiles of the conditional distribution of Y.

Unlike ordinary least squares (OLS) regression, which reduces the sum of squared residuals, QR minimizes a different objective function that involves the sum of absolute deviations. Mathematically, the estimation requires solving the following optimization problem for each quantile $q$ :

\min β_{q} \sum_{i = 1}^{n} p_{q} (y_{i} - X_{i} β_{q})

(2)where n is the observations number,

y_{i}

the observed value of the dependent variable for the

i

-th observation,

X_{i}

the vector of independent variables for the

i

-th observation, and

p_{q} (u)

the quantile loss function, defined as

p_{q} (u) = u (q - I (u < 0))

, where

I (u < 0)

is an indicator function that equals 1 if

u < 0

and 0 otherwise. The solution to this optimization problem provides the estimated coefficients

β_{q}

for each quantile q.

The estimated coefficients $β_{q}$ in QR can be interpreted similarly to OLS regression coefficients. However, instead of representing the change in the mean of Y for a one-unit change in X, estimated coefficients $β_{q}$ represent the change in the $q$ -th quantile of Y for a one-unit change in X. In other words, they describe how the quantiles of Y change with respect to changes in X.

QR has three important advantages over OLS regression: (1) QR is less sensitive to outliers than OLS regression because it minimizes the sum of absolute deviations; (2) QR provides a more complete image of the interrelationship between variables by estimating multiple quantiles, not just the mean; (3) QR can handle datasets with varying levels of dispersion or heteroscedasticity; and also, it is less sensitive to model misspecification. These advantages make QR a powerful tool for exploring and modeling the conditional distribution of a dependent variable.^54,56–58

Although the QR method has the important advantages mentioned above, it does not take into account the relationship between variables that may change over time. To overcome this very important shortcoming of the QR method, we combine traditional QR and rolling subsample windows approaches by following the ideas of.Lee et al.⁵⁹ and Olasehinde-Williams⁶⁰ Our new TVQR method allows us to analyze the time-varying impact of an explanatory variable on the conditional quantiles of a dependent variable.

The empirical application stages of the TVQR method are as follows:

Determine the quantiles (in this study, 0.1, 0.2, …, 0.9) and subsample windows size (in this study, 30).

Set the values of the explanatory and dependent variables for the first subsample window.

Estimate the quantile regression model presented in equation (5) to obtain the slope coefficient $β_{1 q}$ by solving the optimization problem given in equation (6) for each quantile $τ$ for the first subsample window.

Create a new subsample window by moving the window one observation forward, and repeat steps 2 and 3.

Repeat step 4 until the end of the sample period.

Time-varying nonparametric quantile causality

This study employs the time-varying nonparametric quantile causality (TVNQC) technique to examine the dynamic quantile causality between HTI, GASEF, OEF, SGLO, EG, and LF in Saudi Arabia. The nonparametric quantile causality (NQC) approach was introduced by Balcilar et al.,⁶¹ by extending the frameworks of Nishiyama et al.,^62,63 to analyze the nonlinear causal relationship between not only means but also variances of two variables by considering their conditional quantiles.

Balcilar et al.,⁶¹ proposed the following hypotheses to successively test the quantile causality in means and variances of the variables:

H_{0} : P {F_{y_{t}^{c} | Z_{t - 1}} {θ_{q} (Y_{t - 1}) | Z_{t - 1}} = q} = 1, c = 1 or 2

(3)

H_{1} : P {F_{y_{t}^{c} | Z_{t - 1}} {θ_{q} (Y_{t - 1}) | Z_{t - 1}} = q} < 1, c = 1 or 2

(4)where

y_{t}

represents a dependent variable,

Z_{t}

the vector constructed with the dependent and explanatory

x_{t}

variables, q the quantiles, and c the content of the causality analysis (e.g.

c = 1

for the quantile causality in mean and

c = 2

for the quantile causality in variance).

The hypotheses of the NQC method given in equations (3) and (4) can be tested with the feasible kernel-based test statistic for distance measure D calculated as follows:

{\hat{D}}_{T} = \frac{1}{T (T - 1) b^{2 τ}} \sum_{t = τ + 1}^{T} \sum_{δ = τ + 1, δ \neq t}^{T} F (\frac{Z_{t - 1} - Z_{δ - 1}}{b}) {\hat{ϵ}}_{t} {\hat{ϵ}}_{δ}

(5)where F(▪) presents the feasible kernel-based function, b the bandwidth,

T

the sample size,

τ

the lag order, and

{\hat{ε}}_{t}

the estimated unknown error defined as

{\hat{ε}}_{t} = 1 {x_{t} \leq {\hat{θ}}_{q} (X_{t - 1})} - q

, where

{\hat{θ}}_{q} (Y_{t - 1})

represents the forecast of the

q

-th conditional quantile of the dependent variable

y_{t}

Balcilar et al.,⁶¹ state that ${\hat{θ}}_{q} (Y_{t - 1})$ should be estimated by employing the nonparametric Nadarya-Watson kernel estimator, the computational details of which are shown below:

{\hat{J}}_{y_{t} | Y_{t - 1}} (y_{t} | Y_{t - 1}) = \frac{\sum_{δ = τ + 1, δ \neq t}^{T} W (\frac{Y_{t - 1} - Y_{δ - 1}}{b}) 1 {y_{δ} \leq y_{t}}}{\sum_{δ = τ + 1, δ \neq t}^{T} W (\frac{Y_{t - 1} - Y_{δ - 1}}{b})}

(6)where

W (\cdot)

is the kernel function.

The advantages of the NQC approach can be listed as follows: (1) NQC provides a more comprehensive vision about the causal impact of an explanatory variable on a dependent variable as it considers all joint distributions of the variables, not just their means; (2) NQC is less sensitive to misspecification errors because it takes into consideration underlying dependency among variables; (3) NQC allows practitioners to analyze quantile causality among the variables not only in mean but also in variance.⁶⁴

Since the NQC method is able to analyze quantile causality in both mean and variance between the variables by using the whole sample period, it inherently assumes that the relevant causalities do not change over time. However, as Granger (1996) points out, time series exhibit structural changes in certain periods, and these changes can cause shifts in the parameters, so the underlying type of causality among the variables can shift over time.⁶⁵ In order to overcome this shortcoming of the NQC method, Olasehinde-Williams et al.⁶⁰ developed the TVNQC technique by combining the NQC with the rolling subsample windows method. This novel approach allows practitioners to investigate time-varying nonlinear quantile causal relationships in the means and variances of an explanatory and a dependent variable.

In this study, we perform the empirical implementation of the TVNQC method in the following steps:

Determine the quantiles (in this study, 0.1, 0.2, …, 0.9) and size of the subsample windows (in this study, 30).

Set the values of the explanatory and dependent variables for the first subsample window.

Estimate the feasible kernel-based statistics presented in equation (3) to test the hypotheses given in equations (1) and (2) for the first subsample window.^a

Create a new subsample window by moving the window one observation forward, and repeat steps 2 and 3.

Repeat step 4 until the end of the sample period.

To provide a comprehensive overview of the study's analysis, we present the sequential flow of the techniques utilized (see Figure 3).

Figure 3.

Flow of analysis.

Findings and discussion

Analysis of statistical characteristics

Before delving into the analyses of TVQR and TVNQC, we first examine the statistical characteristics of the stationary series that will be employed in our analysis. In Table 3, we present the descriptive statistics for the variables under consideration. The mean values indicate that, over the sample period, the average monthly values of LF, HTI, GASEF, EG, OEF, and SGLO stand at −0.3936, 0.8508, 1.8959, 2.4545, 1.7509, and 1.0115, respectively. Regarding volatility, the standard deviation highlights that LF exhibits a notably higher level of volatility compared to all other variables, whereas EG shows relatively lower volatility. Examining skewness, we observe a positively skewed distribution for all the variables under investigation. Furthermore, the kurtosis values reveal that these variables exhibit platykurtic characteristics, as their values are less than 3. The Jarque-Bera (J-B) values provide insights into the distribution's normality, indicating that all series, except EG, depart from a normal distribution. These findings collectively suggest that methodologies considering the entire distribution, rather than focusing solely on the distribution's center, are more suitable for our analysis. Overall, these observations align with our rationale for employing quantile-based approaches.

Table 3.

Descriptive statistics.

	LF	HTI	GASEF	EG	OEF	SGLO
Mean	−0.3936	0.8508	1.8959	2.4545	1.7509	1.0115
Median	−0.3869	0.8813	1.8773	2.4524	1.7450	1.0132
Maximum	−0.1263	0.9531	1.9770	2.4828	1.8343	1.0742
Minimum	−0.5650	0.7428	1.8543	2.4115	1.6953	0.9298
Standard deviation	0.1307	0.0668	0.0343	0.0176	0.0360	0.0520
Skewness	0.3230	−0.3176	0.8236	−0.2618	0.5155	−0.1481
Kurtosis	1.7573	1.3944	2.3494	2.3095	2.5083	1.3920
Jarque-Bera	10.461	15.899	16.729	4.0044	6.9600	14.257
Probability	0.0053***	0.0003***	0.0002***	0.1350	0.0308**	0.0008***

***P < 0.01, and **P < 0.05

EG: economic growth; GASEF: gas efficiency; HTI: high-tech industry; LF: load capacity factor; OEF: oil efficiency; SGLO: social globalization.

Analysis of nonlinearity and stability

As previously discussed, quantile-based approaches are well-suited to the statistical attributes of the studied variables. However, before proceeding with the primary analysis, it is imperative to assess the appropriateness of the nonparametric approach for these variables. Following the methodology established in prior studies by Balcilar et al.,⁶¹ Olasehinde-Williams et al.,⁶⁰ and Alola et al.,⁵⁰ we employ both the Brock, Dechert, and Scheinkman (BDS) test by Broock et al.,⁶⁶ and the parameter stability test developed by Andrews and Ploberger⁶⁷ and Andrews.⁶⁸ The results of the BDS test, presented in Table 4, are used to evaluate nonlinearity in each variable. These findings indicate that the null hypothesis of independent identical distribution is dismissed across all embedded dimensions. This dismissal offers strong evidence of the presence of nonlinearity in all the indicators. This showcases that the variables under investigation do indeed exhibit nonlinear characteristics. Considering the nature of these series, it becomes clear that our nonparametric methodologies, including TVQR and TVNQC, are exceptionally well-suited for this study. These approaches effectively concurrently tackle both the nonlinearity and non-normality inherent in the dataset.

Table 4.

BDS test results.

	LF	EG	SGLO	GASEF	OEF	HTI
M2	46.520***	38.040***	51.325***	40.028***	36.438***	42.304***
P-value	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
M3	49.390***	39.803***	54.821***	42.784***	38.384***	44.853***
P-value	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
M4	53.068***	42.056***	59.333***	46.135***	40.815***	47.834***
P-value	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
M5	58.580***	45.436***	66.030***	51.031***	44.441***	52.331***
P-value	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
M6	66.178***	50.220***	75.247***	57.897***	49.529***	58.966***
P-value	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)

M denotes the dimension. Values inside () represent showcases P-values. ***P < 0.01.

EG: economic growth; GASEF: gas efficiency; HTI: high-tech industry; LF: load capacity factor; OEF: oil efficiency; SGLO: social globalization.

Table 5 showcases the results of the parameter stability test conducted to assess the stability characteristics of the study series. The findings from the parameter stability BDS test indicate the rejection of the null hypothesis (Ho) of stability for the variables under investigation at a significance level of 1%. This suggests the presence of structural breaks in these variables.

Table 5.

Parameter stability test results.

	LF	EG	SGLO	GASEF	OEF	HTI
Max. LR	424.109***	148.203***	749.068***	761.028***	281.435***	1641.836***
P-value	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
Exp. LR	208.494***	70.521***	370.706***	376.025***	136.534***	816.429***
P-value	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
Ave. LR	232.313***	49.498***	312.238***	222.929***	142.381***	184.303***
P-value	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)

Value inside () represents showcases P-values calculated via the method of Hansen.⁶⁹ ***P < 0.01.

EG: economic growth; GASEF: gas efficiency; HTI: high-tech industry; LF: load capacity factor; LR: long-run; OEF: oil efficiency; SGLO: social globalization.

Stationarity test results

It is worth noting that, as pointed out by Sim and Zhou⁷⁰ and subsequently corroborated by Balcilar et al.,⁶¹ and Das et al.,⁷¹ quantile-based analysis necessitates the use of stationary data. Furthermore, based on the nonlinearity and non-normal distribution of the series, using linear unit root tests inclusing augmented Dickey Fuller (ADF) and Philip Perron (PP) tests will produce misleading results. Thus, the study rest on using nonlinear ADF test suggested by Galvao⁷² with the results presented in Table 6. This test provides estimates relevant to the examination of data stationarity through quantile analysis. Specifically, the quantile unit root test presents estimates concerning the persistence and t-statistics for H0: α(τ) = 1 across a grid of 19 quantiles. The outcomes of the quantile unit root test reveal that HTI, LF, EG, SGLO, GASEF, and OEF exhibit nonstationary behavior when observed at the level series for all quantiles within the conditional distribution. These findings align with the results obtained from the ADF and PP tests, which were discussed previously.

Table 6.

Quantile unit test results.

		0.10	0.20	0.30	0.40	0.50	0.60	0.70	0.80	0.90
LF	CV	−1.135	−2.055	−2.456	−2.335	−2.389	−2.427	−2.282	−2.505	−2.372
LF	t-stats	0.973	−1.263	0.258	−0.629	−1.865	−1.602	−2.032	−1.303	−2.102
HTI	CV	−2.730	−2.847	−2.789	−2.821	−2.751	−2.682	−2.632	−2.511	−2.120
HTI	t-stats	−0.837	−2.141	−1.745	−0.976	0.519	0.117	0.862	1.097	0.557
GASEF	CV	−2.351	−2.668	−2.761	−2.859	−2.842	−2.774	−2.685	−2.499	−2.306
GASEF	t-stats	−0.256	−0.807	−1.458	−1.603	−2.311	−2.398	−2.960	−2.202	−1.957
SGLO	CV	−2.238	−2.521	−2.668	−2.857	−2.761	−2.858	−2.669	−2.634	−2.260
SGLO	t-stats	1.559	−0.735	−1.109	−1.325	−1.809	−1.735	−0.941	−1.228	0.839
OEF	CV	−2.344	−2.634	−2.740	−2.781	−2.857	−2.775	−2.795	−2.683	−2.260
OEF	t-stats	−1.833	−2.563	−2.268	−2.494	−1.659	−1.325	−1.005	−0209	−0.317
EG	CV	−2.209	−2.580	−2.733	−2.723	−2.845	−2.707	−2.450	−2.120	−2.027
EG	t-stats	1.248	0.490	0.198	−0.279	−1.338	−1.597	−2.008	−2.161	−2.011

Table 6 presents the estimated points and t-values at a 5% level of significance. t-values < CV, leads us to dismiss the null hypothesis of α(τ) = 1.s

CV: critical value; EG: economic growth; GASEF: gas efficiency; HTI: high-tech industry; LF: load capacity factor; OEF: oil efficiency; SGLO: social globalization.

Time varying quantile regression results

This study introduces, for the first time, the application of TVQR to explore the relationship between LFs and variables such as the HTI, energy efficiency (GASEF and OEF), and SGLO. This approach represents a significant advancement beyond the traditional quantile regression proposed by Koenker and Bassett.⁴⁸ TVQR allows us to capture the impact of Y on different quantiles of the distribution of X across various subperiods. Figure 4 illustrates the results obtained through TVQR. The vertical axis on the right-hand side presents a color bar ranging from light yellow to dark red (indicating the sign of the relationship). Additionally, the vertical axis on the left-hand side displays the quantile distribution. Meanwhile, the horizontal axis represents the subperiods under consideration. The TVQR results are shown in Figure 3.

Figure 4.

TVQR from HTI, GASEF, OEF, SGLO and EG to LF. (a) Time-varying impact of HTI on LF. (b) Time-varying impact of GASEF on LF. (c) Time-varying impact of OEF on LF. (d) Time-varying impact of SGLO on LF. (e) Time-varying impact of EG on LF. EG: economic growth; GASEF: gas efficiency; HTI: high-tech industry; LF: load capacity factor; OEF: oil efficiency; SGLO: social globalization; TVQR: time-varying quantile regression.

Figure 4(a) illustrates the outcomes of the TVQR analysis pertaining to the impact of the HTI on the LF in Saudi Arabia. Within the middle and high quantiles (ranging from 0.3 to 0.90), particularly spanning from 1992 to 2000, strong evidence emerges of a significant association between the HTI and LF. This suggests that the presence of an HTI contributes positively to ecological quality during this period. Furthermore, from 2001 to 2021, across a range of quantiles (from 0.1 to 0.90), there is discernible but weaker yet positive evidence of a connection between the HTI and LF. This implies that the HTI continues to promote ecological quality in Saudi Arabia over these years. In essence, the development of the HTI appears to be a catalyst for improving ecological quality in the country. These findings align with the conclusions drawn by Xu and Lin,^10,11 who employed a nonparametric additive regression model, as well as those of Shahzad et al.,^12,13 in the case of China, who utilized a dynamic connectedness vector autoregressive model. However, it is noteworthy that these findings run counter to the results reported by Du et al.,³³ in their analysis of the MINT nations.

The rationale behind our findings can be elucidated as follows. Firstly, the HTI has the capacity to enhance energy efficiency through several mechanisms. Technological advancements within the high-tech sector can lead to energy savings by improving the efficiency of equipment and machinery. Simultaneously, it can foster the development of a skilled workforce with a heightened environmental consciousness, leading to more energy-efficient practices.²⁶ Notably, certain segments of the HTI, such as resource utilization and ecological preservation, actively promote the adoption of energy-saving technologies and equipment, including waste generation and automobile exhaust afterburning technologies.^12,13 This contributes to reductions in energy and CO₂. Second, the HTI plays a pivotal role in popularizing a low-carbon lifestyle. High-tech products are progressively reshaping lifestyles and societal operations in various regions. These products can create cleaner and cost-effective living environments, prompting an increasing number of individuals to embrace various high-tech solutions, such as solar heat pipes, wind power motors, biomass energy, coal gasification and geothermal energy.²⁷ Thirdly, the HTI plays a pivotal role in driving the transition to a low-carbon economy. This sector not only provides cutting-edge equipment and technology for the modernization of conventional industries but also offers essential technical support for the manufacturing services industry.²⁶ On the one hand, the HTI facilitates the transformation of traditional sectors, ushering them into the realm of a low-carbon economy. For instance, it significantly enhances the productivity of the equipment manufacturing industry, resulting in reduced energy usage and diminished utilization of natural resources. Consequently, the HTI contributes to the enhancement of ecological quality in Saudi Arabia.

Figure 4(b) provides a visual representation of the TVQR analysis concerning the relationship between GASEF and the LF in Saudi Arabia. Across all quantiles (ranging from 0.10 to 0.90), the impact of GASEF on LF exhibits a consistently positive and robust pattern from 1992 to 2015. Additionally, within the medium and higher quantiles (ranging from 0.40 to 0.90) spanning the years 2015 to 2017, there is evidence of a noteworthy negative effect of GASEF on the LF that cannot be ignored. Furthermore, from 2017 to 2021 and across all quantiles (ranging from 0.10 to 0.90), the influence of GASEF on LF returns to a positive direction. In summary, it can be concluded that GASEF promotes ecological quality in Saudi Arabia by enhancing the LF. This outcome aligns with expectations, given that natural gas is generally considered to be a relatively environmentally friendly energy source when compared to oil and coal. Moreover, improvements in natural GASEF naturally contribute to a more environmentally tolerant energy landscape. Connecting this finding with existing literature²⁹ provides significant evidence that innovative utilization of natural gas and other energy sources, such as nuclear and oil energy, benefits the environment in Finland. However, studies also reveal that in the absence of technological innovation or environmentally focused approaches to natural gas, it can have detrimental effects on environmental quality.^6,7,39,42

Figure 4(c) illustrates the impact of OEF on the LF across various quantiles and subperiods. As anticipated, in the majority of quantiles and subperiods, the influence of OEF on the LF is notably positive. This outcome indicates that the utilization of OEF has mitigated environmental degradation in Saudi Arabia by elevating ecological quality. The plausible explanation for the consistently positive and significant effect of oil energy efficiency, as revealed in this study, demonstrates that when oil energy is used efficiently, it can contribute to ecological improvement. This finding underscores the role of technological advancement as a means to promote energy efficiency. Consequently, it implies that oil energy efficiency can play a pivotal role in ensuring a sustainable environment. This result aligns with past studies initiated by Liu et al.,³⁸ and Jin et al.,^8,9 both of which documented the positive impact of energy efficiency on ecological integrity.

Figure 4(d) depicts the influence of SGLO on the LF in Saudi Arabia across various periods and quantiles. As expected, across the majority of quantiles and subperiods, the impact of SGLO on the LF is generally weak and negative in Saudi Arabia. These research results diverge from previous findings by Olanrewaju et al.,⁴³ Sharif et al.³⁷, and Shi et al.³⁴, which suggested that globalization factors, including SGLO, political globalization, and economic globalization, tend to contribute to increased environmental pollution. However, this study's findings align with the conclusions put forth by Yameogo et al.⁷³ and Wang et al.,³⁵ both of whom affirmed the role of globalization in diminishing ecological excellence. It is worth noting that SGLO often leads to increased trade and EG, which can, in turn, drive higher levels of consumption. This heightened consumption can result in a greater demand for natural resources, including minerals, timber and fossil fuels, potentially leading to overexploitation and resource depletion.

Figure 4(e) provides an overview of the impact of EG on LF in Saudi Arabia. Across all quantiles (ranging from 0.1 to 0.90), particularly from 1992 to 2003, there is evidence of a positive effect of EG on LF, implying that EG has improved ecological quality during this period. However, from 2004 to 2021, across the majority of quantiles and subperiods, the effect of EG on the LF is negative, suggesting that EG has exacerbated ecological deterioration. These results challenge the Load Capacity Curve (LCC) hypothesis for Saudi Arabia. Our study's findings align with the conclusions drawn by Adedoyin et al.,⁷⁴ Ahmed and Galal,⁷⁵ Akram et al.,⁷⁶ Olasehinde-Williams and Özkan,⁷⁷ and Pata et al.,⁷⁸ who conveyed a negative impact of EG on ecological excellence. In light of these findings, there is reason for concern regarding the attainment of the objectives outlined in SDG 13, as the persisting pattern of EG in Saudi Arabia may pose challenges to achieving the objective of climate action. This evidence underscores the unsustainable nature of Saudi Arabia's EG. It underscores the importance of exploring alternative and clean energy solutions, not only for the betterment of ecological integrity but also for ensuring energy security. This issue may also have implications for the attainment of the objectives of SDG 7, as continued dependence on fossil fuels could hinder progress toward the goal of clean and affordable energy.

Nonparametric quantile causality results

Having established the relationship between the LF and its determining factors, we proceeded to employ nonparametric quantile causality (NPQC) to assess how the LF responds to these drivers. The outcomes of this analysis are visually presented in Figure 5. In Figure 5, the black line above the horizontal line signifies the rejection of the null hypothesis at a 10% significance level within the respective quantile. The vertical line represents the test statistic, while the broken red and blue lines depict the mean and variance, respectively. The plots in Figure 5(a) reveal the null hypothesis dismissal, indicating the presence of causality in both mean and variance, as the HTI influences the LF across the middle and higher quantiles (ranging from 0.20 to 0.90). These findings empirically demonstrate that the HTI exerts a statistically significant causal impact on the mean and variance of the LF in Saudi Arabia. Similarly, Figure 5(b) and (c) illustrate the dismissal of the Ho regarding causality in mean and variance, respectively, as it pertains to GASEF and OEF affecting the LF. These results imply that across the majority of quantiles, both GASEF and OEF possess predictive power over the LF. Furthermore, Figures 5(d) and (e) demonstrate the rejection of the null hypothesis concerning causality in mean and variance from SGLO and EG to the LF, spanning the majority of quantiles. These findings suggest that SGLO and EG have a causal influence on the LF across various quantiles.

Figure 5.

NPQC from OEF, GASEF, HTI, EG, and SGLO to LF.

Time-varying nonparametric quantile causality results

After establishing the causal effects of the HTI, energy efficiency (GASEF and OEF), and SGLO on the LF using traditional NPQC for the entire dataset, the examination of their time-varying effects commences through the application of the innovative time-varying non-parametric quantile causality (TVNPQC) method. A moving window size of 7 years, comprising approximately 28 observations, is employed to assess time-varying causality. Within this framework, NPQC is utilized, and test statistics are computed as indicators of time-varying causality. This approach allows for the identification of periods characterized by significant causality, both in terms of returns and volatility, originating from the HTI, energy efficiency (GASEF and OEF), and SGLO toward the LF. These findings can subsequently be linked to crucial developments in the study context.

Figures 6 and 7 illustrate the time-varying non-parametric quantile causality (TVNQC) results across different time periods and quantiles. The estimated test statistics are represented in the heatmaps, progressing from black to bright yellow. Figures 6(a) and 7(a) depict TVNQC results for the causal relationship from the HTI to LF in terms of mean and variance. The findings indicate the rejection of the null hypothesis of “no causality” in mean from HTI to LF within the middle quantiles (ranging from 0.3 to 0.65) during the periods of 1997–2004 and 2009–2014. Conversely, the results demonstrate the dismissal of the null hypothesis of “no causality” in the lower and middle tails (ranging from 0.2 to 0.65) during specific periods such as 1997–2004, 2005–2006, and 2012–2016.

Figure 6.

TVNQC in mean from HTI, GASEF, OEF, SGLO, and EG to LF. (a) TVNQC in mean from HTI to LF. (b) TVNQC in mean from GASEF to LF. (c) TVNQC in mean from OEF to LF. (d) TVNQC in mean from SGLO to LF. (e) TVNQC in mean from EG to LF. EG: economic growth; GASEF: gas efficiency; HTI: high-tech industry; LF: load capacity factor; OEF: oil efficiency; SGLO: social globalization; TVNQC: time-varying nonparametric quantile causality.

Figure 7.

TVNQC in variance from HTI, GASEF, OEF, SGLO, and EG to LF. (a) TVNQC in variance from HTI to LF. (b) TVNQC in variance from GASEF to LF. (c) TVNQC in variance from OEF to LF. (d) TVNQC in variance from SGLO to LF. (e) TVNQC in variance from EG to LF. EG: economic growth; GASEF: gas efficiency; HTI: high-tech industry; LF: load capacity factor; OEF: oil efficiency; SGLO: social globalization; TVNQC: time-varying nonparametric quantile causality.

Figures 6(b) and 7(b) present evidence of causality from GASEF to LF in terms of mean and variance. In the middle quantiles (ranging from 0.35 to 0.60), the null hypothesis of “no causality” from GASEF to LF in mean is rejected between 2009 and 2012. On the other hand, the outcomes reveal the dismissal of the null hypothesis of “no causality” in variance within the lower and middle quantiles from 1997–2004 and 2009–2016. Figures 5(c) and 6(c) display causality in mean and variance from OEF to LF. The results indicate the rejection of the null hypothesis of “no causality” in mean from OEF to LF in the lower tails (ranging from 0.1 to 0.30) during 2007–2006 and in the middle tails (ranging from 0.30 to 0.60) from 2009–2012. Conversely, the Ho of “no causality” from OEF to LF is dismissed in the lower tails during 1997–1998 and in the middle tails (ranging from 0.25 to 0.55) from 2012 to 2016.

Figures 6(d) and 7(d) reveal the causality from SGLO to LF. Causality in mean from SGLO to LF is observed across the majority of the quantiles during 1998–2002, 2006–2012, and in the middle quantile from 2014 to 2018. Conversely, the results show the dismissal of the null hypothesis of “no causality” in variance from SGLO to LF across the majority of the quantiles from 1997 to 2021. Lastly, Figures 6(e) and 7(e) demonstrate causality from EG to LF in terms of mean and variance. Causality is evident mostly across the quantiles (ranging from 0.20 to 0.75) from 1997 to 2017, indicating that EG can predict LF in both mean and variance during these periods. In conclusion, the novel insights uncovered in this study emphasize the critical importance of factoring in the HTI, energy efficiency (both GASEF and OEF), and SGLO when crafting ecological quality policies. These findings underscore that any changes or shifts in these factors can exert a substantial influence on ecological quality.

Conclusion and policy paths

Conclusion

Utilizing quarterly data spanning the period from 1990 to 2021, this study innovatively applied the TVQR methodology to scrutinize the impacts of explanatory variables on distinct quantiles within the distribution of the response variable across diverse time intervals. Furthermore, the research employed the TVNPQC test to assess causal relationships among variables at diverse quantiles and time intervals. In the process of variable selection, this study employed segregated energy efficiency (specifically, OEF and GASEF), the HTI, and SGLO as independent regressors, while the comprehensive LF served as the dependent variable. The analysis conducted through the application TVQR unveiled that both the HTI and OEF and GASEF exert a positive influence on environmental quality, as indicated by their capacity to augment the LF indicator across a substantial portion of quantiles and time intervals. In contrast, EG and SGLO pose a decreasing impact on environmental quality by decreasing the LF. Furthermore, the selected explanatory variables influence LF across different time frames and quantiles.

Policy recommendations

In the context of policy implications, given the HTI's role in advancing environmental sustainability, policymakers should prioritize the reinforcement of R&D efforts, as R&D plays a pivotal role in fostering the growth of the high-tech sector. Consequently, special funds may be established, and substantial investments can be allocated toward the cultivation and promotion of the HTI. Since the HTI is based on knowledge, innovation, and talent, educational institutions can focus on expanding their educational curriculum by offering more subjects related to ecology, marine sciences, vehicle engineering, environmental science, and others. Strategies and initiatives must also be formulated to transition from small emissions-intensive enterprises to high-tech industries, which can be facilitated through the provision of targeted tax incentives and incentives to encourage the establishment of enterprises leveraging advanced and sophisticated technologies. Also, starting a nationalized HTI by offering high incentives to foreign employees and offering them special perks and privileges to retain the talent is necessary as Saudi Arabia currently lacks the required domestic talent for this transformation. Building modern high-speed trains, low-energy vehicles, and infrastructure based on modern technology will consume less energy. However, supportive government policy and high funding and subsidies can increase the production of such HTIs in the country. Enhancing environmental quality necessitates educational initiatives aimed at encouraging residents to embrace energy-efficient lifestyles, which encompass the utilization of electric vehicles, solar-powered heating systems, and the adoption of collective public transportation modes such as buses and subways.

Regarding energy efficiency, the changes in the above behavioral aspects can also enhance energy efficiency. Further, practices like carpooling, regulating speed limits, imposing vehicle fuel economy standards, using hybrid and electronic vehicles, and strengthening public transport can encourage energy efficiency. Notably, the policies related to enhancing R&D can not only drive the HTI but also improve energy efficiency. Saudi Arabia should prioritize a reevaluation of its construction methodologies, incorporating contemporary energy-conservation techniques such as energy-efficient air conditioning, ventilation systems, and advanced heating solutions. The introduction of energy audit practices to curtail excessive energy consumption is also recommended as a beneficial measure for reducing overall energy usage. Enhancing the use of energy-efficient consumer electronics (such as smartphones, smart thermostats, computers, and laptops) and industrial products and machinery (such as highly efficient motors, energy monitoring systems, and modern sensors) can promote energy efficiency, which in turn can decrease the LF.

Apart from this, promoting SGLO with a view to enhancing environmental awareness is necessary. Currently, SGLO is enhancing environmental degradation, underscoring that interaction with the world promotes a more luxurious lifestyle in Saudi Arabia rather than bringing environmental awareness and connected environmental conservation priorities. In this context, practices like increasing the penetration of information technology, introducing more exchange programs in the universities particularly related to energy saving and environmental science, initiating more debates related to environmental protection in the educational institutions with suitable media coverage, promoting travel and mobility to countries with more green environmental practices, and offering incentives for participation in international conferences to students of Saudi Universities, might help to promote environmental conservation aspects of SGLO in the country.

Limitations and future perspective

This study employed a novel TVQR method to comprehensively assess the repercussions of segregated energy efficiency, SGLO, and the HTI on the LF in Saudi Arabia, yielding innovative and previously unreported findings. Nonetheless, this research selectively applied this method to a limited subset of variables. Future investigations may employ this methodology to examine the impacts of a broader array of factors, including technology, economic complexity, energy innovation, and other variables, which could yield interesting and insightful outcomes. This research has the potential to resonate with various international organizations and aid in advancing their objectives, spanning across MINT countries, G7 nations, ASEAN nations, EU nations, and BRICS countries. Consequently, conducting comparative studies could offer valuable insights and foster collaboration toward bolstering global environmental conservation endeavors. Lastly, data from diverse nations can be used for analysis. Furthermore, comparative research studies can serve as a valuable avenue for elucidating novel insights into the relationships among determinants of environmental quality and environmental quality itself.

Footnotes

Data availability

Data is readily available at a request from the corresponding author.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Tomiwa Sunday Adebayo

Zahoor Ahmed

Notes

Tomiwa Sunday Adebayo currently serves as a Faculty Member/Assistant Professor within the esteemed Department of Business Administration at Cyprus International University, situated in Northern Cyprus, Turkey. Additionally, he holds the position of non-resident faculty member at Lebanese American University, as well as the University of Tashkent for Applied Sciences in Tashkent, Uzbekistan. His exceptional scholarly contributions have garnered significant recognition, positioning him within the top 2% of global scientists/researchers, as evaluated by Stanford University in both 2022 and 2023. Moreover, he has been lauded as one of the Top 82 Young Economists by Research Papers in Economics (RePEc), which stands as the preeminent database for economists worldwide. Adebayo also serves as an associate editor for the International Journal of Sustainable Development & World Ecology, a distinguished publication indexed in the Web of Science, with a Q1 rating and an Impact Factor of 5.6. His prolific publication record encompasses numerous reputable journals indexed in Scopus and Web of Science. His areas of interest include Energy Policy, Investments and Portfolio Management, Applied Econometrics, Energy and Environmental Economics, Applied Econometrics.

Oktay Özkan is an associate professor of finance at the Faculty of Economics and Administrative Sciences, Tokat Gaziosmanpasa University, Turkey. He ranked 624th among approximately 1.5 million candidates in the university entrance exam held by the Republic of Turkey Measurement, Selection and Placement Center in 2006. Graduating from TOBB University of Economics & Technology, Ankara/Turkey, he secured the 2nd position in the Department of Business Administration with an average of 3.71/4.00. He was awarded the Scientific and Technological Research Council of Turkey (TÜBİTAK) scholarship, a prestigious accolade granted to very few individuals in Turkey, for both his master's and Ph.D. education. He earned his PhD degree in January 2020 from Tokat Gaziosmanpasa University, Turkey. Remarkably, he achieved the title of Associate Professor in just 1.5 years, receiving the approval of the Inter-Universities Board, Ankara/Turkey, in June 2021. This accomplishment makes him the fastest Associate Professor of Finance in Turkey. He has published several research articles in reputable international academic journals, such as Energy Economics, Journal of Cleaner Production, Energy, Journal of Environmental Management, Resources Policy, and Research in International Business and Finance. His areas of interest include Financial Markets and Institutions, Financial Forecasting and Modeling, Investments and Portfolio Management, Applied Econometrics, Energy and Environmental Economics, Applied Time-Series Econometrics.

Zahoor Ahmed is an Assistant Professor at Bahçeşehir Cyprus University and non resident faculty member at Lebanese American University. He earned his PhD from the prestigious School of Management and Economics at Beijing Institute of Technology. With a profound interest in various domains, his research encompasses an array of critical areas including financial development, globalization, energy economics, environmental economics, financial management, and ICT among others. He has published more than 100 articles in reputed international journals. In recognition of his outstanding contributions to academia, he was included among the Highly Cited Researchers by Clarivate for 2023.

Victoria Olushola Olanrewaju is currently pursuing her PhD at the European University of Lefke. Her scholarly endeavors have resulted in the publication of numerous research articles in esteemed international academic journals, including the Journal of Cleaner Production, Energy, International Journal of Sustainable Development & World Ecology, and Environmental Science and Pollution Research. Her academic focus revolves around Strategic Management, Environmental Economics, and Sustainability, reflecting her dedication to addressing pressing global challenges through rigorous research and analysis.

References

WMO. World Meteorological Organization. Atlas of Mortality and Economic Losses From Weather, Climate and Water Extremes . 2021. Https://public.wmo.int/en/resources/atlas-of-mortality (Issue 1123). Technical Report.

Çitil

İlbasmış

Olanrewaju

, et al. Does green finance and institutional quality play an important role in air quality. Environ Sci Pollut Res 2023. https://doi.org/10.1007/s11356-023-26016-2

Dai

Ahmed

Alvarado

, et al. Assessing the nexus between human capital, green energy, and load capacity factor: policymaking for achieving sustainable development goals. Gondwana Res 2023. https://doi.org/10.1016/j.gr.2023.04.009

IEA. CO2 emissions from fuel combustion: Overview. In International Energy Agency. 2018. https://webstore.iea.org/download/direct/1082?fileName=CO2_Emissions_from_Fuel_Combustion_2018_Overview.pdf

Kanat

Yan

Asghar

, et al. Do natural gas, oil, and coal consumption ameliorate environmental quality? Empirical evidence from Russia. Environ Sci Pollut Res 2022; 29: 4540–4556.

Akram

Chen

Khalid

, et al. Heterogeneous effects of energy efficiency and renewable energy on carbon emissions: evidence from developing countries. J Cleaner Prod 2020a; 247: 119122.

Akram

Chen

Khalid

, et al. Heterogeneous effects of energy efficiency and renewable energy on carbon emissions: evidence from developing countries. J Cleaner Prod 2020b; 247: 119122.

Jin

Ahmed

Korkut

, et al. Do investments in green energy, energy efficiency, and nuclear energy R & D improve the load capacity factor? An augmented ARDL approach. Geosci Front 2023a: 101646. https://doi.org/10.1016/j.gsf.2023.101646

Jin

Ahmed

Pata

, et al. Do investments in green energy, energy efficiency, and nuclear energy R&D improve the load capacity factor? An augmented ARDL approach. Geosci Front 2023b: 101646. https://doi.org/10.1016/j.gsf.2023.101646

10.

Lin

. Investigating the role of high-tech industry in reducing China’s CO2 emissions: a regional perspective. J Cleaner Prod 2018a; 177: 169–177.

11.

Lin

. Investigating the role of high-tech industry in reducing China’s CO2 emissions: a regional perspective. J Cleaner Prod 2018b; 177: 169–177.

12.

Shahzad

Ferraz

Nguyen

, et al. Investigating the spill overs and connectedness between financial globalization, high-tech industries and environmental footprints: fresh evidence in context of China. Technol Forecast Soc Change 2022a; 174: 121205.

13.

Shahzad

Ferraz

Nguyen

H-H

14.

Rudolph

Figge

. Determinants of ecological footprints: what is the role of globalization? Ecol Indic 2017; 81: 348–361.

15.

Ahmed

Wang

Mahmood

, et al. Does globalization increase the ecological footprint? Empirical evidence from Malaysia. Environ Sci Pollut Res 2019; 26: 18565–18582.

16.

Dogan

Pata

. The role of ICT, R&D spending and renewable energy consumption on environmental quality: testing the LCC hypothesis for G7 countries. J Cleaner Prod 2022; 380: 135038.

17.

Shang

Razzaq

Chupradit

, et al. The role of renewable energy consumption and health expenditures in improving load capacity factor in ASEAN countries: exploring new paradigm using advance panel models. Renewable Energy 2022; 191: 715–722.

18.

Adebayo

Pata

Akadiri

. A comparison of CO2 emissions, load capacity factor, and ecological footprint for Thailand’s environmental sustainability. Environ Develop Sustain 2022: 0123456789. https://doi.org/10.1007/s10668-022-02810-9

19.

Akadiri

Adebayo

Riti

, et al. The effect of financial globalization and natural resource rent on load capacity factor in India: an analysis using the dual adjustment approach. Environ Sci Pollut Res 2022; 29: 89045–89062.

20.

Kahouli

Hamdi

Nafla

, et al. Investigating the relationship between ICT, green energy, total factor productivity, and ecological footprint: empirical evidence from Saudi Arabia. Energy Strategy Rev 2022; 42: 100871.

21.

Baloch

, Danish, et al. Nexus between financial development and CO2 emissions in Saudi Arabia: analyzing the role of globalization. Environ Sci Pollut Res 2018; 25: 28378–28390.

22.

OWD. Our World in Data . 2022. Https://ourworldindata.org/grapher/fossil-fuels-per-capita?tab = table .

23.

Statista. Distribution of carbon dioxide emissions worldwide in 2021, by select country . 2021. https://www.statista.com/statistics/271748/the-largest-emitters-of-co2-in-the-world/ .

24.

GFN. Global Footprint Network . 2023a. https://data.footprintnetwork.org/#/.

25.

GFN. Global Footprint Network . 2023b. Http://data.footprintnetwork.org .

26.

Rafique

Schneider

Shahzad

, et al. High-tech industries, financial expansion, and low-carbon energy deployment along the belt and road initiative. Sustain Dev 2022; 30: 1779–1795.

27.

Wang

Gao

Wei

, et al. How FDI and technology innovation mitigate CO2 emissions in high-tech industries: evidence from province-level data of China. Environ Sci Pollut Res 2022; 29: 4641–4653.

28.

Anser

Khan

Umar

, et al. Formulating sustainable development policy for a developed nation: exploring the role of renewable energy, natural gas efficiency and oil efficiency towards decarbonization. Int J Sustain Dev World Ecol 2023; 0: 1–17. https://doi.org/10.1080/13504509.2023.2268586

29.

Akdag

Yıldırım

. Toward a sustainable mitigation approach of energy efficiency to greenhouse gas emissions in the European countries. Heliyon 2020; 6: e03396.

30.

Bisht

Pitchforth

Murray

. Understanding India, globalisation and health care systems: a mapping of research in the social sciences. Global Health 2012; 8: 32.

31.

Sharif

Saqib

Dong

, et al. Nexus between green technology innovation, green financing, and CO2 emissions in the G7 countries: the moderating role of social globalisation. Sustain Dev 2022a; 30: 1934–1946.

32.

Alariqi

Long

Singh

, et al. Modelling dynamic links among energy transition, technological level and economic development from the perspective of economic globalisation: evidence from MENA economies. Energy Rep 2023; 9: 3920–3931.

33.

Jiang

Adebayo

, et al. Asymmetric effects of high-tech industry and renewable energy on consumption-based carbon emissions in MINT countries. Renew Energy 2022; 196: 1269–1280.

34.

Shi

Jia

Mehmood

. Exploring the roles of green finance and environmental regulations on CO2es: defining the roles of social and economic globalization in the next eleven nations. Environ Sci Pollut Res 2023; 30: 62967–62980.

35.

Wang

Chen

Ullah

, et al. Resource curse or blessing? Evaluating the role of natural resource, social globalization, and environmental sustainability in China. Resources Policy 2023; 85: 103749.

36.

Kirikkaleli

Adebayo

. Political risk and environmental quality in Brazil: role of green finance and green innovation. Int J Finance Econ 2023; 5. https://doi.org/10.1002/ijfe.2732

37.

Sharif

Saqib

Dong

, et al. Nexus between green technology innovation, green financing, and CO2 emissions in the G7 countries: the moderating role of social globalisation. Sustain Dev 2022b; 4: 12–24.

38.

Liu

Adebayo

Ramzan

, et al. Do coal efficiency, climate policy uncertainty and green energy consumption promote environmental sustainability in the United States? An application of novel wavelet tools. J Cleaner Prod 2023; 417: 137851.

39.

Alola

Özkan

Usman

. Role of non-renewable energy efficiency and renewable energy in driving environmental sustainability in India: evidence from the load capacity factor hypothesis. Energies 2023b; 16: Article 6.

40.

Naimoglu

Akal

. The relationship between energy technology, energy efficiency, renewable energy, and the environment in Türkiye. J Cleaner Prod 2023; 418: 138144.

41.

Zhang

Sadiq

, et al. Impact of a sharing economy on sustainable development and energy efficiency: evidence from the top ten Asian economies. J Innov Knowledge 2023; 8: 100320.

42.

Adebayo

Ullah

. Towards a sustainable future: the role of energy efficiency, renewable energy, and urbanization in limiting CO2 emissions in Sweden. Sustain Dev 2023; 3. https://doi.org/10.1002/sd.2658

43.

Olanrewaju

Irfan

Altuntaş

, et al. Towards sustainable environment in G7 nations: the role of renewable energy consumption, eco-innovation and trade openness. Front Environ Sci 2022; 10. https://www.frontiersin.org/articles/10.3389/fenvs.2022.925822

44.

Ojekemi

Rjoub

Awosusi

, et al.

Toward a sustainable environment and economic growth in BRICS economies: do innovation and globalization matter?

Environ Sci Pollut Res 2022. https://doi.org/10.1007/s11356-022-19742-6

45.

Awosusi

Ozdeser

Seraj

, et al.

Can green resource productivity, renewable energy, and economic globalization drive the pursuit of carbon neutrality in the top energy transition economies?

Int J Sustain Dev World Ecol 2023; 0: 1–15. https://doi.org/10.1080/13504509.2023.2192007

46.

Oyebanji

Kirikkaleli

Ayobamiji

. Consumption-based CO2 emissions in Denmark: the role of financial stability and energy productivity. Integr Environ Assess Manag 2023; n/a. https://doi.org/10.1002/ieam.4757

47.

Adebayo

. Environmental consequences of fossil fuel in Spain amidst renewable energy consumption: a new insights from the wavelet-based Granger causality approach. Int J Sustain Dev World Ecol 2022; 29: 579–592.

48.

Koenker

Bassett

. Regression quantiles. Econometrica 1978; 46: 33–50.

49.

Pata

Isik

. Determinants of the load capacity factor in China: a novel dynamic ARDL approach for ecological footprint accounting. Resources Policy 2021; 74: 102313.

50.

Alola

Özkan

Usman

. Examining crude oil price outlook amidst substitute energy price and household energy expenditure in the USA: a novel nonparametric multivariate QQR approach. Energy Economics 2023a; 120: 106613.

51.

WB. World Developmental Indicator . 2023. https://data.worldbank.org/country/china

52.

KOF. ETH Zurich KOF Globalisation Index. https://kof.ethz.ch/en/forecasts-and-indicators/indicators/kof-globalisation-index.html (2023, Accessed 20th October 2023).

53.

Ozkan

Sharif

Mey

, et al. The dynamic role of green technological innovation, financial development and trade openness on urban environmental degradation in China: Fresh insights from carbon efficiency. Urban Climate 2023; 52: 101679. https://doi.org/10.1016/j.uclim.2023.101679

54.

Koenker

. A note on studentizing a test for heteroscedasticity. J Econom 1981; 17: 107–112.

55.

Akhayere

Kartal

Adebayo

, et al. Role of energy consumption and trade openness towards environmental sustainability in Turkey. Environmental Science and Pollution Research 2023; 30(8): 21156–21168. https://doi.org/10.1007/s11356-022-23639-9

56.

Koenker

Hallock

. Quantile regression. Journal of Economic Perspectives 2001; 15(4): 143–156.

57.

. The relationship between economic growth, energy consumption, and CO2 emissions: Empirical evidence from China. Science of the Total Environment 2004; 542: 360–371.

58.

Olanipekun

Ozkan

Olasehinde-Williams

. Is renewable energy use lowering resource-related uncertainties. Energy 2023; 2(6): 126–137.

59.

Lee

C-C

Olasehinde-Williams

Özkan

. Geopolitical oil price uncertainty transmission into core inflation: evidence from two of the biggest global players. Energy Econ 2023; 126: 106983.

60.

Olasehinde-Williams

Özkan

Akadiri

. Effects of climate policy uncertainty on sustainable investment: a dynamic analysis for the U.S. Environ Sci Pollut Res 2023; 30: 55326–55339.

61.

Balcilar

Gupta

Pierdzioch

. Does uncertainty move the gold price? New evidence from a nonparametric causality-in-quantiles test. Resources Policy 2016; 49: 74–80.

62.

Nishiyama

Hitomi

Kawasaki

, et al. A consistent nonparametric test for nonlinear causality—specification in time series regression. J Econom 2011; 165: 112–127.

63.

Jeong

Härdle

Song

. A consistent nonparametric test for causality in quantile. Econ Theory 2012; 28: 861–887.

64.

Birindelli

Miazza

Paimanova

, et al. Just “blah blah blah”? Stock market expectations and reactions to COP26. International Review of Financial Analysis 2023; 8(8): 102–122.

65.

Balcilar

Ozdemir

. Economic growth and energy consumption causal nexus viewed through a bootstrap rolling window. Energy Economics 2010; 32(6): 1398–1410.

66.

Broock

Scheinkman

Dechert

, et al. A test for independence based on the correlation dimension. Econom Rev 1996; 15: 197–235.

67.

Andrews

DWK

Ploberger

. Optimal tests when a nuisance parameter is present only under the alternative. Econometrica 1994; 62: 1383–1414.

68.

Andrews

DWK

. Tests for parameter instability and structural change with unknown change point. Econometrica 1993; 61: 821–856.

69.

Hansen

B E

. Approximate asymptotic p values for structuras-change tests. Journal of Business and Economic Statistics 1997; 15(1): 60–67.

70.

Sim

Zhou

. Oil prices, US stock return, and the dependence between their quantiles. J Bank Financ 2015; 55: 1–8.

71.

Das

Murshed

Rej

, et al. Can clean energy adoption and international trade contribute to the achievement of India’s 2070 carbon neutrality agenda? Evidence using quantile ARDL measures. Int J Sustain Dev World Ecol 2023; 30: 262–277.

72.

Galvao

. Unit root quantile autoregression testing using covariates. J Econom 2009; 152: 165–178.

73.

Yameogo

CEW

Omojolaibi

Dauda

ROS

. Economic globalisation, institutions and environmental quality in sub-Saharan Africa. Res Global 2021; 3: 100035.

74.

Adedoyin

Gumede

Bekun

, et al.

Modelling coal rent, economic growth and CO2 emissions: does regulatory quality matter in BRICS economies?

Sci Total Environ 2020; 710: 136284.

75.

Ahmed

Galal

. Using artificial neural networks to recognize the determinants of economic growth: new approach ‘evidence from developing countries’. Indian J Econ Business 2021; 20. https://doi.org/10.5281/zenodo.5410524

76.

Akram

Umar

Xiaoli

, et al. Dynamic linkages between energy efficiency, renewable energy along with economic growth and carbon emission. A case of MINT countries an asymmetric analysis. Energy Rep 2022; 8: 2119–2130.

77.

Olasehinde-Williams

Özkan

. A consideration of the environmental externality of Turkey’s integration into global value chains: evidence from dynamic ARDL simulation model. Environmental Science and Pollution Research 2023; 30(11): 29154–29163.

78.

Pata

Kartal

Ullah

. Enhancing environmental quality in the United States by linking biomass energy consumption and load capacity factor. Geoscience Frontiers 2023; 14(3): 101–121.