1 Introduction

Depletion of natural assets breaks the sustainability of almost every sector, especially economic activities. Conducting economic activities through conventional sources (e.g., fossil fuels) is the main driver of greenhouse gases (GHG) that deteriorate the natural environment in the same way. Therefore, managing sustainability in the environment and economic activities has become a global priority. Taking into consideration such challenges, all United Nations member countries agreed in 2015 to maintain sustainability in various sectors by 2030. Particularly, ecological sustainability has been considered an essential goal of sustainable development goals (SDGs). Moreover, the Paris agreement regarding climate change and the recent Glasgow Climate Pact have considerable relevance for different economies. Disasters pertinent to climate change have increased over the last decade. As a result, around 2 million people have lost their lives, as well as around 3.60 trillion USD in economic losses (WMO 2021). As mentioned by the IEA (2018), environmental degradation has almost doubled since the 1970s. Economies were passionate about increasing their growth by overlooking environmental degradation due to massive GHG emissions. Due to ecological degradation, economies are vigilant to consider environmental sustainability as an important factor. In the wake of this, economies are becoming more eco-friendly by considering green growth, energy efficiency, and green technology innovation. Such factors may have a significant impact on environmental quality. Therefore, this research reveals the impact of green growth, energy efficiency, and green technology innovation on environmental sustainability.

Due to the drastic increase in global warming and other ecological problems, the subject of environmental sustainability has been a major concern for policy analysts and academicians as a GHG emissions mitigation tool (Long et al. 2018). The advancements in green technological innovation (GTI), energy efficiency (EE), and green growth (GG) have achieved much importance in the literature of environmental economics. Numerous works have been done to inspect this debate in relation to different factors that may stop pollution secretions through advancement in GTI, GG, and EE (Khan et al. 2023a, b; Guloglu et al. 2023; Pata and Samour 2023; Jin et al. 2023; Deng et al. 2023; Wenlong et al. 2023; Jahanger et al. 2023; Farooq 2023). These studies concluded that environmental sustainability techniques play a significant role in reducing pollution emissions. Moreover, such techniques also assist in increasing the biocapacity of land. Mitigating the detrimental effects of ecological footprints and increasing the biocapacity of land have been highlighted at different international treaties and meetings. The report by Brundtland (1985) noted that the human economy is undergoing ecological deterioration, and he further suggested that we should do economic expansion within ecologically acceptable limits. The consumption of traditional sources of energy (oil, gas, and coal) triggers ecological catastrophe and a reduction in biocapacity (Shahbaz et al. 2017a, b). The panacea to these challenges lies in introducing changes in the energy sector through GTI, GG, and EE. For this purpose, businesses introduce green technology innovation, energy efficiency, and green growth approaches that promote the application of innovative solutions by having a minimum impact on load capacity factor. Moreover, various scholars have shed light on the importance of green technology innovation for businesses and the environment, but they have failed to generalize this trend by disclosing the overall impact of green growth, energy efficiency, and green technology innovation on load capacity factors. Economies have taken actions to reduce environmental impact and increase biocapacity by addressing environmentally sustainable regulations in recent decades (Akdag and Yildirim 2020; Zhuang et al. 2021). Initially, economies prioritize economic development rather than environmental sustainability, which reveals a U-shaped liaison between environmental deterioration and economic development. However, as economies reach a higher income level, they intend to invest in green technologies to become more environmentally conscious. We refer to this concept as the load capacity curve (LCC) hypothesis.

Energy efficiency relates the quantity of useful energy output to the overall energy input, indicating the effectiveness of a process in consuming energy sources. It includes measures and technologies intended to reduce energy waste and augment energy utilization, thereby mitigating ecological impact, and improving sustainability. Similarly, green growth exemplifies the notion of economic development that enhances ecological sustainability, accentuating strategies and policies that stimulate growth while mitigating negative environmental consequences. It involves pursuing economic practices that prioritize the prominence of resource efficiency, pollution mitigation, and environmental preservation, leading to long-term affluence without compromising ecological integrity. Green technological innovation refers to the deployment and advancement of modern technologies aimed at incorporating ecological challenges and enhancing sustainability. It entails the construction of innovative solutions, e.g., technologies regarding renewable energy, sustainable materials, and efficient production processes, to alleviate ecological deterioration and advocate the transition towards a more sustainable and greener future.

Additionally, most of the earlier studies have focused on CO2 emission and ecological footprint (EFP) (Pata and Ertugrul 2023; Pata et al. 2023b, a; Guloglu et al. 2023) as an indicator to examine the environmental performance. However, these two aspects suffer from significant limitations. For instance, CO2 emissions only refer to air pollution by neglecting water and soil pollution and do not offer a significant opportunity to evaluate environmental degradation in a complete manner. This is due to the fact that ecological concerns encompass more than just air pollution (Caglar et al., 2023). On the other hand, the EFP determines the extent to which human demands on nature surpass nature's ability to meet those demands. This particular environmental indicator only displays the ways in which the environment has degraded as a result of the pressure that humans have placed on natural resources. Consequently, EFP only focuses on the demand side and has a tendency to overlook the influence of nature's supply side. Therefore, environmental evaluations should include biocapacity, which is the supply of naturally occurring resources that can meet demand as it is right now (Pata et al. 2023b). As a result, we utilize the load capacity factor measured as ecological footprint/biocapacity as an indicator of environmental quality. This indicator refers to a more comprehensive proxy of environmental sustainability because it considers both the demand and supply sides of the earth. In other words, pollution is something that an environment can accept without significant deterioration. Such factors determine the level at which resource extraction or pollution secretions can be sustained without depleting resources.

Further, Fig. 1 below discloses the LCF levels of the top 10 manufacturing economies between 1990 and 2019. The red line in Fig. 1 denotes the sustainability limit, which means that if an economy stands above the red line, natural assets and ecological conditions are up to par to meet human needs. Nevertheless, if an economy stands below the red line, that means the environmental quality is not good. It further highlights that natural assets and environmental conditions are not enough to meet human needs. LCF simultaneously considers both ecological footprint and biocapacity. In 1990, although India, China, Korea, and Mexico were below the sustainability line, they still had a decreasing trend, which shows the unsustainability of the environment. However, the UK and Germany have an increasing trend, but unfortunately, they are also below the sustainability line, which shows the unsustainability of the environment. Moreover, the USA, Korea, Japan, and Italy do not have an increasing or decreasing trend, but they remain at the same level over time.

Fig. 1
figure 1

Load capacity factor in top 10 manufacturing economies (Source; Authors own calculation)

Full size image

When discussing the most appropriate metric to gauge ecological conditions and global advancements in tackling pressing issues, it is crucial not to disregard the leading manufacturing economies and their contribution to environmental catastrophes. The top ten manufacturing economies are as follows: China, the United States of America, Japan, Germany, France, the United Kingdom, Italy, Canada, India, Mexico, and South Korea. These economies are accountable for the majority of ecological issues that exist today. These nations account for more than 65 percent of global capital and approximately 55 percent of global GDP (Climate Transparency 2018). The world's leading nations are involved in large-scale commercial activities that are contributing to an increasing number of ecological crises (OECD 2019). For analysis purposes, we considered the top 10 manufacturing economies because they are production-oriented countries. As a result, they are contributing to environmental degradation on a large scale. For instance, China has $4.7 trillion in gross value added (GVA) in the manufacturing sector, the USA has $2.3 trillion, Japan has $1.0 trillion, Germany has $798 billion, India and South Korea have $395 billion, Italy has $356 billion, and the UK has $304 billion in GVA. Due to the high production and consumption levels of these economies, they spend hefty funds on R&D to improve environmental quality by mitigating environmental degradation. Moreover, these economies have significant ecological impact, high resource utilization, worldwide influence, policy relevance, economic importance, data availability, and a wide-reaching supply chain. Studying under-analyzed economies allows for a deep-down examination of opportunities, challenges, and strategies regarding ecological sustainability. Such economies collectively demonstrate a significant chunk of the global carbon emissions contribution. In 2022, China led the pack with roughly 11 gigatons of CO2 emissions and produced 30.30% of global emissions, fundamentally due to its dependence on coal and heavy industrial activities. Following closely, the U.S. contributed collectively 6 gigatons and produced 13.43% of global CO2 emissions, reflecting its high energy utilization and reliance on fossil fuels. In the meantime, India, with about 3 gigatons and producing 8% of global emissions, confronted difficulties in balancing economic growth with ecological sustainability. However, Germany and Japan, known for their technological advancement, emitted around 0.9 and 1.3 gigatons, respectively, emphasizing the importance of transitioning towards clean energy sources. Germany produces 1.8% of the global emissions. Similarly, the U.K. and France showed relatively lower CO2 secretions, each contributing about 0.4 gigatons. The United Kingdom contributed 1.2% of total global CO2 emissions. Italy, Mexico, and Korea fell within the range of 0.5–0.7 gigatons, highlighting varying degrees of ecological policy effectiveness and industrial progress. Italy and Mexico produced 1.3 and 1.6 percent of the worldwide CO2 emissions (Jahanger et al. 2023). Overall, such economies not only bear responsibility for significant environmental degradation, but they also have the capacity to drive determination towards alleviating climate change through environmentally sustainable practices. In brief, industrial-oriented economies are concerned about environmental sustainability through green growth, energy efficiency, and green technological innovation.

Based on the above-mentioned information, the characteristics of the world's top manufacturing economies and the enormous influence of these nations on environmental sustainability make it a worthy case to investigate which factors exacerbate environmental deterioration and which ones reduce their negative impact on the environment. Therefore, this study examines the impact of green growth, energy efficiency, and green technological innovation on load capacity factor. We also propose that there can be many non-environmental factors that affect the environmental quality of top 10 manufacturing economies. Therefore, we have included non-environmental control variables like gross domestic product, trade openness, and human development. For empirical analysis, we considered the data of the top ten manufacturing economies from 1990 to 2019. Further, we employed the cross-sectional autoregressive distributed lag model (CS-ARDL) to check short- and long-term relationships. To the best of our understanding, this is the first research in the literature to investigate the short- and long-run impacts of green growth, energy efficiency, and green technology innovation on the load capacity in the top 10 manufacturing nations. This study delves into the mechanism through which the process of ecological sustainability contributes to favorable ecological outcomes by underscoring the following research objectives:

  • To investigate whether there exists a stable cointegration connection over the long run between the load capacity factor and the study's regressors in the top 10 manufacturing nations.

  • To evaluate the efficacy of green growth policies in lessening environmental degradation and enhancing load capacity factor in underlying economies.

  • To examine the role of energy efficiency in diminishing environmental degradation and promoting load capacity factor in underlying nations.

  • To access the connection between green technology innovation and improvement in load capacity factor across the top 10 manufacturing economies.

  • To evaluate the causal connections and the direction of causality between green growth, energy efficiency, green technological innovation, economic growth, trade openness, human capital, and load capacity factor in underlying nations.

Based on the above objectives, we build the following research hypotheses:

  • There is a significant relationship between green growth and the load capacity factor.

  • There is a significant relationship between energy efficiency and the load capacity factor.

  • There is a significant connection between green technological innovation and the load capacity factor.

The current research contributes to the existing literature in several ways. Firstly, this study focuses on the top 10 manufacturing nations, which account for a sizeable share of the world's population and make a considerable contribution to both global output and global environmental degradation. Second, this research introduces a new idea in the stream of previous literature regarding the substantial role of green growth, energy efficiency, and green technology innovation in enhancing the level of load capacity factor in the top 10 manufacturing economies. Third, most of the work in the literature focused on the demand side of the ecosystem, such as ecological footprint and CO2 emissions. However, in this current study, we consider LCF a more comprehensive indicator that discusses both the demand side and the supply side of the planet. Furthermore, this assortment of variables has never appeared in any other study in the literature. Fourth, this study utilizes the most recent 2nd generation econometric approaches, e.g., cross-sectional dependence, CIPS, CADF unit root test, Westerlund (2007) cointegration test, cross-sectional ARDL, AMG, and CCEMG approaches. In addition, we also applied D-H causality to find the liaison between variables. Lastly, the outcomes of this study provide real-world insights that inform decision-making, policy design, and the implementation of effective strategies for achieving sustainable goals. It further highlights that green practices regarding environmental sustainability play a vital role in enhancing biocapacity and decreasing ecological degradation.

This research organizes the remaining sections as follows: Sect. 2 highlights the review of existing literature. Section 3 outlines the variables and methodology used for analysis. Section 4 reports and discusses the study's empirical results. Finally, Sect. 5 discusses the study's conclusion, policy recommendations, and research limitations.

2 Review of literature

The interrelation between ecological sustainable practices and environmental degradation have been a challenge in environmental economics over the last few decades. In the wake of global warming, the adoption of sustainability techniques is believed to have a substantial impact on ecological sustainability while reducing GHG emissions in the atmosphere. Therefore, numerous studies have been done to find the drivers of environmental. In this section, we will highlight prior studies' outcomes.

2.1 Energy efficiency and environmental sustainability

There appears to be a shortage of literature about the liaison between energy efficiency and load capacity factor so far. A very few scholars worked to explore the energy efficiency influence on some other factors e.g., economic growth and CO2 emissions instead of load capacity factors. The work of Brookes (1990) found that there is a significant impact of energy efficiency on economic growth; however, it has no significant impact on CO2 emissions. Another study conducted by Wang et al. (2017) explored the nexus between energy efficiency and mitigation of CO2 emissions in the context of China. They further suggested that reduction in GHG emissions methods lead to an upsurge in energy efficiency. Emir and Bekun (2019) noted in their study that energy efficiency is an important mechanism to reduce energy consumption and global warming. The work of Akdag and Yıldırım (2020), Ehsanullah et al. (2021), and Wenlong et al. (2023) found an inverse connection between energy efficiency and GHG emissions. The work of Chien et al. (2022b) has explored that renewable energy and energy efficiency affected CO2 emissions positively in a panel of Asian economies. Their core findings recommended that renewable energy and energy efficiency minimize carbon emissions in the long run. However, it is only effective in the short-run in case of India and Nepal. The study of Akram et al. (2020a) explored the impact of energy efficiency on CO2 emissions by constructing the EKC hypothesis for a panel of 66 developing states. They found an adverse relationship between energy efficiency and GHG emissions by employing panel OLS and fixed effect panel quantile regression. However, the study of Guloglu et al. (2023) considered 26 OECD economies by finding the determinants of load capacity factors. To attain their purpose, they employed quantile common correlation effects mean group method. Their findings suggest that renewable energy, energy efficiency, and human capital assist to expand the supply side of nature (bio-capacity). Shang et al. (2022) recently described that ecological footprint is not enough to judge environmental quality. In short, based on the above studies, we may develop our first hypothesis that.

H1

There is a significant relationship between energy efficiency and load capacity factor.

2.2 Green growth and environmental sustainability

In this contemporary era, it is the prime concern of the world to make sure sustainable development. In this respect, achieving green growth is extensively acknowledged as the more suitable pathway. Green growth introduces saving technologies and pollution emissions mitigation approaches to decouple ecological disruption from economic growth (Acemoglu et al. 2016). Ecological technologies assist in diminishing CO2 secretions and are considered an essential factor of green growth (Ulucak 2020). Technologies regarding clean energy and research and development expenditure require more concentration to attain sustainable development goals (Li et al. 2022). The notion related to green growth is not new in the existing literature. This concept emerged after the worldwide financial crisis of 2008. The study of Hao et al. (2021a) explored the relationship between green growth on GHG emissions in the context of G-7 economies throughout 1991–2017. Their study highlights that green growth sustains ecological quality. Furthermore, it enhances the bio capacity of the land. The study of Lee (2011) found a positive link between green growth and bio capacity however, there is an inverse connection between green growth and ecological footprint. They further recommended that green growth assures economic and environmental sustainability. Moreover, green growth is promoted through advancement in environmentally friendly technologies (Su et al. 2020; Umar et al. 2020). Green technology has an inevitable role to reduce CO2 emissions and enhance bio capacity by sustaining green economic growth (Sohag et al. 2019; Yin et al. 2015). Above mentioned studies proved that green growth has a significant role to reduce ecological footprint and to strengthen bio capacity. On behalf of existing literature, we may predict that.

H2

There is a significant connection between green growth and load capacity factor.

2.3 Green technology innovation and environmental sustainability

Technological advancement has been considered a mitigating factor of pollution emission worldwide because it has been acknowledged as a significant and new determinant in recent studies on ecological sustainability (Chien et al. 2021a, b, c; Chien et al. 2022a, b; Nawaz et al. 2021a, b). Numerous works in previous literature have shown a relationship between technological advancement and environmental quality. For instance, the work of Long et al. (2018) found how technological innovation affects GHG emissions in China throughout 1997–2014. They concluded that technological advancement has a significant contribution to improve the environmental quality which discloses an inverse connection between innovation and CO2 emissions. However, the study of Yii and Geetha (2017) reported that technological advancement has a negative influence on CO2 emissions in Malaysia over the period from 1971 to 2013. Similarly, Dauda et al. (2021) noted that an innovation leads to increased pollution emissions but square of innovations decreased GHG emissions in the context of 9 African countries. However, they found an inverted U-shaped liaison in Morocco, South Africa, and Egypt. In the same vein, the work of Hodson et al. (2018) suggested that innovation declines CO2 emissions by providing efficient energy consumption and cost-efficient way of producing lesser CO2 emissions. Cheng et al. (2022) found that green technological innovations play a significant role to curb CO2 emissions. The study of Hao et al. (2021b) considered STRIPAT and EKC frameworks while finding the relationship among innovations, economic development, and environment in 25 Asian under-developed economies throughout 1998–2019. Their findings highlighted that innovations in green technology and economic development help to protect the environment while mitigating CO2 emissions. They further corroborated that EKC exists in Asian economies. Similarly, the work of Omri and Hadj, (2020) was done to find the effect of foreign investment, governance, and innovations on CO2 emissions while considering data of 23 emerging economies from 1996 to 2014. They employed the GMM estimation technique and found an inverse relationship between technological innovations and load capacity factor. Relevant works conducted by (Khan et al. 2020a, 2020b; Shahbaz et al. 2020). However, there exist some studies which found a positive link between technological innovations and ecological footprint. It also means that technological advancement accelerates the emissions of GHG. For example, the study of Ganda (2019) explored the impact of technological innovation on secretions of lethal gases where he found a positive association between innovations and CO2 secretion in the context of OECD nations. Similarly, the study of Yu and Du (2019) observed a positive connection between innovations and CO2 emissions. In brief, above mentioned works have revealed both positive and negative link between technological innovation and load capacity factors. Therefore, we may hypothesize that.

H3

There is a significant relationship between green technological innovation and load capacity factors.

2.4 Economic growth and environmental sustainability

The nexus between economic development and load capacity factor has been discussed in the literature of environmental economics and financial economics by using different proxies. A recent study of Awosusi, et al. (2022a, b) found an inverted U-shaped liaison between economic growth and environmental degradation in the context of BRICS economies from 1990 to 2018 by employing panel quantile regression. Similarly, Adebayo et al. (2022) found the same results as Awosusi et al. (2022a, b) study in the context of MINT economies. Moreover, the study of Kirikkaleli et al. (2022) highlighted the environmental degradation determinants in the context of Chile throughout 1985–2018 by considering ARDL technique. Their outputs unveil that a boost up in economic growth leads to devastating environments in the form of ecological footprint and bio capacity. In addition, the work of Alola et al. (2022) explored the impact of real growth and CO2 secretions in China by employing QQ technique. They further revealed that ecosystem deterioration is caused by an increment in GDP in the context of China. There are some other studies which described that an increment in GDP growth leads to bringing environmental harmony. For instance, the work of Usman et al. (2020) noted that an improvement in GDP growth appeases environmental devastation by mitigating ecological footprint in the USA. Similarly, a study by Adebayo et al. (2022) concluded that economic growth abates environmental degradation in the context of Sweden. On behalf of above studies, we may expect positive, negative, and non-linear relationships between economic development and load capacity factors. Therefore, we may predict that.

H4

There is a significant relation between GDP growth and load capacity factor.

2.5 Trade openness and environmental sustainability

In this competitive era, almost every economy has been able to do business across boundaries as a result of globalization. Over the time, different studies in the current literature have investigated the liaison between trade liberalization and environmental quality. The study of Nasir and Rehman (2011) and Ozturk and Al-Mulali (2015) described that an upsurge in economic activities among the countries lead to rise in pollution emissions which resultantly decrease the bio capacity. However, the work of Hossain (2011) and Dogan and Seker (2016) found a direct connection between trade openness and carbon secretions which further disclose that an occurrence of business activities mitigates CO2 emissions which strengthen the supply side of nature. An increment in export activities hoist CO2 emissions (Atici 2012). The study of Shahbaz et al. (2014) examined an exciting liaison among trade openness, income, energy consumption, and GHG emissions by employing ARDL estimation technique. They explored a positive connection between energy consumption and trade liberalization that speeds up pollution emissions in the long run. Moreover, the study of Chien et al. (2021a) noted that energy utilization and trade openness are significant determinants of CO2 emissions. Similarly, the work of Zameer et al. (2020) and Huang et al. (2022) found the impact of trade openness, technological advancement, FDI influx, economic development, and energy utilization on ecological footprint in the context of India throughout 1985–2017. They used VECM and ARDL estimation techniques to run the regression and observed that trade openness and energy utilization affect lethal gases (CO2 emissions) positively however, technological advancement and an influx of FDI affect CO2 emissions negatively. Thus, on the basis of existing literature, we may suppose that.

H5

There is a significant relationship between trade liberalization and load capacity factor.

2.6 Human capital and environmental sustainability

The factors of production, which include land, labour, capital, and entrepreneurship, are the major building stones of an economy. All factors are important, but labour has a significant contribution in economic development and growth. Although human capital has a history of 2.5 centuries, this concept began to be utilized in empirical works and economic analysis in the 1960s (Mincer 1981). It comprises knowledge, health, and skills that people achieved over their whole life. Moreover, the study of Zafar et al. (2019) considered the USA to explore the impact of human capital on environmental degradation throughout 1970–2015 by employing ARDL technique. They found that human capital has a negative impact on the ecological footprint that shows that human capital contributes to enhancing the environmental quality by mitigating ecological footprint. Similarly, the study of Ahmed and Wang (2019) found that human capital considerably lessens ecological footprint by validating EKC hypothesis. Ahmed et al. (2020a) also found a direct link between human capital and environmental quality in G7 economies. They observed that human capital improves environmental quality by enhancing bio capacity. In another study of Ahmed et al. (2020b) found similar results that human capital has a positive effect on ecological quality. Based on above-mentioned works, we may hypothesize that.

H6

There is a positive and significant relationship between human capital and load capacity factor.

2.7 Literature gap

To sum up, the analysis of the aforementioned studies found a broad range of conflicting and inconclusive findings concerning the relationships between energy efficiency, green growth, green technological innovation, and environmental quality. Additionally, there is a dearth of research that explores the relationship between energy efficiency, green growth, and green technological innovation in a single study, specifically within the context of the top 10 manufacturing nations. Furthermore, most of the earlier studies frequently utilized CO2 emissions and EFP as indicators of environmental sustainability. Contrary to the current literature results, there is a noticeable absence of literature that comprehensively delves into the influence of energy efficiency, green growth, and technological innovation on the load capacity factor, especially within the framework of the top 10 manufacturing economies. Consequently, this current research endeavours to address this gap in the literature by examining the impact of energy efficiency, green growth, green technology innovation, economic growth, trade openness, and human capital on the load capacity in the top 10 manufacturing countries.

3 Material and methods

3.1 Data

Enhancing environmental quality has emerged as a fundamental concern for many nations in this age of industrialization and globalization as they seek to fulfill their commitments to achieving sustainable development goals. The study utilized data from various databases to examine the empirical relationship between energy efficiency (EE), green growth (GG), and green technological innovation (GTI) concerning the interplay with the load capacity factor (LCF). Additionally, the study aimed to investigate the impact of control variables such as economic growth (GDP), trade openness (TRD), and human capital (HC) on LCF. The relevant data sources for this analysis were as follows: LCF data in global hectares (BC/EFP) was obtained from the Global Footprint Network (GFN 2023). Green growth and green technological innovation datasets were sourced from the Organization for Economic Cooperation and Development (OECD 2023). Energy efficiency, economic growth, and trade openness data were obtained from the World Development Indicators provided by the World Bank. The human capital data, which is computed by considering education schooling years and returns to education as key factors and was acquired from the Penn World Tables. The analysis focused on the top ten manufacturing countries, namely China, France, Italy, Germany, India, USA, UK, Korea, Mexico, and Japan, for the period between 1990 and 2019. The dataset was limited to this timeframe due to availability of data. Table 1 briefly explains the study variables, symbols, unit of measurement, data sources, and relevant literature.

Table 1 Description of variables
Full size table

3.2 Model and theoretical underpinnings

In light of the load capacity factor idea discussed in recent studies by (Shang et al. 2022; Ni et al. 2022; Pata and Balsalobre-Lorente 2022; Li et al. 2023; Pata and Ertugrul 2023), this study examined the impact of energy efficiency, green growth, green technology innovation, economic growth, trade openness, and human capital on the environmental quality (load capacity factor) in the top 10 manufacturing countries. The adaptation of the Cobb–Douglas production function into the environmental quality function is described as follows:

$$LCF = f \left( {EE, GG, GTI, GDP, TRD, HC} \right)$$
(1)

In this context, LCF represents the load capacity factor, which is calculated as the bio capacity divided by the ecological footprint (in global hectares per person), serving as a measure to assess environmental quality, EE indicates energy efficiency, GG explains green growth GTI refers to green technological innovation,, TRD trade openness, GDP is economic growth, and HC is human capital respectively. The functional form of the model is as follows:

$$LCF_{it} = \varphi_{0} + \varphi_{1} EE_{it} + \varphi_{2} GG_{it} + \varphi_{3} GTI_{it} + \varphi_{4} GDP_{it} + \varphi_{5} TRD_{it} + \varphi_{6} HC_{it} + \varepsilon_{it}$$
(2)

where \({\varphi }_{0}\) explains the intercept, \({\varphi }_{1}\) to \({\varphi }_{6}\) are coefficients of explanatory and control variables, while \({\varepsilon }_{it}\) indicate error terms or residuals. \(i\) denotes the time period of this study (from 1990 to 2019), whereas \(t\) indicates the countries or cross-sections (top 10 manufacturing countries).

The growing environmental challenges associated with development have sparked an increased interest in research linking environmental quality to economic growth and other factors. This ongoing research seeks to address the complexities of environmental challenges by considering both human consumption patterns and the planet's capacity to regenerate resources, thus promoting more sustainable and informed decision-making. On the other hand, enhancing energy efficiency is considered the best approach to improving environmental quality because a rise in energy efficiency promotes energy savings, ultimately leading to an increase in the LCF. Additionally, by improving energy efficiency, the negative environmental impacts of energy consumption can be minimized (Pata and Samour 2022). Similarly, increasing energy efficiency means utilizing less energy for a given level of gross domestic product (GDP), resulting in reduced energy consumption, and improved environmental quality (e.g., Yao et al. 2021; Shahbaz et al. 2015; Shokoohi et al. 2022; Tajudeen et al. 2018). Empirically, a significant increase in energy efficiency corresponds to an enhanced performance of LCF, establishing a positive relationship between the two factors, such as \(\frac{LCF}{EE}\)> 0.

Green growth, which falls within the framework of sustainable economic development, aims to address resource allocation and environmental degradation issues by promoting coordinated socio-economic progress and transforming consumption and production patterns (Hao et al. 2021a). By leveraging environmentally friendly technologies and innovations in energy generation green growth facilitates efficient energy production and conservation of natural resources, ultimately fostering a higher-quality environment (Umar et al. 2020; Su et al. 2020). With its potential to mitigate environmental degradation and promote energy savings, green growth has gained widespread recognition as a strategy for ecological control (Sandberg et al. 2019; Ling Guo et al., 2017). Consequently, a positive nexus has been primarily advanced in the literature. Hence, the hypothetical nexus is depicted as, \({\boldsymbol{\varphi }}_{2}\) \(\frac{LCF}{GG}\)> 0.

Green technology innovation is a form of technological advancement directly associated with environmental protection. Its implementation reduces the negative impact on the environment and enhances the competitiveness of businesses, leading to a win–win situation where both environmental and economic benefits are achieved (Paramati et al. 2022; Doğan et al. 2022). By reducing ecological costs and introducing environmentally friendly technologies, green innovation ensures environmental security by mitigating ecological damage (Galli et al. 2020; Yilanci et al., 2019). Several studies have examined the relationship between green technological innovation and environmental quality (Bashir et al. 2022; Saleem et al. 2022; Wen et al. 2021; Bilal et al. 2021; Hao et al. 2021a). Based on these arguments, we propose a direct link between GTI and LCF, implying that an increase in green innovation will positively impact LCF, as follow \({\boldsymbol{\varphi }}_{3\boldsymbol{ }}\) is positive \(\frac{LCF}{GTI}\)> 0.

Further, this research is grounded in the theoretical aspects of economic growth, which in the context of the load capacity factor is referred to as the Load Capacity Curve according to studies by (Dogan and Pata 2022; Pata and Kartal 2023). These studies propose that LCF serves as a measure of environmental quality. Consequently, they argue that there is a negative correlation between LCF and the initial stages of economic growth, but as economic growth progresses, LCF increases at a higher level. We expect to observe a U-shaped curve between the load capacity factor and economic growth. Empirically, as economic growth significantly rises, there is a corresponding decrease in LCF, thus establishing this relationship \({\boldsymbol{\varphi }}_{4}\) is negative one such as \(\frac{LCF}{EG}\)< 0.

In recent decades, the proportion of international trade has increased in countries worldwide, particularly in developed economies and manufacturing nations. This increased trade activity has led to accelerated economic growth and contributed to environmental degradation by intensifying production and increasing the volume of goods traded (e.g., Apergis et al. 2018; Zhang et al. 2017). Consequently, regardless of the production process, technology employed, or quality of export products, the expansion of trade has been associated with a more significant environmental impact (Dogan and Turkekul 2016; Zhang et al. 2017). Therefore, we expect a negative relationship between trade openness and load capacity factor (LCF), implying that as trade openness increases, LCF \({\boldsymbol{\varphi }}_{5}\) is likely to decrease, \(\frac{LCF}{TRD}\)< 0.

Additionally, human capital (HC) plays a crucial role in fostering environmental awareness, leading individuals, and organizations to adopt strategies for conserving energy and resources (Desha et al. 2015). Chankrajang and Muttarak (2017) demonstrates that education encourages people to embrace green energy products and enhances environmental quality. Zen et al. (2014) find that education promotes active participation in recycling activities. Furthermore, Desha et al. (2015) highlight how education fosters adherence to environmental laws, while Godoy et al. (1998) show its effectiveness in curbing deforestation trends. These collective findings underscore the significant impact of education on driving positive environmental outcomes. As a result, the overall quality of the environment improves due to advancements in HC. Consistent with previous studies (Zhang et al. 2021; Zafar et al. 2019; Ahmed et al. 2020a; Hao et al. 2021a; Lin et al. 2021), we hypothesize a positive relationship between HC and LCF as a measure of environmental quality the anticipated sign of \({\boldsymbol{\varphi }}_{6\boldsymbol{ }}\) is positive \(\frac{LCF}{HC}\)> 0.

3.3 Empirical strategy

The empirical analysis of the present study consists of six steps. In the first step, we investigate the parameters slope coefficient heterogeneity (SCH). The second step examines the issue of cross-sectional dependence (CSD). The third step is to evaluate the stationarity features of the indicators. The long-term cointegration is probed in the fourth step of the analysis. In the fifth step, we analyze the long- and short-run relationships among the variables. The final step uncovers the direction of causal association between variables. More specifically, Fig. 2 displays the empirical strategy of the present study.

Fig. 2
figure 2

Empirical strategy of the present study

Full size image

3.3.1 Slope heterogeneity

The first step of the estimation procedure is to ensure that the parameters estimated in the model are not biased or inconsistent, the slope homogeneity test is frequently used in panel regression analysis to determine the presence of heterogeneous coefficients (Compello et al. 2019). In this study, we use a modified version of the Swamy (1970) model developed by Pescaran and Yamagata (2008) to test for the heterogeneity of slope coefficients. Based on the sample size, the authors propose two different test statistics shown in Eqs. (3) and (4), respectively.

$$\tilde{\Delta }sch = \left( N \right)^{1/2} \left( {2K} \right)^{{ - \left( {1/2} \right)}} \left( {\frac{1}{N}\tilde{s} - K} \right)$$
(3)
$$\tilde{\Delta }Adj - sch = \left( N \right)^{1/2} \left( {\frac{2K(T - K - 1}{{T + 1}}} \right)^{ - 1/2} \left( {\frac{1}{N}\tilde{s} - 2K} \right)$$
(4)

where SCH and Adj-SCH represent the delta and adjusted delta tildes, respectively, for which the H0 presumes the slope coefficients are homogeneous.

3.3.2 Cross-sectional dependence

The occurrence of cross-sectional dependence (CSD) is a frequent issue when dealing with panel data. Traditional panel econometric approaches make the assumption of cross-sectional independence. However, evidence suggests that a shock in a single country or region could affect the macroeconomic performance of multiple countries or regions due to their shared characteristics and geographical proximity. This fundamental issue of CSD has been overlooked by conventional approaches, which may deteriorate the effectiveness of panel data and lead to erroneous outcomes (Pescaran 2004; Chudik and Pescaran 2015; Flores 2019). Therefore, we employ the CSD tests proposed by Pescaran (2004) and Pescaran (2015) prior to model estimation to ensure that there is no CSD. The CSD can be determined as follows:

$$CSD = \sqrt {\frac{2T}{{N\left( {N - 1} \right)}}} \mathop \sum \limits_{i = 1}^{N - 1} \mathop \sum \limits_{j = i + 1}^{N} T_{ij}$$
(5)

When conducting a panel regression, the CSD test requires the alternative hypothesis to presume interdependence, whereas the null hypothesis does not.

3.3.3 Stationarity tests

After the observation of CSD, the next task is to analyze the integration features of the variables. In the presence of CSD and SCH, the traditional first-generation unit root test does not perform well and provides inaccurate estimates (Danish and Wang 2019). The present study uses the 2-s-generation unit root test suggested by Pesaran (2007) namely "Cross-Sectional Im-Pesaran-Shin" (CIPS) and "Cross-Sectional Augmented Dicky-Fuller" (CADF). These tests tackle the CSD and SCH issues and produce valid and more reliable outcomes. The CIPS and CADF mathematical formulas are as follows:

$$\Delta Y_{it} = \alpha_{i} + \beta_{i} Y_{i,t - 1} + c_{i} \overline{Y}_{t - 1} \mathop \sum \limits_{j = 0}^{p} d_{i} \Delta \overline{Y}_{t - j} + \mathop \sum \limits_{j = 0}^{p} \theta_{ij} \Delta \overline{Y}_{i,t - j} + \varepsilon_{it}$$
(6)
$$\widehat{CIPS} = N^{ - 1} \mathop \sum \limits_{i = 0}^{n} CDF_{i}$$
(7)

3.3.4 Cointegration test

The next step is to probe the long-term cointegration among the variables after the establishment of CSD and variable stationarity. We employed the powerful second-generation test provided by Westerlund’s (2007) to test the study’s variables. This test is more appropriate and selected because it is applicable even in the presence of CSD and does not suffer from the effects of a common factor in producing valid and reliable results (Kapetanios et al. 2011). Westerlund’s test statistics are given as follows:

$$G_{t} = \frac{1}{N}\mathop \sum \limits_{i = 1}^{N} \frac{{\alpha_{i} }}{{SE\left( {\alpha_{i} } \right)}}$$
(8)
$$G_{a} = \frac{1}{N}\mathop \sum \limits_{i = 1}^{N} \frac{{T\alpha_{i} }}{{\alpha_{i} \left( 1 \right)}}$$
(9)
$$P_{t} = \frac{\alpha }{SE\left( \alpha \right)}$$
(10)
$$P_{t} = T\alpha$$
(11)

3.3.5 CS-ARDL model

The current study employed the "Cross-Sectional Autoregressive Distributed Lag" (CS-ARDL) estimator, initially developed by Pescaran and Smith (1995) and later refined by Chudik and Pesaran (2015). The occurrence of CSD is a common issue in the panel data, especially when the cross-sections all come from the same geographical area and are subject to the same economic and political stresses, and so on. The existence of CSD in the data set may lead to inaccurate regression outcomes (Tao et al. 2021). The CS-ARDL technique well addresses the issues of CSD and SCH. This approach also takes into account the weak exogeneity, structural breaks, and endogeneity issues that arise due to the lag of the independent variables. Further, when the time period exceeds the number of cross-sections, as in the present study, this estimator is also considered valid (Westerlund 2007). Based on common correlated effects, the CS-ARDL mathematical demonstration begins with the following step:

$$LCF_{i,t} = \mathop \sum \limits_{i = o}^{{p_{u} }} \alpha_{i,t} LCF_{i,t - 1} + \mathop \sum \limits_{i = 0}^{{p_{v} }} \delta_{i,t} V_{i,t} + \varepsilon_{it}$$
(12)

Equation (1) is the depiction of the ARDL model. However, due to the existence of CSD in the dataset, following Eq. (12) may result in inaccurate estimates and conclusions. This is because Eq. (12) completely ignores the interdependencies that exist between the cross-sections, which in this case are the top 10 manufacturing economies. As a result, Eq. (13) offers an expanded form of Eq. (12) by using the cross-section averages of the study’s predictor variables, based on the theoretical recommendations made by Chudik and Pescaran (2015). In addition, it has been observed that using average cross-sections is helpful in reducing the influence of CSD.

$$LCF_{i,t} = \mathop \sum \limits_{i = o}^{{p_{u} }} \alpha_{i,t} LCF_{i,t - 1} + \mathop \sum \limits_{i = 0}^{{p_{v} }} \delta_{i,t} V_{i,t} + \mathop \sum \limits_{i = 0}^{{p_{w} }} \gamma_{i} \overline{W}_{t - 1} + \varepsilon_{i,t}$$
(13)
$$\overline{W}_{t - 1} = \left( {\overline{LCF}_{t - 1} ,\overline{V}_{t - 1} } \right)$$

To be more explicit, \({LCF}_{i,t}\) displays the values of the explained variable, which is the load capacity factor in the present study, while the values of the explanation variables, which are titled (EE, GG, and GTI.), are covered by \({V}_{i,t}\). Further, \({p}_{u}\), \({p}_{v}\), \({p}_{w}\) can be utilized to ascertain the lag values for underlying variables. In addition, \({\overline{W} }_{t-1}\) expresses the average values for the explained, explanatory, and control variables, which eliminates any issue that may arise as a result of CSD caused by spillover effects (Liddle 2018). The CS-ARDL approach offers long-run estimates of the given set of variables, which are derived by using the values of the coefficients of their short-term relationship. In addition to this, the long-run estimates along with the associated mean group estimators are reported in the following manner:

$$\hat{\theta }_{CD - ARDL.i} = \frac{{\mathop \sum \nolimits_{I = 0}^{Pv} \hat{\delta }_{Ii} }}{1 = \Sigma = 0}$$
(14)

Further, the mean group is defined as:

$$CIPS = \hat{\theta }_{MG} = \frac{1}{N} \mathop \sum \limits_{i = 1}^{N} \hat{\delta }_{i}$$
(15)

Based on this, the short-term coefficient approximation is reported as:

$$\Delta LCF_{i,t} = \theta_{i} \left[ {LCF_{i,t - 1} - \theta_{i} V_{i,t - 1} } \right] - \mathop \sum \limits_{i = o}^{{p_{u - 1} }} \alpha_{i,t} \Delta_{i} LCF_{i,t - 1} + \mathop \sum \limits_{i = 0}^{{p_{v} }} \delta_{i,t} \Delta_{i} V_{i,t} + \mathop \sum \limits_{i = 0}^{{p_{w} }} \gamma_{i} \overline{W}_{t - 1} + \varepsilon_{i,t}$$
(16)

where

$$\Delta_{i} = t - \left( {t - 1} \right)$$
(17)
$$\hat{\tau }_{i} = - \left( {1 - \mathop \sum \limits_{i = o}^{{p_{u} }} \alpha_{i,t} } \right)$$
(18)
$$\hat{\theta }_{i} = \frac{{\mathop \sum \nolimits_{i = o}^{{P_{v} }} \hat{\delta }_{i,t} }}{{\hat{\tau }_{i} }}$$
(19)
$$\hat{\theta }_{MG} = \frac{1}{N} \mathop \sum \limits_{i = 1}^{N} \hat{\delta }_{i}$$
(20)

The CS-ARDL framework uses the term "Error Correction Mechanism" (ECM(-1)), which is analogous to the term "Pooled Mean Group" (PMG), to describe the rate of speed of adjustment when an economy is moving towards an equilibrium point.

3.3.6 Robust AMG and CCEMG

In addition, the present study uses the two additional approaches to verify the robustness of the CS-ARDL model. Using conventional methods when CSD and SCH remain in a study can lead to misleading conclusions (Chudik and Pescara 2015; Liddle 2018). In light of the aforementioned problems and structural break, it is advised to employ the AMG ("Augmented Mean Group") by Eberhardt et al. (2010) and the CCEMG ("Common Correlated Effect Mean Group") by Pesaran (2006). Despite the presence of unobserved common components and non-stationarity, the findings generated by using these two methods tend to be more stable. Specifically, CCEMG allows for the averages of the studied variables in order to eliminate the spillover effects that arise due to the presence of CSD, and it also addresses time-variant unobservables in addition to addressing the SCH without highlighting identification. Furthermore, AMG serves as an alternative to CCEMG by including the dummies on an annual basis to address any unobservable along with the issues of CSD, SCH, and other discussed concerns that produce inaccurate estimates.

3.3.7 Granger causality

In the final step, it is imperative to identify the direction of causal relationship between the variables in addition to the application of CS-ARDL, AMG, and CCEMG. However, the results from these techniques provide useful information, but these approaches lack the directional flow between the variables, which is crucial for drawing policy conclusions. To investigate the directional flow and causal relationship between the variables, this study used the panel causality test made popular by Dumitrescu and Hurlin (2012). When the error terms are cross-sectionally dependent, this test is especially useful and important. The formula for the test is as follows:

$$Z_{i,t} = \sigma_{i} + \mathop \sum \limits_{k = 1}^{p} \alpha_{i}^{k} Z_{i, t - k} + \mathop \sum \limits_{k = 1}^{p} \psi_{i}^{k} T_{i, t - k}$$
(21)

where \({\sigma }_{i}\) represents the intercept, k denotes the order of lag, \({\alpha }_{i}\) and \({\psi }_{i}\) are the autoregressive coefficients.

4 Empirical results and discussion

4.1 Descriptive statistics

Understanding the characteristics of the data set used in this work is made more accessible through the descriptive statistics. Table 2 offers a comprehensive breakdown of descriptive information for this investigation. Table shows that mean value for all the variables ranges from 10.009 to −1.284. More precisely, GDP has the highest mean value (10.009) followed by TRD (3.748), GTI (2.139), EE (2.104), HC (1.076), GG (0.579), and LCF (−1.284). Besides this, GDP, TRD, and GG have the highest maximum values. Furthermore, results illustrate that all the variables are negatively skewed.

Table 2 Descriptive statistics
Full size table

4.2 Slope heterogeneity results

The next phase of the analysis is testing the parameters SCH. This study relied on Pescaran and Yagamat's (2008) test for this aim. The absence of heterogeneity among the slope parameters is supported by the null hypothesis (H0). However, the existence of slope variation in the slope coefficients is a central tenet of the alternative hypothesis. Table 3 displays the results, which indicate that the estimated test scores for delta and adjusted delta tildes are (10.374) and (12.114), respectively. These results indicate the rejection of H0 and conclude that H1 is accepted because there is substantial evidence of slope heterogeneity in the data.

Table 3 Slope heterogeneity test
Full size table

4.3 Cross-sectional dependence results

As a result of the discussion in the preceding section, it has been determined that the CSD test is of the utmost significance in the panel data in order to produce accurate outcomes. Therefore, the presence of CSD is evaluated using the Pescaran (2007) test in this investigation. In contrast to H1, which argues that observational cross-sectional units are connected, H0 assumes that there is no cross-sectional dependence. The results reported in Table 4 reject H0 (absence of CSD) at the 1% significance level and support H1. As a result, the presence of CSD has been demonstrated and proven. The existence of CSD and SCH necessitates the use of 2nd-generation stationarity and cointegration tests.

Table 4 CSD results
Full size table

4.4 Stationarity results

Since the results of Pescaran and Yagmata's (2007) SCH and Pescaran's (2007) CSD tests confirm the issue of CSD and CSH, the use of the first-generation unit root test may lead to an inaccurate estimate in the existence of these issues. Therefore, we examined the stationarity characteristics of all the variables by using 2nd-generation unit root tests, namely CIPS and CADF, which overcome the issues of CSD and SCH and provide reliable outcomes. The H0 of both tests specifies the non-stationary behaviour of data, whereas H1 accepts that data is stationary. The results of both tests are shown in Table 5, which endorses that the underlying variables are integrated at I(0) and 1(1) and none of the variables are stationary at 1(2).

Table 5 CIPS and CADF outcomes
Full size table

4.5 Cointegration results

In addition, testing for a long-term cointegration relationship is essential before investigating the influence of a set of explanatory variables on an explained variable. This study performed the Westerlund’s (2007) cointegration test to uncover the long-term linkages between the variables. The premise of no cointegration underlies the H0 of the stated test, while the H1 confirms the existence of cointegration. The results of the cointegration test summarized in Table 6 illustrate that all of the test statistics (Gt, Ga, Pt, and Pa) are significant at the 1% and 5% level. These outcomes demonstrate that the null statement is false and approve the acceptance of H1. As a result, we conclude that a long-term cointegration relationship exists among the research variables, i.e., EE, GG, GTI, GDP, TRD, HC, and LCF. We next proceeded to estimate the required regression model according to the objective of the study.

Table 6 Cointegration results
Full size table

4.6 CS-ARDL estimates

After establishing the existence of cointegration, the study then looks into the long- and short-run dynamics of green growth, energy efficiency, green technology innovation, economic growth, trade openness, human capital with load capacity factor. The CS-ARDL evaluation developed by Chudik and Pesaran (2015) is used for this investigation. The long- and short-term CS-ARDL estimates are shown in Table 7. The long-run outcomes illustrate that, as expected, energy efficiency soars the level of LCF in the top 10 manufacturing countries. It is evident that with other explanatory variables remaining constant, an escalation of 1% in the EE surges LCF by (0.252%) in the long-term. This findings provides strong evidence for EE as a long-term solution to reduce environmental degradation in top 10 manufacturing countries. This positive impact of EE on the environmental quality is also confirmed by (Abudureheman et al. 2022; Mirza et al. 2022; Tajudeen et al. 2018; Tajudeen. 2015; Jahanger et al. 2023; Qing et al. 2023; Adebayo and Ullah 2024; Sarkodie and Strezov 2019; Wang et al. 2018; Akram et al. 2020b). They documented that EE is best for improving environmental quality. Energy production via the combustion of fossil fuels accounts for the vast bulk of worldwide emissions of greenhouse gases. By enhancing energy efficiency, we can reduce the quantity of energy needed to carry out certain activities or deliver certain services. As a result, fewer fossil fuels are used and fewer greenhouse gases, especially carbon dioxide (CO2), are released into the atmosphere. Furthermore, limited natural resources such as oil, coal, and natural gas are mostly utilized in the production of energy. Increasing efficiency saves energy since it minimizes the need for it, which in turn reduces the amount of energy used. As a result, energy efficiency safeguards ecosystems, maintains biodiversity, prevents habitat damage caused by resource extraction, and improves sustainability.

Table 7 CS-ARDL results
Full size table

Besides, the long-run estimated coefficient of green growth exerts a positive and statistically significant impact on the LCF, inferring that a 1% upsurge in GG in the underlying countries improves the environmental sustainability by enhancing the level of LCF by (0.065%) over the tested period. Parallel outcomes are also supported by (Chien et al. 2021a, b, c; Saleem et al., 2022; Hao et al. 2021a; Wei et al. 2024). It has been widely acknowledged that green growth is an important method for attaining sustainable development and reducing poverty. Because it has the capacity to address both the environmental sustainability and the economic development of a country at the same time. Green growth encourages the conservation of natural resources while simultaneously fostering economic development. Further, it pushes for more eco-friendly methods of manufacturing and consumption, like reducing waste and using less energy. Green growth aids conservation efforts and reduces pollution levels by cutting down on both resource exploitation and waste production.

Similarly, the long-term observed findings of the CS-ARDL model demonstrated that GTI had a significant positive impact on the LCF among the top 10 manufacturing countries over the examined period. The research outcomes portrayed that a 1% positive rise in GTI led to a positive impact on environmental health by increasing the LCF by (0.285%). This positive effect of green technology innovation on LCF supports the idea that the rise and development of GTI trigger ecological sustainability by increasing the level of LCF. Sun et al. (2020) argued that adoption and development of green technological innovation are among the many tools and strategies needed to address environmental and energy challenges. The development of clean energy and the increased utilization of renewable resources are two additional ways in which green technological advancements boost environmental quality (Kongbuamai et al. 2020). The outcome of our study is also consistent with the earlier studies (Hao et al. 2021a; Meng et al. 2022; Herzer 2022; Javed et al. 2023a; Jahanger et al. 2023; Koseoglu et al. 2022; Sadiq and Wen 2022; Bekun 2024; Aydin et al. 2024). This infers that development of GTI in the top 10 manufacturing countries play a vital role in improving ecological health.

On the other hand, long-run estimations of the CS-ARDL model demonstrated a statistically significant inverse relationship between economic expansion (GDP) and LCF among the top 10 manufacturing countries over the examined period. More specifically, a rise in GDP of 1% appears to result in (0.645%) decrease in the LCF. This finding is justified on the ground that the underlying countries have experienced enormous growth over the past couple of decades. The manufacturing and transportation sectors of the underlying economies significantly rely on highly carbon-intensive energy sources such as fossil fuels. The significant economic gain is the result of massive energy consumption, production, and excessive utilization of natural resources, which in turn increases the ecological footprint, increases waste generation, and reduces the sustainability limit. As a result, it infers that the economic growth strategies of the underlying economies contradict the environmental conservation objectives. This negative connection between GDP and LCF is consistent with the outcomes of a number of investigations conducted by (Shahbaz et al. 2013, 2021; Pata et al. 2023a; Khan et al. 2023a, b; Shang et al. 2022; Pata and Balsalobre-Lorente 2022; Ni et al. 2022; Akadiri et al. 2022; Awosusi et al. 2022a, b; Li et al. 2023; Javed et al. 2023b). They all documented the detrimental impact of GDP on the quality of the environment.

Likewise, the long-run analysis also portrays that TRD is negatively linked with LCF over the examined period. Apparently, the outcomes illustrate that keeping all the other research factors constant, a rise of 1% in trade activities in underlying countries decreases the LCF by (0.435%). This negative impact of TRD on LCF indicates that a significant rise in the level of trade activities causes environmental damage by lowering the LCF level in the investigated countries. The findings of the current study lend credence to the outcomes of previous studies (Javed et al. 2023a, b; Akhayere et al. 2023; Shahbaz et al. 2017a, b; Ali et al. 2020; Hadom and Fuinhas 2020). The mentioned studies confirmed that trade openness negatively affects ecological sustainability. However, the results are disagreed with the claims suggested by (Huilan et al. 2022 and Pata et al. 2023b). This negative association between trade openness and LCF is justified by the fact that the underlying countries are the world's largest industrialized economies and contribute at a larger scale to production and consumption activities. Therefore, an increase in manufacturing and trading activities demands more utilization of energy consumption, land, water, and other natural resources. The rising demand for natural resources may place further stress on ecosystems, which may result in the loss of habitat, the cutting down of trees, an increase in ecological footprint, and, as a result, reduce the sustainability limit. Thus, increased commercial activity has unfavourable consequences for ecological sustainability. Hence, it can be concluded that augmented trading activities in the top 10 manufacturing countries are harmful for ecological sustainability due to the mounting pressure on biocapacity followed by uncontrolled utilization of ecosystem resources.

Next, contrary to GDP and TRD, human capital has a statistically significant long-term advantageous impact on the LCF. This infers that a rise in HC stimulates the LCF in the top 10 manufacturing countries. According to the estimates, an increase of 1% in HC leads to a (0.632%) intensification in LCF. On the basis of these outcomes, we can conclude that HC escalates the level of LCF and thereby improves environmental quality. Our results are also supported by the research conducted by (Dai et al. 2023; Pata and Ertugural 2023; Pata and Isik 2021; Ahmad et al. 2020a; Jahanger et al. 2022). These studies revealed that HC is an important factor to enhancing environmental quality. However, the empirical outcomes of the present study disagree with the outcomes discovered by (Zhou et al. 2022 and Chen et al. 2022). Our research suggests that environmental consciousness can be raised in society through a high level of education. It is evident that societies with higher levels of education are more likely to advocate for environmental cleanliness and use natural resources cautiously. Increased human capital makes it easier for people to adopt energy-efficient and innovative technologies. The rise in recycling rates, interest in renewable energy sources, and other environmentally positive behaviours have all been related to rising levels of education. In addition, those with more education have the potential to earn more money, which may then be used to purchase more environmentally friendly products.

Further, Table 7 also displays the CS-ARDL short-run estimates to illustrate the relationship between energy efficiency, green growth, green technology innovation, economic growth, trade openness, and human capital with load capacity factor for the top 10 manufacturing countries. The short-term outcomes of the CS-ARDL model are consistent with the long-term ones. The short-term coefficients predict that EE, GG, GTI, and HC improve the level of LCF and enhance environmental quality, while GDP reduces the LCF and degrades the environment. In addition, the significant and negative coefficient value of (ECT) indicates the LCF speed of adjustment towards the long-term equilibrium point due to any shock in the short run. A negative value of ECT, such as (−0.683), indicates that the equilibrium point is regained at a rate of (68.3%) within a year.

4.7 Robustness estimates

After estimating the long- and short-run impacts between the research variables by using the CS-ARDL estimator, It is important to validate the model results for robustness. The present study used the AMG and CCEMG estimators to verify the CS-ARDL estimates following the prior studies (Meng et al. 2022; Amin et al. 2023; Shang et al. 2023). Table 8 reports the findings of both AMG and CCEMG estimates. The outcomes indicate that energy efficiency and green growth have a positive impact on LCF in both of the aforementioned models. This outcome implies that both EE and GG enhance the quality of the environment by increasing the LCF. Similarly, GTI is associated with an increase in LCF, suggesting that adoption and the development of GTI will help to promote environmental quality in sample countries. In addition, results show that GDP and TRD in both models contribute to environmental degradation by lowering the LCF. Finally, the outcomes of human capital provide sufficient evidence that a rise in HC is beneficial to attaining environmental sustainability in the top 10 manufacturing countries. The results of AMG and CCEMG are robust to the outcomes of CS-ARDL finings reported in Table 7. In addition, Fig. 3 presents the summary of CS-ARDL, AMG, and CCEMG estimates.

Table 8 AMG and CCEMG results for robustness
Full size table
Fig. 3
figure 3

Graphical summary of CS-ARDL, AMG, and CCEMG estimates

Full size image

4.8 Granger causality results

In the presence of CSD among the cross-sections, conventional causality tests cannot guarantee reliable causal results. Therefore, the present study employed the Dumitrescu and Hurlin (2012) panel causality approach to evaluate the direction of causal association between the variables in the study. Table 9 provides the empirical outcomes of the causal connection between the variables. The outcomes illustrate a unidirectional association is evident between energy efficiency and LCF, implying that policies implemented to improve EE could significantly improve LCF. Further, a bidirectional causal connection (feedback) running from green growth, green technology innovation, and economic growth to LCF, suggesting that changes in GG, GTI, and GDP could have far-reaching effects on the quality of the environment. Additionally, outcomes indicate one-way causal connections running from trade openness and human capital to the load capacity factor, indicating that variations in these factors would result in significant changes in LCF. The results of causality analysis guarantee the success of CS-ARDL estimates. The results confirm the substantial impact of GG, EE, GTI, GDP, TRD, and HC on LCF in the top 10 manufacturing countries.

Table 9 D-H panel causality outcomes
Full size table

5 Conclusion and policy recommendations

Over the past few decades, countries rapid industrialization, economic growth, rising energy consumption, trading activities, changing lifestyles, and consumption of natural resources at an unsustainable rate have contributed to a variety of environmental problems, including pollution of air and water, depletion of natural resources, loss of biodiversity, a rising ecological footprint, and climate change, amongst others. Despite being among the world’s most industrialized nations, however, the top 10 manufacturing countries are facing tremendous environmental issues. Therefore, regardless of their economic success, these nations have failed to ensure the health of their natural resources. As a result, it is very important for these nations to identify the structural elements that can ameliorate mounting ecological challenges. The role of energy efficiency, green growth, and the development of green technology innovations in achieving social sustainability and environmental degradation reduction goals has become a focal point of debate about the best path towards a more sustainable future. In this perspective, the prime objective of the present study is to evaluate the impact of energy efficiency, green growth and green technology innovation, along with economic growth, trade openness, and human capital, on the load capacity factor of the top 10 manufacturing nations between 1990 and 2019. We used LCF as a measure of environmental viability, which considers both the demand and supply sides of the ecosystem. Before proceeding to the analysis of long- and short-run relationships among variables, we performed some preliminary second-generation econometric tests. These include the Pescaran (2004), Pescaran (2015) and Pesaran, and Yamagata (2008) tests for CSD and SCH issues, respectively. Furthermore, the stationarity qualities of the variables utilized in the analysis are checked by employing the Pesaran (2007) CIPS and CADF unit root tests. However, the long-term cointegration is verified by using Westerlund's (2007) cointegration test. The CS-ARDL approach established by Chudik and Pesaran (2015) is utilized to estimate the long- and short-term coefficients. We also considered the AMG and CCEMG estimators to test the robustness of the long-run coefficients. Finally, the directions of the causal relationships are then inferred by using the D-H causality test. The CS-ARDL model estimates explained that energy efficiency, green growth and green technology innovation have a positive association with LCF. These long-term coefficient values assert that energy efficiency, green growth, and green technology significantly improve the sustainability limit in the top 10 manufacturing countries. On the other hand, economic growth and trade openness reduce the LCF and elevate environmental degradation. Besides, human capital was found to improve the sustainability limit by increasing the LCF. In a similar vein, the short-term results of CS-ARDL supported the long-term findings. Furthermore, AMG, CCEMG, and causality analysis also verified the estimates of the CS-ARDL model. On the basis of the proven association between green growth, energy efficiency, and green technology innovation with load capacity factor, the study concludes that GG, EE, and GTI improve the LCF and encourage environmental sustainability.

The findings of this study have a number of policy ramifications for a sustainable future environment. The results indicate that energy efficiency, green growth, and technological innovation improve environmental quality in the examined countries. The underlying countries are highly industrialized, and their higher energy consumption and resource utilization exert a negative impact on the environment. Thus, the decision-makers in these nations should prioritize policies linked to energy efficiency and the transition towards clean energy. It is imperative that residential and industrial machinery meet the minimum energy efficiency standards and that older technology be phased out. In addition, the construction and transportation industries should be offered tax incentives to encourage them to use less energy. The government should provide financial and policy assistance for initiatives like making zero-emission vehicles and creating energy-efficient buildings. Similarly, providing a line of credit at low interest rates helps attract domestic and foreign investors to invest in an energy-efficient project. In order to achieve the global climate and sustainable development goals, governments should promote energy efficiency regulations and collaborate to promote energy efficiency. They need to work together on initiatives that will increase awareness and encourage the adoption of energy-saving devices. More specifically, governments of rising nations in the sample countries need to devote greater resources to R&D in order to create fresh breakthroughs and continue strengthening partnerships with advanced nations at the forefront of technological innovation.

The findings of this study also suggest that the underlying economies should encourage green growth. Because green growth is an economic expansion strategy that reduces negative impacts on the environment and conserves natural resources, it acknowledges the connection between economic, environmental, and social sustainability. Green growth strategies stress the replacement of fossil fuels with renewable energy sources. This adjustment lessens the effects of climate change and enhances the quality of the environment. In addition, in order to maximize efficiency and reduce waste, green growth advocates for a circular economy. Reducing waste and the need to harvest virgin resources is accomplished through recycling, reusing, and repurposing. A circular economy helps the environment by decreasing waste, conserving materials, and leaving less impact on the planet. Further, to lessen the environmental damage caused by industrialization in the top 10 manufacturing nations, the creation of eco-friendly technologies is crucial. In order to improve energy efficiency and lower their dependence on fossil fuels, the governments of these countries should allocate sufficient funds for R&D in the development of technologies that increase the efficiency of renewable energy consumption and decrease the price of alternative energy sources in order to lessen their reliance on fossil fuels. Besides, policymakers should support the efforts of academics and organizations working to improve energy efficiency.

Further, outcomes illustrate that trade openness harms the environmental quality of the underlying countries. It is imperative that environmental considerations be incorporated into trade laws in order to reduce the adverse effects that an open trading environment can have on the surrounding ecosystem. This can be accomplished through the implementation of various measures, including the incorporation of environmental standards into trade agreements, the promotion of sustainable production and consumption practices, the strengthening of environmental regulations and enforcement, and the support of green technological advancements that reduce resource consumption and pollution during the manufacturing of trade commodities.

The present research has some limitations, which pave the way for crucial new avenues for future research on the topic of environmental sustainability. We used the load capacity factor as an indicator of environmental sustainability, which is a more comprehensive indicator because it considers both the demand and supply sides of our ecosystem. However, a few studies in the literature have utilized LCF as an indicator. This opens the door for further research into the impact of other macroeconomic factors such as urbanization, financial integration, globalization, institutional quality, and environmental policy stringency on the LCF. Another limitation of the present study is that we utilized the linear methodology to examine the impact of green growth, energy efficiency, and green technology innovation on the LCF. Further studies can utilize quantile regression approaches to better understand the roles of GG, EE, and GTI on the various quantiles of LCF. Besides, the present study is limited to the top 10 manufacturing economies. This study can be extended to other regions and blocks such as BRICS, MINT, ASEAN, G-20, and Next-11 countries to get more insights and to design effective policy implications to achieve environmental sustainability.