Exploring the impact of digital technology, circular technology and public management on environmental sustainability in emerging economies Authors: Yifang Wei, Muhammad Sibt e Ali, Mohd Abass Bhat Published: 2025 ABSTRACT Environmental degradation significantly hampers economic sustainability by causing severe economic problems, prompting researchers to focus on environmental sustainability. Global economies aim to achieve net-zero emissions by 2060 to address climate change issues. Achieving this by 2050 depends on the transition from fossil fuels to renewable energy. Recently, the circular economy (CE) has emerged as a key concept for expediting sustainability. Prior research has emphasized CE practices in the corporate sector, overlooking the public sector. This study uses panel IFE and D-CCE tests to analyze the impact of CE and public management on CO2 emissions in E-7 economies from 2000 to 2023, considering GDP per capita, urbanization, digital technologies, and energy transition. It examines long-term cointegration and heterogeneity among variables and employs a panel quantile regression (PQR) model to validate the findings. The results show that digital technology, CE, and public management significantly reduce carbon emissions, while GDP per capita, urbanization, and energy transition are crucial for long-term CO2 emissions. Further analysis indicates that public management mediates the negative impact of digital technology, CE, and energy transition on carbon emissions in E-7 economies. These findings provide policymakers with insights into managing CE, energy transition, digital technology, and public administration to reduce CO2 emissions without hindering economic growth and sustainable development. The practical implications concerning carbon neutrality and structural transitions are also discussed. Introduction Currently, the most pressing challenges are climate change and environmental degradation, which are primarily driven by CO2 and greenhouse gas (GHG) emissions. According to BP statistics from 2019, CO2 emissions in 2018 amounted to 33,890.8 million tons, a significant increase from the 11,190 million tons recorded in 1965, all attributed to fossil fuel consumption. The emission of CO2 has been recorded at its fastest rate in the last seven years, with an average growth of 2.0% per annum, according to BP statistics from 2019. Climate change and environmental degradation have profound and unprecedented effects on human activities and life. Furthermore, it has been noted that human health is facing significant challenges as a result of extreme weather events and air pollution1. To minimize the impact of climate and environmental issues, it is essential for every country to invest in environmental sustainability and the effective utilization of clean energy, circular economy (CE) practices, and innovations in green technology2. Over the past ten years, Garrido-Prada3 have made substantial contributions to the fields of low-carbon economic development, energy technologies, and energy sources, offering valuable insights for various stakeholders and economies. Simultaneously, the global economy is undergoing a severe economic shutdown as a result of the ongoing conflict between Russia and Ukraine4. Nonetheless, decreasing reliance on fossil fuels and decoupling energy consumption from economic expansion may effectively address significant ecological degradation5. The integration of circular economy (CE) principles, public management (PM), and digital technology (DT) is essential for attaining environmental sustainability and promoting sustainable development while minimizing adverse effects on the natural world6. The pursuit of carbon neutrality and ecological sustainability, along with various other sustainable development goals (SDG), has garnered considerable attention from scholars, policymakers, and other stakeholders in both developing and developed economies. However, the emission of CO2 from various energy sources and trade-related into the natural atmosphere affects carbon emissions. Currently, there are several discussions regarding circular economy principles, public management, and digital technology, all aimed at reducing carbon emissions and boosting environmental sustainability. However, in the context of society, the environment, and the economy, CE, PM, and DT have undergone a complete transformation regarding sustainable green solutions. Various nations have implemented a range of strategies and policies concerning the circular economy, green energy, and innovations in green technology to promote the sustainable development of the natural environment. For example, as part of the United Nations (UN) 2030 agenda, China established the Green Technology Bank (GTB) to facilitate and advance green energy initiatives and foster innovation in green technologies7. Similarly, the United Kingdom (UK) initiated a direct financial support program for various distinct industries via a green investment bank aimed at promoting DT and environmental sustainability8. To create a mutually beneficial scenario for both the environment and economy, practices involving CE, PM, and DT are being implemented. In the context of environmental resource utilization, the prospects for innovation and green energy present both detrimental and optimistic aspects9. This study develops an integrated conceptual model linking digital technology (DT), circular economy (CE), public management (PM), and energy transition (ENT) with environmental sustainability outcomes in E-7 economies. The model assumes that DT and CE directly lower CO₂ emissions by improving resource efficiency and reducing waste, while PM strengthens these effects through effective regulation, policy enforcement, and institutional coordination. Moderating terms (PM × DT, PM × CE, PM × ENT) are therefore included to test whether governance enhances technological and circular effects on emission reduction. Figure 2 summarizes this conceptual framework and guides the empirical testing. The circular economy (CE) is acknowledged as a mechanism for attaining environmental sustainability at international, national, and local levels. It has attracted the interest of scholars, policymakers, and multinational corporations10,11. Nonetheless, the circular economy has received comparatively less focus in low- and middle-income nations, with the significant exception of China12. The academic community and global development professionals13 have recently focused considerable attention on CE and leveraged its diverse elements for environmental sustainability. Furthermore, CE is characterized by distinct actions and practices, such as industrial symbiosis, repair, and product sharing14, along with refurbishment, remanufacturing, eco-design, and reuse15. There is a conceptual overlap between the practices, concepts, and principles of the circular economy and other approaches such as industrial ecology and cleaner production. Furthermore, for sustainable consumption and production, as well as the transformation within current systems, practices related to the circular economy are a crucial component16. Scholars argue that the investigation of environmental change and institutions in E-7 countries is inherently intertwined with the scrutiny of policies and economic transformations. In this context, the relationship between the effectiveness of governance and the conservation of a sustainable environment becomes relevant. It offers a persuasive rationale to investigate the importance of good governance in relation to CO2 emissions in E-7 countries. Public administration (PM) at the macro level encompasses three primary functions: the appointment and succession of governments, the capability of the administration to make and apply efficient strategies, and the preservation of citizens’ dignity and the integrity of the establishments that govern financial and public affairs among them17. However, the efficacy of public administration at the macro level may be crucial in formulating appropriate policies that aid in reducing environmental damage. Attaining enhanced environmental governance is essential for protecting environmental integrity and guaranteeing the safe and sustainable use of resources. The concept of government includes several indicators that primarily focus on particular aspects of government practices. Environmental conservation and the appropriate use of natural resources are focal points of public management18. Diverse measures of governance result in unique effects on carbon dioxide emissions19. While the impact of various aspects of governance on the environment may vary, the caliber of institutions significantly affects the decrease in CO2 emissions. Both the direct and indirect influences of institutional quality on carbon dioxide emissions have been investigated. The efficiency of regulatory mechanisms and the influence of environmental organizations significantly contribute to the optimization of air chemistry. Socioeconomic and democratic factors are also associated with environmental quality. Indeed, the importance of institutional quality for the environment has been confirmed by Xinyu20, and implementing efficient environmental governance is advantageous in guaranteeing the sustainable use of resources. Corruption adversely affects environmental quality, both directly and indirectly, by weakening the effectiveness of institutions, promoting rent-seeking behavior, and hindering the effective implementation of environmental policy. Nevertheless, corruption management plays a role in mitigating emissions and has been demonstrated to directly and indirectly lower CO2 emissions by acting as a moderator. Existing literature has established a non-linear relationship between regional quality and the environment, suggesting that a significant degree of regional quality does not result in enhanced environmental efficiency6. Through the promotion of greater justice in wealth and power distribution, public administration contributes to the mitigation of pollution and enhancement of the environment. An effective institutional structure empowers the government to exercise effective control over emission levels. The efficacy of institutions plays a pivotal role in mitigating pollution levels by means of democratic processes and curbing corruption, thus enabling the more effective formulation of regulatory measures targeted at pollution reduction. The rule of law is a fundamental expression of a governance system that effectively limits the consequences of market failures through a well-operating judicial system, property rights, and other associated elements. By implementing suitable environmental restrictions, it is feasible to decrease the pollution level without encountering any difficulty in adhering to the rules. Otherwise, the existence of regulatory loopholes would create difficulties for the enforcement of laws designed to decrease CO2 emissions21. Fundamental vulnerability in the governance structure enables fraudulent activities within the environmental sector, thereby directly or indirectly affecting environmental stress. Moreover, public management has a vital function in governance, including the regulation of CO2 emissions, the fight against corruption, and the grey economy6. Considering the previously described viewpoint, the present study demonstrates the influence of public management and CE on CO2 emissions in E-7 economies from 2000 to 2023. Considering this background, this study significantly enhances the existing body of knowledge. This work contributes three novel discoveries to existing academic scholarly research. The present study is one of the few investigations to assess the effect of CE on carbon emissions (CO2) inside the E-7 (Turkey, Russia, Mexico, India, Brazil, China, Indonesia, and Mexico) economies. The results of our study will be valuable to policymakers for the development and implementation of green technologies that promote environmental sustainability. The second contribution focuses on the collection of the indicators used in this study. This study contributes to the literature on the correlation between digital technology and public management by presenting empirical information on the influence of CE on carbon emissions (CO2) in E-7 economies. Nevertheless, the current corpus of literature has investigated this previously expounded correlation. Nevertheless, these conflicting findings create a deceptive portrayal of the correlation in E-7 economies. Therefore, this study provides a clear demonstration of the previously explained correlation in different economic systems. We assess the long-term heterogeneity of E-7 economies by employing empirical and scientific methodologies. The econometric methods used in this study include the second-generation panel unit-root test, cross-sectional dependency test, and second-generation cointegration tests. In addition, lung run estimation was performed using the Interactive Fixed Effect (IFE) and D-CCE approaches. Additionally, this study employed panel quantile regression (PQR) approaches to derive robust findings. Policy proposals based on scientific data will contribute to the attainment of high levels of sustainable development outcomes nationwide by promoting clean energy quality. The main goal of this study is to bridge existing research gaps in three ways. First, prior studies have rarely analyzed the combined effects of CE, DT, and PM on CO₂ emissions, especially for E-7 countries that face rapid urbanization and industrialization. Second, few works have empirically tested the moderating role of governance quality in enhancing technological and circular transitions. Third, the study applies second-generation econometric techniques and a panel quantile regression (PQR) approach to capture distributional heterogeneity offering more robust and policy-relevant insights. The remainder of this paper is structured as follows. Building upon the “Introduction” section, the “Literature review” section provides a meticulous examination of previous studies on public administration, CE, and carbon emissions. The "Data and Methodology" section elucidates the procedure and visually displays the data. The section under "Empirical results and discussions" presents the empirical outcomes and discussions. The section "Conclusions and policy implications" concludes by providing a concise summary of the primary findings and wider policy consequences. Literature reviews CE and carbon emission The impact of energy usage on carbon emissions and production-based accounting techniques have all been studied in the past22. Prior studies have clarified the increase in carbon-reducing measures, such as the implementation of carbon levies and emission-trading systems. A second contentious academic study concerning indirect carbon emissions of nations further emphasizes this widespread issue. First, it is essential to understand the differences between direct and indirect emissions to successfully address this problem. Direct greenhouse gas (GHG) emissions (Scope 1) are produced directly from the production of goods, such as factory aerosols, irrespective of whether the reporting business owns or manages the source. By contrast, indirect greenhouse gas emissions (scopes 2 and 3) are produced by the activities of a reporting organization but occur at sources that are owned or managed by a different entity. Indirect emissions encompass the emissions generated by the commercial consumption of energy, steam, and heating/cooling as well as the emissions produced by suppliers within the supply chain. Recent research by Singh23 indicates that the sole application of energy-saving techniques is insufficient for achieving net-zero outcomes. Although prior research24 has investigated additional factors influencing scope 3 carbon emissions, including business travel, procurement, waste, and water, these features have not been extensively analyzed. Su25 examined the influence of dynamic material flow on productivity, recycling, and iron ore consumption within the Chinese steel industry from 2001 to 2010. This examination considers the generation of garbage, retrieval and reprocessing of plastic, and subsequent reutilization of plastic after recycling. An extensive meta-analysis undertaken by Cheng26 examines more than 300 CE scenarios, with a specific emphasis on the impact on gross domestic product, employment generation, and CO2 emissions. These studies reveal a direct relationship between CO2 emissions and CE indicators. Unfortunately, a considerable proportion of this research focuses on much shorter periods, therefore overlooking temporal complexities in carbon dioxide emissions. Moreover, in the present empirical research, challenges persist in adequately addressing endogeneity. There are two divergent viewpoints regarding the influence of CE on CO2 emissions. Existing research indicates that the adoption of a CE may significantly reduce CO2 emissions, thereby improving environmental quality27. Hence, countries with strong circular policies dedicate significant resources to mitigating CO2 emissions. Bag et al.28 studied the link between the CE and carbon dioxide emissions across the European Union (EU). Data analysis indicated that a 1% increase in the recycling rate of municipal waste (RMW) led to a 0.5% decrease in average CO2 emissions. Their study model indicates that recycling the 15 most high-demand polymers in Europe reduces plastic CO2 emissions by 73%, or 200 million metric tons of CO2 equivalent. In a recent study undertaken by Schöggl et al.29, the potential benefits of recycling steel, nonferrous metals, plastic, and paper wastes were examined in relation to energy generation (electricity) and environmental footprint (GHG emissions and air pollutants). Employing the model equations method, the research results demonstrated that solid waste recycling successfully prevented the release of 4665.8 billion kg of carbon dioxide and 23.504 billion kg of methane. Patra30 investigated the cointegration between energy efficiency, financial growth, carbon emissions, and recycling of MSW using bootstrapping autoregressive distributed lag models. The analysis was conducted using quarterly data from 1990 to 2017. The results suggest that a 1% rise in MSW recycling might have a positive effect on economic development and result in a decline in carbon emissions by 0.417% (0.257%) and 0.309% (0.187%) in the long term (short-term). In contrast, the second body of study contends that a CE has an insignificant impact on CO2 emissions. By analyzing the economic interaction between China and Nigeria, Alkhuzaim31 investigated the potential of a CE to reduce energy use in countries that heavily depend on mining and extractive sectors. Moreover, Ding32 argue that the implementation of chemical recycling technology will not essentially reduce the use of fossil fuels or the production of greenhouse gases. This is because their current minimum environmental impacts surpass those of the established benchmark waste treatment technologies. Moreover, Agyabeng-Mensah33 investigated the recycling practices of steel, plastics, and paper materials in China. The research findings indicated that the recycling of these three commodities had an adverse influence on the environment, particularly in relation to emissions. The causative analysis indicated that there was no statistically significant correlation between the rate of recycling, renewable energy consumption, and carbon dioxide emissions. CE is an ecologically sound approach to waste management that focuses on three key areas: waste reduction, reuse improvement, and ecological biodiversity protection. In recent decades, the concept of CE has clarified the efficient use of trash by economies to reduce emission levels. CE is now acknowledged as a new phenomenon that forms the foundation of green development. Hence, CE might be the most effective strategy to reduce overdependence on non-sustainable energy sources, thereby resulting in major benefits for the economy, environment, and society34. However, few empirical studies have attempted to explore the correlation between emissions and CE in different worldwide regions. Furthermore, a study of the Angolan real economy using time series analysis clarified the association between CE and emission levels by Wei35. However, they chose to use the landfill method and found a direct association between CE and the decrease in biodiversity. Zhu36 unveiled the substantial influence of CE on the long-term viability of the United States economy between 1990 and 2018. The goals of this study were evaluated using the ARDL estimator, which demonstrated a significant decrease in environmental strain by repurposing recyclable waste. This association has been established because of the widespread use of the recycling technique, instead of the landfiling method, in CE, which has successfully integrated sustainability at the household level. Although several recent case studies have examined the influence of CE in sectors such as economics and energy, they have not explicitly addressed the concept of sustainability. Notable research undertaken by Ahmad37 in Nigeria, Hang38 in Bangladesh, Raihan39 in Malaysia, Bashir40 in China, and Wang41 is included in these studies. The operationalization of the circular economy varies widely across studies ranging from recycling rates and material circularity indicators to waste-to-energy recovery42. In this paper, CE is proxied by “energy recovery from waste,” which captures one critical but partial dimension of circularity. While this measure reflects the emission-reduction potential of energy substitution, it does not cover the full spectrum of circular practices such as reuse, redesign, or remanufacturing. This limitation is recognized in the discussion and addressed in the study’s limitations section. Public management and carbon emission (CO2) In pursuit of this objective, the World Bank has devised measures to enhance the beneficial impact of governance on both the economy and environment43. Public management refers to the coordination and administration of resources and services by governments, institutions, and economic entities within a legal framework that promotes sustainable development by ensuring compliance with the legal rules. Nevertheless, research on the function of public management in the environment is limited. Nevertheless, the literature has not specifically addressed the role of governance in environmental pollution, and there has been limited research on the connection between government management and CO2 emissions44. Nevertheless, existing studies concur that a correlation exists between the two variables. Yadav45 examined a group of high- and upper-middle-income nations between 1985 and 2011. The researchers did not observe a statistically significant influence of corruption control on the association between energy, environment, and economic growth. Therefore, they contended that the effective management of corruption had minimal significance in reducing CO2 emissions. Yu46 examined how corruption moderates the association between CO2 emissions and financial growth in a sample of BRICS nations from 1996 to 2016. Their findings indicate that corruption weakens the relationship between economic development and CO2 emissions. The researchers concluded that the effective management of corruption leads to a reduction in air pollution. Wei and Liu47 used experts’ survey evaluations as proxies for the institutions of Switzerland and Norway among European Union nations. These findings indicate that pollution is reduced by national environmental rules, environmental subsidies, and corporate social responsibility. Luo48 employed a panel consisting of 17 nations from the Middle East and North Africa, observed from 1980 to 2005. Their findings validated the moderating influence of the quality of democratic institutions on SO2 pollution but did not show statistical significance for CO2 emissions. Erdiwansyah49 assessed the moderating influence of political and institutional criteria on the elasticity of GDP for CO2 emissions in 140 countries from 1972 to 2014. The findings revealed a moderating influence, albeit marginal, of democratic organizations and zero corruption, and a more significant effect of civil society involvement. The observed pattern of moderation indicates that in nations characterized by low per capita GDP, institutional elements play a role in generating greater CO2 elasticity in relation to per capita GDP. The influence of this phenomenon varies as per capita GDP increases, and CO2 emissions are not linked to financial growth in affluent countries. The carbon tax policy is a carbon emission (CO2) reduction program that the government is currently considering. The nonpartisan game developed by Rajendran50 involved retailers and manufacturers. The objective of this study was to determine the most effective carbon tax formulation that would lead to the greatest possible decrease in carbon emissions. From the standpoint of central decision-makers who aim to maximize social welfare, Agnihotri51 examined the development of a carbon tax for the most effective reduction of carbon emissions in a two-party game including retailers and producers. The implementation of a more stringent carbon price by the government has the potential to incentivize firms to actively minimize their carbon emissions. However, it impedes the carbon emission (CO2) reduction behavior of competitive firms52. Practically, where investment is justifiable, the cost subsidy is more effective in reducing carbon emissions than the carbon tax policy53. Prior research has demonstrated that corporations can successfully encourage investment in renewable energy through government subsidy programs. Government subsidies are an essential and indispensable means of survival for small-scale renewable energy enterprises54. The government plays a central role as a management mechanism in the carbon emission reduction process of building companies. The policies enacted by the government directly affect the degree of carbon emission (CO2) reduction throughout the entire process. Effective policy implementation serves as a catalyst, expediting the process of reducing CO2 emissions from buildings and enhancing collaboration among construction companies throughout the construction process. It also encourages the sustainable behavior of construction enterprises in reducing CO2 emissions throughout the supply chain. In contrast, the enforcement of inadequate policies will severely limit the carbon emission reduction initiatives of construction companies and impede coordination and collaboration among them. Indeed, the government has not been the primary focus of investigation in prior 1studies on reducing carbon emissions from buildings55. However, this work adopts a novel methodology by focusing on the government as the sole subject of investigation. The objective was to identify a balanced approach to attain the best possible level of CO2 reduction throughout the industrial supply chain. Prior research has exclusively focused on the government’s subsidy program for construction costs, resulting in a reduction in the government’s importance56. This study examines the impact of government intervention, specifically the provision of cost subsidies to firms or the imposition of penalties on carbon emissions by building enterprises. These kinds of improvements can more accurately demonstrate the government’s focus on reducing carbon emissions from buildings and underscore the government’s role in overseeing and managing them. Furthermore, through a comparison and analysis of the equilibrium strategy for reducing carbon emissions in buildings under three different policies, this study identifies the most effective approach to decrease carbon emissions. This finding offers theoretical legitimacy for determining suitable subsidy and tax rates for CO2 reduction costs. From a theoretical standpoint, the moderating influence of public management on environmental outcomes can be explained through institutional and innovation-systems perspectives. Effective governance enhances technology adoption and circular transitions by enforcing environmental standards, allocating green subsidies, co-investing in infrastructure, and signaling credible long-term climate commitments. Well-functioning institutions reduce uncertainty for private actors and amplify the benefits of technological innovation and circular initiatives. Consequently, the interaction terms (PM × DT, PM × CE, PM × ENT) represent these enabling or constraining institutional effects on CO₂ mitigation. Data and methodology Data The purpose of this study is to examine the impact of CE and public management on carbon emissions (CO2) in E-7 economies using a dataset stretching from 2000 to 2023. The variables CO2, GDPC, URB, DT, CE, PM, and ENT indicate carbon emissions, GDP per capita, urbanization, digital technology, CE, public management, and energy transition, respectively. The dependent variable in the study is carbon emissions (CO2), which is quantified in kilotons (kt). The World Development Indicator (WDI) provides data on GDP per capita, urbanization (% of total urban population), digital technology assessed by the number of internet users, and the CE evaluated by energy recovery from wastage (thousand tons) obtained from national statistical agencies and and cross-verified with UNEP’s Global Waste Management Outlook (2022) and OECD Environmental Statistics. Public administration refers to the capacity of the government to create and enforce effective rules and regulations, as well as guide the growth of the private sector. The World Bank served as the data source for independent variables. Finally, the energy transition is quantified as a percentage of the total final energy consumption, including data obtained from the WDI (Table 1). Figure 1 shows the bees-plots of the study variables.Table 1Variable descriptions and data source.VariablesDefinitionSourceCO2Carbon emission (kt)WDIGDPCGDP per capita (constant 2015 US$)WDIURBUrbanization (% of total urban population)WDIDTDigital technology (number of internet users)WDICEMaterial circularity, recycling rate, energy recovery (Thousand Tonnes)National statistical agenciesPMPublic management (the capacity of government to devise and execute effective policies and regulations while fostering private sector growthWDIENTEnergy transition (% of total final energy consumption)WDIFor non-EU E-7 countries, data on energy recovery were compiled from national statistical agencies and cross-verified with UNEP’s Global Waste Management Outlook (2022) and OECD Environmental Statistics. These values were harmonized to ensure comparability across countries.Fig. 1Beeswarm plots of study variables. Each panel shows the distribution of the variable across E-7 countries (2000–2022). Axes are labeled in natural-log form: CO₂ (kt), GDP per capita (USD 2015), urbanization (%), etc. Table 1Variable descriptions and data source.VariablesDefinitionSourceCO2Carbon emission (kt)WDIGDPCGDP per capita (constant 2015 US$)WDIURBUrbanization (% of total urban population)WDIDTDigital technology (number of internet users)WDICEMaterial circularity, recycling rate, energy recovery (Thousand Tonnes)National statistical agenciesPMPublic management (the capacity of government to devise and execute effective policies and regulations while fostering private sector growthWDIENTEnergy transition (% of total final energy consumption)WDIFor non-EU E-7 countries, data on energy recovery were compiled from national statistical agencies and cross-verified with UNEP’s Global Waste Management Outlook (2022) and OECD Environmental Statistics. These values were harmonized to ensure comparability across countries. Variable descriptions and data source. For non-EU E-7 countries, data on energy recovery were compiled from national statistical agencies and cross-verified with UNEP’s Global Waste Management Outlook (2022) and OECD Environmental Statistics. These values were harmonized to ensure comparability across countries. Beeswarm plots of study variables. Each panel shows the distribution of the variable across E-7 countries (2000–2022). Axes are labeled in natural-log form: CO₂ (kt), GDP per capita (USD 2015), urbanization (%), etc. Model specification This study specifically examined the following models:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{CO}}_{{2}} \, = \,{\mathrm{f}}\left( {\beta_{0} ,{\text{ GDPC}}^{{\beta {1}}} ,{\text{ URB}}^{{\beta {2}}} ,{\text{ DT}}^{{\beta {3}}} ,{\text{ CE}}^{{\beta {4}}} ,{\text{ PM}}^{{\beta {5}}} ,{\text{ ENT}}^{{\beta {6}}} } \right)$$\end{document} Carbon emissions, GDP per capita, urbanization, digital technology, CE, public management, and energy transition are abbreviated as CO2, GDPC, URB, DT, CE, PM, and ENT, respectively. Consequently, the observational model that requires estimation is as follows:2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{CO}}_{{{2},{\mathrm{t}}}} \, = \,\beta_{0} \, + \,\beta_{{1}} {\mathrm{GDPC}}_{{{\mathrm{i}},{\mathrm{t}}}} \, + \,\beta_{{2}} {\mathrm{URB}}_{{{\mathrm{i}},{\mathrm{t}}}} \, + \,\beta_{{3}} {\mathrm{DT}}_{{{\mathrm{i}},{\mathrm{t}}}} \, + \,\beta_{{4}} {\mathrm{CE}}_{{{\mathrm{i}},{\mathrm{t}}}} \, + \,\beta_{{5}} {\mathrm{PM}}_{{{\mathrm{i}},{\mathrm{t}}}} \, + \,\beta_{{6}} {\mathrm{ENT}}_{{{\mathrm{i}},{\mathrm{t}}}} \, + \,\varepsilon_{{{\mathrm{i}},{\mathrm{t}}}}$$\end{document} Let i represent economics and t represent a year. The error term, denoted by εit, is subject to a normal distribution with a mean of zero and minimal volatility. The variables lnCO2, lnGDPC, lnURB, lnDT, lnCE, lnPM, and lnENT in Eq. (3) represent the natural logarithm of carbon emissions, GDP per capita, urbanization, digital technology, CE, public management, and energy transition, respectively. Regression analysis was the basis for this study.3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${lnCO}_{2 i,t}={\beta }_{0}+{\beta }_{1}ln{GDPC}_{i,t}+{\beta }_{2}ln{URB}_{i,t}+{\beta }_{3}{lnDT}_{i,t}+{\beta }_{4}ln{CE}_{i,t}+{\beta }_{5}ln{PM}_{i,t}+{\beta }_{6}ln{ENT}_{i,t}+{\epsilon }_{i,t}$$\end{document} Another aim of this study is to observe the mild influence of public management on digital technology, CE, energy transition, and their effect on carbon emissions (CO2). Furthermore, Eqs. 4, 5, and 6 for the moderate effect are included below:4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${lnCO}_{2 i,t}={\beta }_{0}+{\beta }_{1}ln{GDPC}_{i,t}+{\beta }_{2}ln{URB}_{i,t}+{\beta }_{3}{lnDT}_{i,t}+{\beta }_{4}ln{CE}_{i,t}+{\beta }_{5}ln{PM}_{i,t}+{\beta }_{6}ln{ENT}_{i,t}+{\beta }_{7}(ln{PM X DT)}_{i,t}+{\epsilon }_{i,t}$$\end{document}5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${lnCO}_{2 i,t}={\beta }_{0}+{\beta }_{1}ln{GDPC}_{i,t}+{\beta }_{2}ln{URB}_{i,t}+{\beta }_{3}{lnDT}_{i,t}+{\beta }_{4}ln{CE}_{i,t}+{\beta }_{5}ln{PM}_{i,t}+{\beta }_{6}ln{ENT}_{i,t}+{\beta }_{7}(ln{PM X CE)}_{i,t}+{\epsilon }_{i,t}$$\end{document}6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${lnCO}_{2 i,t}={\beta }_{0}+{\beta }_{1}ln{GDPC}_{i,t}+{\beta }_{2}ln{URB}_{i,t}+{\beta }_{3}{lnDT}_{i,t}+{\beta }_{4}ln{CE}_{i,t}+{\beta }_{5}ln{PM}_{i,t}+{\beta }_{6}ln{ENT}_{i,t}+{\beta }_{7}(ln{PM X ENT)}_{i,t}+{\epsilon }_{i,t}$$\end{document} This study displays the long-run URB, DT, CO2, CE, PM, GDPC, and ENT elasticities. The coefficient of URB > 0 implies that the level of urbanization increases carbon emission (CO2) (β2 = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{\beta {\mathrm{CO}}2}}{{\beta {\mathrm{URB}}}}$$\end{document} > 0). The process of urbanization in E-7 countries, which heavily depend on fossil fuels, might be linked to unsustainable methods of generating economic output, which hinders the goal of establishing green growth. Therefore, given the significant reliance of E-7 countries on fossil fuels, a rise in the rate of urbanization can potentially accelerate the growth of their CO2 emissions at a comparatively quicker rate than the growth of their economic production levels. One result of digital technology is the ability to make more effective investments in environmentally sustainable industrial processes and technologies, leading to increased allocation of funds towards such opportunities. This study shows a negative effect of digital technology on carbon emission (β3 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{\beta {\mathrm{CO}}2}}{{\beta {\mathrm{DT}}}}$$\end{document} < 0). The CE coefficient is negatively associated with carbon emission (CO2) (β4 = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{\beta {\mathrm{CO}}2}}{{\beta {\mathrm{CE}}}}$$\end{document} < 0). The public management coefficient is negatively associated with carbon emission (CO2) (β5 = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{\beta {\mathrm{CO}}2}}{{\beta {\mathrm{PM}}}}$$\end{document} < 0). Last of all, we predict a positive association between the energy transition and carbon emission, i.e., (β6 = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{\beta {\mathrm{ENR}}}}{{\beta {\mathrm{URB}}}}$$\end{document} > 0). Previous research has recognized that a growth in this proportion can be seen as a successful execution of policies aimed at transitioning to renewable energy, thereby enhancing the level of reliance on fossil fuels57. Further implementation of energy transition, coupled with rapid reduction of fossil fuel use, is expected to facilitate the uptake of carbon energy resources, significantly augmenting the reliance of E-7 countries on fossil fuels. Consequently, this can enhance their rates of economic output development while simultaneously propelling their CO2 emissions growth rates. Therefore, the shift towards renewable energy generation might be associated with the emission of CO2 in E-7 countries. Estimation strategy The estimation pattering has been shown in Fig. 2.Fig. 2Estimation pattering. Estimation pattering. Cross-sectional dependence test The cross-sectional dependence (CD) study conducted by Pesaran (2015) is of utmost importance for panel data, as it has the potential to generate imprecise and conflicting findings. Various routes, including economic, social, political, and bilateral trading and sharing boards, provide tangible instances of international connections. The emergence of CD may be attributed to several manifestations of international interaction. To achieve this objective, we employ the CD test, which was advanced by Pesaran58, and the Lagrange Multiplier (LM) test, which was established by Baltagi59. The equation utilized in CD testing to detect the presence of CD in the data is as follows:7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathrm{CD}=\sqrt{\frac{2T}{N(N-1)({\sum }_{J=1+1}^{N}Pji)}}$$\end{document} A cross-section is represented by the numerical value N and the temporal dimension T. To gain insight into the diverse correlations among random fluctuations, we might examine the following: This equation is utilized in the subsequent LM test example to examine the panel data for the CD-dependent variables.8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{Yit}} = \alpha {\mathrm{i}} + \beta {\mathrm{ixit}} + \varepsilon {\mathrm{it}}$$\end{document} We define T as the temporal interval and I as the intermediate values. Both estimation methods rely on the null hypothesis that CD is absent in panel data, while alternative hypotheses consider CD. Cross-sectional unit root test Therefore, when confronted with CD, the unit root tests of the first generation are extremely inadequate. The goal of this project is to create a single framework that is flexible enough to incorporate the most beneficial parts of the Pesaran-Shin and Dickey-Fuller models. The CIPS and CADF models were developed by Baltagi60. The heterogeneity present in both cross-sectional and panel data is taken into consideration by both evaluations. We used unit tests of the second generation to determine whether the variables were consistent with one another.9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\Delta X}_{{{\mathrm{i}},{\mathrm{t}}}} = \alpha_{{{\mathrm{i}},{\mathrm{t}}}} + \beta {\mathrm{iX}}_{{{\mathrm{i}},{\mathrm{t}} - 1}} + \rho {\mathrm{iT}} + \sum\nolimits_{{{\mathrm{j}} = 1}}^{{\mathrm{n}}} {{\uptheta }_{{{\mathrm{i}},{\mathrm{t}}}} \cdot {\Delta X}_{{{\mathrm{i}},{\mathrm{t}} - {\mathrm{j}}}} + {\upvarepsilon }_{{{\mathrm{i}},{\mathrm{t}}}} }$$\end{document} If t denotes the time interval, i stands for cross-sections, r stands for the model’s residuals, and xit is the variable of interest. The alternative hypothesis assumes stationarity, whereas the null hypothesis accounts for non-stationarity. Westerlund panel cointegration Using the error correction-based cointegration approach, which considers the CSD as presented by Westerlund61, we investigate the long-term relationship that exists between the variables that have been selected. The capacity of the Westerlund test to handle disruptions caused by the endogeneity of the reactants was a primary factor in its adoption. Subsequent to the establishment of four null hypotheses denoting a lack of cointegration, this test employed two statistical analyses to evaluate the overall cointegration of the panel. The tests for the CSD and non-strictly exogenous regressors are predicated on the subsequent error correction equations, articulated as follows:10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${y}_{\mathrm{it}}={\alpha }_{\mathrm{0i}}+{\alpha }_{\mathrm{1it}}+{Z}_{\mathrm{it}}$$\end{document}11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${X}_{\mathrm{it}}={x}_{\text{i, t-1}}+{\mu }_{\mathrm{it}}$$\end{document}12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\delta }_{\mathrm{i}}(L){Z}_{\mathrm{it}}={\delta }_{\mathrm{i}}({Z}_{\mathrm{it-1}}+{\gamma }_{1}\mathrm{,}{\mathrm{x}}_{\mathrm{i,t-1}})+{\beta }_{\mathrm{i}}(L){\mu }_{\mathrm{it}}+{\varepsilon }_{\mathrm{it}}$$\end{document}13\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\delta }_{\mathrm{i}}(L)\Delta {y}_{\mathrm{i,t}}={\theta }_{\mathrm{0i}}+{\theta }_{\mathrm{1it}}+{\delta }_{\mathrm{i}}({y}_{\text{i, t-1}}-{\gamma }_{1},{x}_{\text{i, t-1}})+{\beta }_{\mathrm{i}}(L){\mu }_{\mathrm{it}}+{\varepsilon }_{\mathrm{it}}$$\end{document} Long Run estimation of Interactive Fixed Effect (IFE) and D-CCE tests Prior research on the quantification of regression models with factor error structures can be categorized into two distinct classes. In the first section, we will discuss the D-CCE approach, an enhancement of the work of Li62 that is commonly employed to measure the common factor of the error structure. In 2006, Pesaran introduced the CCE technique, which utilizes linear static panel models that incorporate both homogeneous and heterogeneous coefficients. When using econometric methods, substituting the cross-sectional average of the descriptive and explanatory variables for common factors may be difficult because the estimator cannot accurately measure the factor component. The D-CCE protocol, devised by Naor63, was employed in this study for many purposes. Additionally, it accommodates heterogeneity within the ED, as demonstrated by Le Quéré64. Bai’s IFE proposal represents the second approach. The present methodology employs an iterative procedure to concurrently compute both the factor element and regression coefficient. Equation (14) was used to ascertain the suitability of this approach for the present study.14\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{Y}}_{{{\mathrm{i}},{\mathrm{t}}}} = {\mathrm{X}}_{{,{\mathrm{t}}\alpha }} + \varphi_{{\mathrm{a}}} + {\mathrm{s}}_{{\mathrm{t}}} + {\mathrm{e}}_{{{\mathrm{i}},{\mathrm{t}}}}$$\end{document}15\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{e}}_{{{\mathrm{i}},{\mathrm{t}}}} = \Gamma_{i}^{\prime } {\mathrm{f}}_{{\mathrm{t}}} + \mu_{{{\mathrm{i}},{\mathrm{t}}}}$$\end{document} In Eq. (14), Γ′i,t represents the error term, Ft is a vector of unobserved time-specific common stocks multiplied by factor r, and µi, t is a vector of factor loadings that reflects how each unit reacts to standard shocks. The regression coefficients and factoring procedure were proposed by Wang65. Which individual employed the technique of reducing the sum of residual squares (SSR), as described in Eq. (16)?16\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{SSR}} = \sum\nolimits_{i = 1}^{N} {\sum\nolimits_{t = 1}^{T} {\left( {Y_{i,t} - X_{i,t} \alpha - \varphi_{i} - S_{t} - \Gamma_{i}^{\prime } f_{t} } \right)} }$$\end{document}17\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{SSR}}_{1} = \sum\nolimits_{i = 1}^{N} {\sum\nolimits_{t = 1}^{T} {\left( {{\mathrm{Y}}_{{{\mathrm{i}},{\mathrm{t}}}} \left( 1 \right) - {\mathrm{X}}_{{{\mathrm{i}},{\mathrm{t}}}} \alpha - \varphi_{{\mathrm{i}}} - {\mathrm{S}}_{{\mathrm{t}}} } \right)^{2} } }$$\end{document}18\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{SSR}}_{2} = \sum\nolimits_{i = 1}^{N} {\sum\nolimits_{t = 1}^{T} {\left( {Y_{i,t} \left( 1 \right) - \Gamma_{i}^{\prime } f_{t} } \right)^{2} } }$$\end{document}19\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Y\left( 2 \right)_{I,t} = Y_{i,t} - X_{i,t} \beta - \varphi_{i} - S_{t}$$\end{document} Robust check by Panel quantile regression model The panel quantile regression (PQR) method is employed because it allows coefficient estimation across different points of the conditional CO₂ distribution rather than only the mean. Compared with traditional mean-based estimators, PQR effectively captures country-level heterogeneity and is robust to outliers, non-normal residuals, and heteroskedasticity. Given the diverse economic structures of E-7 countries, this approach provides a richer understanding of how the effects of DT, CE, and PM vary across low- and high-emission regimes. This study examines the factors that affect carbon emissions across the conditional distribution of panel nations. Similar to the conditional technique, the PQR method does not conform to the normality assumption66. Unlike traditional techniques, the PQR algorithm effectively addresses outliers and produces robust results. With different quantiles, the impact of the predicted factors on the exogenous variables is clearly shown by PQR. PQR analyses unobserved heterogeneity and measures many parameters at different quantiles within each cross-section. However, conventional regression methods prioritize average effects, leading to the potential underestimation or overestimation of the crucial coefficient and the inability to identify significant variations67–70. Quantile regression has attracted considerable interest from scholars and is widely used in environmental economics. PQR approach for different quantiles of yit given xit is as follows.20\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${y}_{it}={\theta }_{i}+{x}_{it}+{e}_{it}$$\end{document} In this context, θi represents the individual fixed effects, yit is the dependent variable, xit refers to the independent variables, and eit represents the error term. The estimator for the coefficient of the τth conditional quantile function is provided as21\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Q_{{y_{it} }} \left( {\tau |\theta_{i} ,x_{it} } \right) = \theta_{i} + x_{it}^{\prime } \psi \left( \tau \right) + \left( \tau \right)e_{t}$$\end{document} Results and discussion Descriptive statistics test The variables’ descriptive statistics, including the mean, median, minimum, maximum, and standard deviation, are displayed in Table 2. Upon linking the means of the various categories, it is evident that public management (PM) exhibits the lowest mean value, whereas GDPC demonstrates the highest. This suggests that the Gross Domestic Product per capita (GDPC) value of the selected economies remains elevated. Furthermore, the lack of an outlier is indicated by the absence of a statistically significant difference between the mean and median. The main measurements for CO2 were 12.374, GDPC was 21.048, URB was 13.276, DT was 19.945, CE was 7.016, PM was 5.286, and ENT was 9.137.Table 2Descriptive statistics.VariablesMeanMedianMaximumMinimumlnCO212.37412.09522.1975.726lnGDPC21.04820.70634.8327.494lnURB13.27613.14515.6243.615lnDT19.94519.23625.3176.920lnCE7.0166.8599.7521.475lnPM5.2865.1906.3062.386lnENT9.1378.93612.4264.590 Table 2Descriptive statistics.VariablesMeanMedianMaximumMinimumlnCO212.37412.09522.1975.726lnGDPC21.04820.70634.8327.494lnURB13.27613.14515.6243.615lnDT19.94519.23625.3176.920lnCE7.0166.8599.7521.475lnPM5.2865.1906.3062.386lnENT9.1378.93612.4264.590 Descriptive statistics. Pairwise correlation matrix test Prior to using multivariate regression models, it is crucial to evaluate the existence of multicollinearity within a particular subset of panel data. We confirmed the absence of multicollinearity in our model using a pairwise correlation matrix. Pairwise correlation can lead to significant results, as demonstrated in (Table 3). The analysis shows a strong and meaningful positive association between GDP per capita (GDPC) and urbanization, and the emission of CO2 at a level of significance of 1%. Moreover, when considering a significance threshold of 1%, there is an unfavorable association observed between digital technology, CE, and public management with the explanatory factors. Furthermore, there was a statistically significant positive link (with a significance level of 5%) between the process of ENT sources and the amount of carbon dioxide emissions. This suggests that the rise in CO2 levels can be attributed to the growing levels of energy transition.Table 3Correlation matrix results.VariableslnCO2lnGDPClnURBlnDTlnCElnPMlnENTlnCO21.000lnGDPC0.367**1.000lnURB0.510**0.065***1.000lnDT − 0.322*** − 0.146**0.624***1.000lnCE − 0.657***0.215***0.516**0.214***1.000lnPM − 0.289***0.496**0.183***0.625**0.352**1.000lnENT0.421***0.590***0.451**0.291*** − 0.476***0.195***1.000*** and ** indicate p < 1% and p < 5%, respectively. Table 3Correlation matrix results.VariableslnCO2lnGDPClnURBlnDTlnCElnPMlnENTlnCO21.000lnGDPC0.367**1.000lnURB0.510**0.065***1.000lnDT − 0.322*** − 0.146**0.624***1.000lnCE − 0.657***0.215***0.516**0.214***1.000lnPM − 0.289***0.496**0.183***0.625**0.352**1.000lnENT0.421***0.590***0.451**0.291*** − 0.476***0.195***1.000*** and ** indicate p < 1% and p < 5%, respectively. Correlation matrix results. *** and ** indicate p < 1% and p < 5%, respectively. Cross-sectional dependency test It is assumed that the cross sections are entirely independent of one another. Nonetheless, panel data may exhibit cross-sectional interdependence due to international trade and travel. Neglecting CSD in the estimate may lead to the production of inefficient and biased estimators. Consequently, the present investigation included the use of CSD analyses, and the results are presented in Table 4. The presence of CSD was identified. The results of the first-generation unit root test conducted by Ma et al.71 may be biased and misleading when there is CSD. The discovery we made requires us to deal with the problem of CSD by using a different method for testing unit roots, namely the second-generation method developed by Stanislavská72, referred to as the CIPS and CADF algorithms.Table 4CSD results.VariablesPesaran’s testFrees’testFriedman’s testlnCO215.876 (0.021)6.290 (0.000)132.736 (0.000)lnGDPC9.629 (0.003)3.816 (0.000)75.594 (0.025)lnURB12.039 (0.005)4.208 (0.000)145.625 (0.009)lnDT6.971 (0.000)2.742 (0.004)72.032 (0.001)lnCE5.725 (0.000)3.865 (0.002)52.104 (0.000)lnPM20.413 (0.001)7.926 (0.015)195.126 (0.005)lnENT7.625 (0.009)4.217 (0.024)67.430 (0.000)*** and ** indicate p < 1% and p < 5%, respectively. Table 4CSD results.VariablesPesaran’s testFrees’testFriedman’s testlnCO215.876 (0.021)6.290 (0.000)132.736 (0.000)lnGDPC9.629 (0.003)3.816 (0.000)75.594 (0.025)lnURB12.039 (0.005)4.208 (0.000)145.625 (0.009)lnDT6.971 (0.000)2.742 (0.004)72.032 (0.001)lnCE5.725 (0.000)3.865 (0.002)52.104 (0.000)lnPM20.413 (0.001)7.926 (0.015)195.126 (0.005)lnENT7.625 (0.009)4.217 (0.024)67.430 (0.000)*** and ** indicate p < 1% and p < 5%, respectively. CSD results. *** and ** indicate p < 1% and p < 5%, respectively. Findings of the unit root test This paper offers an extensive examination of CIPS and CADF. The results of both unit root tests provide a robust rejection of the nnon-stationarity of nonstationary, with p-values below 0.01 for all selected variables. Thus, although explanatory variables exhibit stationarity at this level, environmental variables display differenced stationarity. Table 5 clearly demonstrates that both tests yielded a consistent diagnosis. The CIPS test results for the sampled economies reveal that the CO2, GDPC, URB, CE, and PM variables show a unit root problem at the aggregate level. Nonetheless, the principles of Digital Transformation (DT) and energy transition remain constant at the macro level. If none of the variables display a unit root under the assumption of first-difference outcomes, it indicates that the variables of interest are integrated of order one. Furthermore, the conclusions reached by CADF were not markedly different from those of CIPS.Table 5Findings of CADF and CIPS unit root tests.VariablesCADF unit root testCIPS unit root testLevel1st DifferenceLevel1st DifferencelnCO2− 1.287− 3.920***− 1.287− 3.920***lnGDPC− 3.094− 5.617***− 3.094− 5.967***lnURB− 1.725− 3.824**− 1.725− 3.251***lnDT− 1.064***− 3.319***− 1.064**− 3.209***lnCE− 1.526− 3.490***− 1.526− 3.564***lnPM− 2.387− 4.827***− 2.387− 4.917***lnENT− 1.762***− 3.245***− 1.762***− 3.452****** and ** indicate p < 1% and p < 5%, respectively. Table 5Findings of CADF and CIPS unit root tests.VariablesCADF unit root testCIPS unit root testLevel1st DifferenceLevel1st DifferencelnCO2− 1.287− 3.920***− 1.287− 3.920***lnGDPC− 3.094− 5.617***− 3.094− 5.967***lnURB− 1.725− 3.824**− 1.725− 3.251***lnDT− 1.064***− 3.319***− 1.064**− 3.209***lnCE− 1.526− 3.490***− 1.526− 3.564***lnPM− 2.387− 4.827***− 2.387− 4.917***lnENT− 1.762***− 3.245***− 1.762***− 3.452****** and ** indicate p < 1% and p < 5%, respectively. Findings of CADF and CIPS unit root tests. *** and ** indicate p < 1% and p < 5%, respectively. Westerlund panel cointegration test This study assessed the cointegration relationship between both explosive dynamics (ED) models utilizing the ECM cointegration methodology established by Westerlund61. Table 6 presents the results of the cointegration analysis, encompassing both constant and trend values. The bootstrapped significant p-values reveal that the panel and group statistics refute the null hypothesis of no cointegration. The findings indicate that cointegration exists among the variables lnCO2, lnGDPC, lnURB, lnDT, lnCE, lnPM, and lnENT. Consequently, the null hypothesis of no cointegration is rejected, confirming the presence of cointegration among the variables of interest. Following the verification of cointegration, the subsequent step involves analyzing the association between the variables.Table 6Westerlund cointegration.StatisticsValueZ-valueP-valueRobust P-valueGt− 2.5760.9170.8260.152Ga− 7.1420.5241.0000.000Pt− 12.3900.4390.0050.005Pa− 6.5860.4721.0000.326 Table 6Westerlund cointegration.StatisticsValueZ-valueP-valueRobust P-valueGt− 2.5760.9170.8260.152Ga− 7.1420.5241.0000.000Pt− 12.3900.4390.0050.005Pa− 6.5860.4721.0000.326 Westerlund cointegration. Long-run estimation of IFE and D-CCE The objective of this study is to analyze the association between GDP per capita and carbon emissions (CO2), with a specific focus on the E-7 countries. An extended examination of GDP per capita demonstrated a robust and statistically significant positive influence on CO2 emissions. The research indicates that for E-7 countries, an increase of one unit in GDPC levels leads to an increase of 0.065%, 0.154%, 0.216%, and 0.267% in CO2 levels in the IFE test and 0.180%, 0.374%, 0.491%, and 0.205% in the D-CCE test. The findings suggest an inverse association between increased economic growth in E-7 nations and a decrease in environmental sustainability. Our conclusion is corroborated by several studies conducted in specific polluted economies, such73,74, which have verified the positive correlation between GDP per capita and CO2 emissions. The industrial sector, which uses a lot of energy, is what E-7 countries towards moving toward. Nevertheless, the upward trend in economic growth is linked to the fulfillment of human needs through consumption and development, resulting in heightened levels of pollution, waste, and environmental degradation75. Hence, economic activities seem to be conducive to environmental preservation and progress rather than posing a threat to long-term ecological well-being76. This discovery implies that E-7 countries rely on energy as a primary source of production, and their economic progress is crucial for providing the necessary resources for sustained development. Implementing renewable energy sources will not only enhance the economic growth of E-7 countries and reduce environmental deterioration, but also enable the country to assume a leading role in the global economic system and enhance its financial competitiveness with more advanced nations. At a significance threshold of five percent, the computed coefficient of URB showed a statistically significant positive correlation with CO2 emissions in E-7 countries, in line with the preliminary expectation. In the long run, a percentage increase in urbanization, assuming all other factors remain constant, resulted in an increase of around 0.221%, 0.105%, 0.254%, and 0.415% in the IFE test and 0.119%, 0.157%, 0.125%, and 0.141% in the D-CCE test in CO2 emissions. This result illustrates that the urbanization of E-7 countries made a substantial contribution to the rise in energy consumption, mostly fueled by fossil fuels, therefore exacerbating environmental degradation. This conclusion is not unexpected, considering that CO2 emissions are generated by these networks, including transportation, demand for electrical appliances, and the construction of commercial and residential buildings. Instead of prioritizing long-term environmental sustainability, civilizations in the initial phases of urbanization often prioritize maximizing comfort by depending on personal automobiles. The study outcomes are consistent with the findings of Issatayeva77, which show a positive correlation between urbanization and CO2 emissions. The results suggest that for every one-unit increase in digital technology (DT), there is a corresponding decrease in carbon emissions (CO2) of 0.526%, 0.390%, 0.451%, and 0.186% in the IFE test, and 0.390%, 0.651%, 0.314%, and 0.167% in the D-CCE test. This, in turn, leads to a substantial improvement in energy-use efficiency and a direct or indirect reduction in carbon emissions. As a result, the use of conventional fossil fuels is gradually being replaced by RE, the growth of RE is continuously encouraged, systems for producing and consuming energy can be optimized and improved, and overall urban carbon emissions are significantly reduced . The causal factor for this outcome can be attributed to the prevalence of Internet shopping facilitated by sophisticated ICT technologies, decreased energy usage, enhanced web accessibility, accelerated downloading speed, online payment transactions, and efficient product shipment. Second, these countries possess highly advanced financial systems and sectors that have played a significant role in the advancement and effective utilization of the Internet and mobile cellular technologies. Another potential explanation for this phenomenon may be the effective and ecologically sustainable use of fossil fuels in the economies of E-7 countries. All these elements can elucidate the beneficial correlation between mobile and Internet usage and environmental quality. Previous scholarly research has mostly concentrated on the macro-level factors of how digital technology contributes to achieving carbon neutrality at national or international scales. Existing literature has focused on several macro aspects of distributed technology (DT), such as emissions forecasting models78, ensemble prediction systems that offer both point and interval estimates of future emissions79, and predictive models that can estimate global warming80. In recent years, policy analysts have shown significant interest in CE, which aims to repurpose the resources utilized for various reasons, including energy production. Hence, this study evaluates CE as an environmentally friendly indicator of sustainability and reveals intriguing results. As an illustration, an increase of 1% in this variable (CE) would cause a reduction of CO2 emissions by 0.27%, 0.5099%, 0.3266%, 0.6111%, and 0.6277%, 0.6011%, 0.4600%, 0.6788% at a significance level of 5%. Furthermore, the analysis of IFE and D-CCE tests confirms that any variation in the CE would contribute to the reduction of emissions in E-7 economies. Consequently, the increasing upward trajectory of greenhouse gas emissions resulting from extensive inefficiency in human and economic activities has emerged as a grave menace to humanity. To mitigate the decline in biodiversity, highly structured economies have embraced an environmentally friendly approach of repurposing such waste for energy production. From this stage, substantial energy demand can be reduced with minimal loss of biodiversity. Fundamentally, CE is seen as a crucial element in the era of development to address increasing environmental strain81. Furthermore, CE not only safeguards environmental integrity but also fosters the preservation of resources and the attainment of sustainable development in E-7 countries. The findings suggest that a one-unit increase in PM resulted in a reduction in carbon emissions (CO2) levels by 0.263%, 0.612%, 0.162%, and 0.525% in the IFE test and by 0.622%, 0.716%, 0.159%, and 0.234% in the D-CCE test. Furthermore, economic governance can also contribute to ecological quality through public management, contingent on the effectiveness of governance. An analysis of the data indicates that a rise in the efficiency of public management will lead to a proportional reduction in carbon emissions (CO2) in the economies of E-7 countries. In summary, this study demonstrated that effective governance has a detrimental influence on CO2 emissions, thereby facilitating the regulation of its generation and consequences, in line with the findings of other scholars82. Evidence indicates that E-7 economies are facing ecological degradation as an outcome of the performance of public management institutions and the implementation of environmental protection measures. These nations’ economic activities are significantly impacted by the operational efficacy of their institutions. One could hypothesize that improving economic productivity through public management could increase trade and financial activities, reduce inequality, and increase economic productivity, all of which would increase the impact on CO2 emissions. Indicators that government agencies may prioritize include the political and legal framework, adequate funding, feedback mechanisms, perceived ease of citizen engagement, and influential individuals who promote public participation. These factors contribute to the generation of greater value through open government data, which strengthens institutional quality and enhances environmental welfare. The results of this study are similar to those reported by Alola and Adebayo83 for E-7 economies (Table 7). According to Tanveer84, achieving an equitable distribution of income and power indicates improved institutional quality and facilitates more effective management of pollution and other environmental issues. With a robust institutional framework, the governments of E-7 economies may legalize and regulate emissions more effectively. This is because they can implement various measures without external pressure from any particular party. The rule of law is crucial for policy enforcement, as it compels industries and firms to adhere to the principles of environmental regulations to mitigate environmental deterioration.Table 7Long Run IFE and D-CCE Tests Results.VariablesModel 1Model 2Model 3Model 4lnGDPC0.065 *** (0.000) [0.056, 0.074]0.154 *** (0.006) [0.142, 0.166]0.216 *** (0.000) [0.201, 0.231]0.267 *** (0.005) [0.251, 0.283]lnURB0.221 *** (0.000) [0.198, 0.244]0.105 *** (0.001) [0.093, 0.117]0.254 ** (0.019) [0.220, 0.288]0.415 *** (0.005) [0.397, 0.433]lnDT − 0.526 *** (0.000) [− 0.558, − 0.494] − 0.390 ** (0.012) [− 0.422, − 0.358] − 0.451 *** (0.005) [− 0.477, − 0.425] − 0.186 ** (0.012) [− 0.210, − 0.162]lnCE − 0.197 *** (0.009) [− 0.223, − 0.171] − 0.509 ** (0.014) [− 0.545, − 0.473] − 0.326 *** (0.005) [− 0.352, − 0.300] − 0.611 *** (0.000) [− 0.637, − 0.585]lnPM − 0.263 *** (0.004) [− 0.287, − 0.239] − 0.612 *** (0.007) [− 0.634, − 0.590] − 0.162 *** (0.001) [− 0.181, − 0.143] − 0.525 ** (0.032) [− 0.558, − 0.492]lnENT0.142 ** (0.019) [0.118, 0.166]0.425 *** (0.000) [0.405, 0.445]0.210 ** (0.012) [0.186, 0.234]0.317 *** (0.000) [0.299, 0.335]lnPM × DT − 0.054 *** (0.005) [− 0.066, − 0.042]lnPM × CE − 0.138 *** (0.005) [− 0.150, − 0.126]lnPM × ENT − 0.423 ** (0.012) [− 0.447, − 0.399]Long-run D-CCE test resultslnGDPC0.180 *** (0.000) [0.165, 0.195]0.374 *** (0.005) [0.352, 0.396]0.491 *** (0.005) [0.469, 0.513]0.205 *** (0.005) [0.187, 0.223]lnURB0.119 *** (0.005) [0.107, 0.131]0.157 *** (0.005) [0.141, 0.173]0.125 *** (0.001) [0.113, 0.137]0.141 *** (0.000) [0.129, 0.153]lnDT − 0.390 *** (0.000) [− 0.414, − 0.366] − 0.651 *** (0.000) [− 0.679, − 0.623] − 0.314 ** (0.021) [− 0.340, − 0.288] − 0.167 ** (0.015) [− 0.189, − 0.145]lnCE − 0.127 *** (0.000) [− 0.143, − 0.111] − 0.601 ** (0.024) [− 0.629, − 0.573] − 0.460 *** (0.005) [− 0.486, − 0.434] − 0.678 *** (0.000) [− 0.702, − 0.654]lnPM − 0.622 *** (0.004) [− 0.646, − 0.598] − 0.716 *** (0.007) [− 0.740, − 0.692] − 0.159 *** (0.001) [− 0.175, − 0.143] − 0.234 ** (0.000) [− 0.252, − 0.216]lnENT0.039 *** (0.006) [0.027, 0.051]0.015 *** (0.000) [0.003, 0.027]0.134 *** (0.000) [0.118, 0.150]0.107 *** (0.000) [0.091, 0.123]lnPM × DT − 0.412 *** (0.005) [− 0.430, − 0.394]lnPM × CE − 0.142 *** (0.001) [− 0.158, − 0.126]lnPM × ENT − 0.290 ** (0.012) [− 0.312, − 0.268]Robust standard errors clustered by country are reported in parentheses. 95% confidence intervals are provided in brackets. ***p < 0.01, **p < 0.05. Table 7Long Run IFE and D-CCE Tests Results.VariablesModel 1Model 2Model 3Model 4lnGDPC0.065 *** (0.000) [0.056, 0.074]0.154 *** (0.006) [0.142, 0.166]0.216 *** (0.000) [0.201, 0.231]0.267 *** (0.005) [0.251, 0.283]lnURB0.221 *** (0.000) [0.198, 0.244]0.105 *** (0.001) [0.093, 0.117]0.254 ** (0.019) [0.220, 0.288]0.415 *** (0.005) [0.397, 0.433]lnDT − 0.526 *** (0.000) [− 0.558, − 0.494] − 0.390 ** (0.012) [− 0.422, − 0.358] − 0.451 *** (0.005) [− 0.477, − 0.425] − 0.186 ** (0.012) [− 0.210, − 0.162]lnCE − 0.197 *** (0.009) [− 0.223, − 0.171] − 0.509 ** (0.014) [− 0.545, − 0.473] − 0.326 *** (0.005) [− 0.352, − 0.300] − 0.611 *** (0.000) [− 0.637, − 0.585]lnPM − 0.263 *** (0.004) [− 0.287, − 0.239] − 0.612 *** (0.007) [− 0.634, − 0.590] − 0.162 *** (0.001) [− 0.181, − 0.143] − 0.525 ** (0.032) [− 0.558, − 0.492]lnENT0.142 ** (0.019) [0.118, 0.166]0.425 *** (0.000) [0.405, 0.445]0.210 ** (0.012) [0.186, 0.234]0.317 *** (0.000) [0.299, 0.335]lnPM × DT − 0.054 *** (0.005) [− 0.066, − 0.042]lnPM × CE − 0.138 *** (0.005) [− 0.150, − 0.126]lnPM × ENT − 0.423 ** (0.012) [− 0.447, − 0.399]Long-run D-CCE test resultslnGDPC0.180 *** (0.000) [0.165, 0.195]0.374 *** (0.005) [0.352, 0.396]0.491 *** (0.005) [0.469, 0.513]0.205 *** (0.005) [0.187, 0.223]lnURB0.119 *** (0.005) [0.107, 0.131]0.157 *** (0.005) [0.141, 0.173]0.125 *** (0.001) [0.113, 0.137]0.141 *** (0.000) [0.129, 0.153]lnDT − 0.390 *** (0.000) [− 0.414, − 0.366] − 0.651 *** (0.000) [− 0.679, − 0.623] − 0.314 ** (0.021) [− 0.340, − 0.288] − 0.167 ** (0.015) [− 0.189, − 0.145]lnCE − 0.127 *** (0.000) [− 0.143, − 0.111] − 0.601 ** (0.024) [− 0.629, − 0.573] − 0.460 *** (0.005) [− 0.486, − 0.434] − 0.678 *** (0.000) [− 0.702, − 0.654]lnPM − 0.622 *** (0.004) [− 0.646, − 0.598] − 0.716 *** (0.007) [− 0.740, − 0.692] − 0.159 *** (0.001) [− 0.175, − 0.143] − 0.234 ** (0.000) [− 0.252, − 0.216]lnENT0.039 *** (0.006) [0.027, 0.051]0.015 *** (0.000) [0.003, 0.027]0.134 *** (0.000) [0.118, 0.150]0.107 *** (0.000) [0.091, 0.123]lnPM × DT − 0.412 *** (0.005) [− 0.430, − 0.394]lnPM × CE − 0.142 *** (0.001) [− 0.158, − 0.126]lnPM × ENT − 0.290 ** (0.012) [− 0.312, − 0.268]Robust standard errors clustered by country are reported in parentheses. 95% confidence intervals are provided in brackets. ***p < 0.01, **p < 0.05. Long Run IFE and D-CCE Tests Results. Robust standard errors clustered by country are reported in parentheses. 95% confidence intervals are provided in brackets. ***p < 0.01, **p < 0.05. The study found that the energy transition variable has a profound and substantial impact on the long-term rise in CO2 emissions in the E-7 economies. Each incremental score or 5% increase in ENT resulted in a corresponding increase in CO2 growth of 0.142%, 0.425%, 0.210%, 0.317%, and 0.039%, 0.015%, 0.134%, and 0.107% in the long term, respectively. Furthermore, the impact is comparatively less for the average and highest quantiles. The observed result can be attributed to the persistently high demand for fossil fuels in the E-7 economies. Notwithstanding the allocation of resources towards renewable energy, fossil fuels continue to constitute the bulk of investments in energy sources. Industries such as transportation and manufacturing continue to rely significantly on fossil fuel resources, which account for over 65% of the CO2 emissions from conventional energy sources85. Significantly, the energy revolution accelerated with time. The decline in CO2 emissions caused by the COVID-19 pandemic will primarily contribute to the continuation of the previously documented increasing trend in a flat manner, which has been ongoing for the past 20 years86. Therefore, it is clear that the long-term effects of renewable energy transition on reducing CO2 emissions are far greater than those observed in the short term. This indicates that a continuous rise in per capita energy transition consumption is expected to consistently reduce CO2 emissions in E-7 economies. These study results align with the findings documented in previous research conducted by87–89 for the E-7 economies. In the long-run estimation of the D-CCE test, the outcomes show that a 1 unit increase in the moderating role of public management (PM) with digital technology, CE, and ENT would lead to a decrease in the level of carbon emission (CO2) in E-7 economies by 0.412%, 0142%, 0.290%, and 0.054%, 0.138%, and 0.423%, respectively. These findings support the theoretical expectation that governance capacity moderates technological and circular effects. In economies with stronger public management, environmental regulations and administrative efficiency enhance the ability of digital innovation and circular initiatives to deliver measurable emission reductions consistent with institutional-theory explanations presented earlier. Panel quantile regression (PQR) for heterogeneity analysis Guo90 proposes that time-period fixed-effect models should include any time-specific, spatial-invariant variables that, if removed, could potentially bias the results in typical time-series studies. Therefore, we place greater importance on the results of the model incorporating two-way fixed effects. In addition, the autocorrelation test was employed. We employed the one-step methodology recommended by91,92 to conduct thorough examinations of our conditional panel quantile findings. In addition, we utilized the two-way ordinary least squares (OLS) estimation method proposed by Slorach93. To account for distributional heterogeneity, the fixed-effect quantile regression method developed by Shipalana94 was employed. The findings of PQR for several percentile levels: 15th, 30th, 45th, 60th, 75th, and 90th are shown in Table 8. The findings suggest that there is a positive correlation between GDP per capita and urbanization and the level of carbon emissions (CO2) in all quantiles of the E-7 economies. Digital technology (DT) exerts a detrimental and substantial influence on carbon emissions (CO2).Table 8Panel Quantile regression results.Variables15th30th45th60th75th90thlnGDPC0.190 *** (0.000) [0.175, 0.205]0.231 *** (0.009) [0.213, 0.249]0.346 *** (0.005) [0.330, 0.362]0.245 *** (0.001) [0.229, 0.261]0.219 *** (0.000) [0.203, 0.235]0.472 *** (0.002) [0.452, 0.492]lnURB1.428 *** (0.001) [1.392, 1.464]1.378 *** (0.000) [1.340, 1.416]1.690 *** (0.000) [1.654, 1.726]1.612 *** (0.007) [1.578, 1.646]1.521 *** (0.001) [1.493, 1.549]1.934 *** (0.000) [1.896, 1.972]lnDT − 0.129 *** (0.000) [− 0.143, − 0.115] − 0.175 *** (0.005) [− 0.191, − 0.159] − 0.196 ** (0.012) [− 0.212, − 0.180] − 0.209 *** (0.007) [− 0.225, − 0.193] − 0.146 ** (0.021) [− 0.164, − 0.128] − 0.085 *** (0.006) [− 0.099, − 0.071]lnCE − 0.184 *** (0.000) [− 0.200, − 0.168] − 0.168 *** (0.001) [− 0.182, − 0.154] − 0.155 *** (0.000) [− 0.169, − 0.141] − 0.140 *** (0.000) [− 0.154, − 0.126] − 0.159 *** (0.000) [− 0.173, − 0.145] − 0.161 ** (0.005) [− 0.177, − 0.145]lnPM − 0.265 *** (0.000) [− 0.279, − 0.251] − 0.321 *** (0.001) [− 0.337, − 0.305] − 0.349 ** (0.020) [− 0.365, − 0.333] − 0.257 *** (0.000) [− 0.271, − 0.243] − 0.134 *** (0.005) [− 0.150, − 0.118] − 0.492 ** (0.015) [− 0.510, − 0.474]lnENT0.019 ** (0.012) [0.003, 0.035]0.189 ** (0.000) [0.173, 0.205]0.215 ** (0.005) [0.199, 0.231]0.172 *** (0.000) [0.156, 0.188]0.166 ** (0.025) [0.148, 0.184]0.071 ** (0.000) [0.055, 0.087]Robust standard errors clustered by country are reported in parentheses. 95% confidence intervals are provided in brackets. ***p < 0.01, **p < 0.05. Table 8Panel Quantile regression results.Variables15th30th45th60th75th90thlnGDPC0.190 *** (0.000) [0.175, 0.205]0.231 *** (0.009) [0.213, 0.249]0.346 *** (0.005) [0.330, 0.362]0.245 *** (0.001) [0.229, 0.261]0.219 *** (0.000) [0.203, 0.235]0.472 *** (0.002) [0.452, 0.492]lnURB1.428 *** (0.001) [1.392, 1.464]1.378 *** (0.000) [1.340, 1.416]1.690 *** (0.000) [1.654, 1.726]1.612 *** (0.007) [1.578, 1.646]1.521 *** (0.001) [1.493, 1.549]1.934 *** (0.000) [1.896, 1.972]lnDT − 0.129 *** (0.000) [− 0.143, − 0.115] − 0.175 *** (0.005) [− 0.191, − 0.159] − 0.196 ** (0.012) [− 0.212, − 0.180] − 0.209 *** (0.007) [− 0.225, − 0.193] − 0.146 ** (0.021) [− 0.164, − 0.128] − 0.085 *** (0.006) [− 0.099, − 0.071]lnCE − 0.184 *** (0.000) [− 0.200, − 0.168] − 0.168 *** (0.001) [− 0.182, − 0.154] − 0.155 *** (0.000) [− 0.169, − 0.141] − 0.140 *** (0.000) [− 0.154, − 0.126] − 0.159 *** (0.000) [− 0.173, − 0.145] − 0.161 ** (0.005) [− 0.177, − 0.145]lnPM − 0.265 *** (0.000) [− 0.279, − 0.251] − 0.321 *** (0.001) [− 0.337, − 0.305] − 0.349 ** (0.020) [− 0.365, − 0.333] − 0.257 *** (0.000) [− 0.271, − 0.243] − 0.134 *** (0.005) [− 0.150, − 0.118] − 0.492 ** (0.015) [− 0.510, − 0.474]lnENT0.019 ** (0.012) [0.003, 0.035]0.189 ** (0.000) [0.173, 0.205]0.215 ** (0.005) [0.199, 0.231]0.172 *** (0.000) [0.156, 0.188]0.166 ** (0.025) [0.148, 0.184]0.071 ** (0.000) [0.055, 0.087]Robust standard errors clustered by country are reported in parentheses. 95% confidence intervals are provided in brackets. ***p < 0.01, **p < 0.05. Panel Quantile regression results. Robust standard errors clustered by country are reported in parentheses. 95% confidence intervals are provided in brackets. ***p < 0.01, **p < 0.05. Empirical research suggests that a single-unit increase in digital technology is associated with a decrease in carbon emissions (CO2) across all quantile levels. Research indicates that CE significantly and negatively impacts carbon emissions in E-7 economies. The results indicate that an increase of one unit in public management activities would result in a decrease in carbon emissions (CO2) across all quantile levels in the E-7 economies. Ultimately, the findings indicate that the transition to renewable energy sources has a beneficial impact on the decrease in carbon emissions (CO2) in the E-7 economies. The findings suggest that an increase of one unit in energy transition leads to a proportional increase in carbon emissions. The findings suggest that incorporating digital technology, CE, and energy transition into the interaction model of public administration would cause a decrease in carbon emissions (CO2) in E-7 economies. Although positive coefficient for the energy transition appears counterintuitive, similar transitional dynamics have been observed in the emerging economies. Three mechanisms may explain this result: (a) transitional emissions generated during manufacturing of the renewable infrastructure (e.g., solar panels, wind turbines); (b) rebound effects, where improved energy access stimulates the additional consumption; and (c) measurement limitations, as renewable shares may increase while total energy demand also rises, masking absolute fossil-fuel growth. Recognizing these dynamics underscores complexity of the decarbonization process in the rapidly developing contexts. Conclusion and policy recommendations Conclusion Primary objective of this study was to quantitatively examine how key country-specific factors GDP per capita, urbanization, digital technology (DT), circular economy (CE), public management (PM), and the energy transition (ENT) influence the carbon dioxide (CO₂) emissions in E-7 economies from 2000 to 2023. To ensure about the robustness, the study adopted second-generation econometric techniques that account for cross-sectional dependence and heterogeneity among nations. After confirming existence of the cross-sectional dependency, the Westerlund cointegration method was applied, and results verified a long-run equilibrium relationship among selected variables. Estimates obtained through Interactive Fixed Effects (IFE) and Dynamic Common Correlated Effects (D-CCE) models has demonstrated that the economic growth, urbanization, and energy transition tend to increase CO₂ emissions, while the digital technology, circular economy practices, and public management significantly reduce them. Panel Quantile Regression (PQR) has approached further validated these outcomes by showing the consistent results across various quantiles of the CO₂ distribution. Importantly, interaction terms lnPM × DT, lnPM × CE, and lnPM × ENT has revealed that the strong governance amplifies positive environmental effects of digitalization, circular economy activities, and the energy transition, thereby confirming moderating role of the institutional quality in sustainable development. Findings yield several significant implications for environmental policy in E-7 economies. Policymakers must prioritize to the formulation of integrated low-carbon strategies that balance thr environmental sustainability with economic expansion. Establishing comprehensive ecological frameworks can reduce CO₂ emissions while sustaining growth. To achieve this, governments should create the long-term regulatory incentives for businesses to invest in the emission-reducing technologies, carbon capture and storage (CCS), renewable-energy deployment, and emission trading systems. These mechanisms can help to shift production and consumption patterns toward the more sustainable trajectories. Enhancing the quality of public management is another critical prerequisite for sustainability. Effective institutions and transparent governance structures are necessary for implementing the environmental policies that work in practice, not just in theory. In the E-7 context, improvements in anti-corruption measures, regulatory enforcement, and administrative efficiency can enhance to the institutional environment in which green technologies operate. Moderating relationships identified in this study particularly lnPM × DT, lnPM × CE, and lnPM × ENT has demonstrated that the good governance acts as a catalyst, converting the potential benefits of the technological and circular innovations into tangible environmental outcomes. Transition to a digital economy represents a further avenue for the reduction of emissions. Governments in E-7 economies should strengthen the financial mechanisms that lower the initial investment barriers faced by firms seeking to adopt the advanced technologies such as blockchain, the Internet of Things (IoT), and cloud computing. These technologies can enhance the transparency, improve energy efficiency, and facilitate the carbon tracking. For instance, China’s Digital Green Certificate System (State Council, 2023) employs blockchain to verify the renewable-energy transactions, illustrating that how digital transformation and governance interact synergistically to improve sustainability performance. Equally important is the promotion of circular economy principles across industrial sectors. Empirical evidence has shown that circular economy practices such as waste recovery, recycling, and resource reuse are effective tools for reducing CO₂ emissions. Governments should encourage private sector to transition from a linear “produce-use-dispose” model toward a circular one by offering the targeted incentives, including tax credits, grants, and low-interest financing. Brazil’s National Solid Waste Policy (Law 12.305/2010), which has institutionalized the government oversight of waste management and recycling, is an example of successful PM × CE integration that other E-7 nations can emulate. These initiatives not only conserve the natural resources but also create employment opportunities within green industries. Urbanization remains a significant source of the emissions, yet it can also serve as a platform for innovation if managed wisely. Governments of E-7 economies should promote the sustainable urbanization through improved spatial planning, renewable-energy infrastructure, low-carbon transportation, and energy-efficient construction. India’s Smart Cities Mission exemplifies how real-time IoT-based monitoring systems can optimize to the urban resource use and minimize environmental impact. Through similar initiatives, cities can become laboratories of sustainability that has contributed to national emission-reduction targets rather than undermining them. Transforming the traditional energy systems of E-7 economies is equally vital. Policymakers must increase the use of renewable energy technologies and simultaneously has reduce the dependence on fossil fuels. Introducing progressive tariffs on petroleum imports, eliminating fossil-fuel subsidies, and investing in indigenous renewable resources can strengthen the energy security while achieving emission-reduction objectives. Temporary rise in emissions during early stages of energy transition, as found in this study, reflects the transitional dynamics rather than policy failure. Over the time, as renewable capacity expands and efficiency improves, these transitional emissions are expected to decline. Finally, E-7 governments should invest in the research and development (R&D) that fosters innovation in energy efficiency, green manufacturing, and sustainable production. Collaborative networks between academia, industry, and government can accelerate to the creation and diffusion of clean technologies. In summary, coherent governance frameworks, targeted financial support, and institutional reforms are essential for maximizing joint environmental benefits of the digital technology, circular economy, and energy transition in the emerging economies. Limitations and future studies While study offers the novel empirical insights, it is not without limitations. The analysis focuses exclusively on E-7 economies due to the data availability, and its conclusions may not fully apply to other regional or economic contexts. Future research should broaden scope to include the additional country groups such as G-7, OECD, BRICS, MENA, and other Asian nations to test generalizability of the results. Moreover, circular economy was measured the primarily through energy recovery from waste, which captures only a narrow aspect of the circularity. Future work should employ more comprehensive indicators, including material circularity rates, recycling intensities, and resource-productivity measures, to provide the more complete understanding of circular-economy performance. Another limitation lies in exclusion of the potentially influential variables such as trade openness, foreign direct investment, energy-price subsidies, and sectoral composition. Incorporating these factors in the future analyses could enhance model accuracy and better capture the multifaceted drivers of the emissions. Furthermore, although econometric techniques used IFE, D-CCE, and PQR account for the cross-sectional dependence and heterogeneity, future studies might consider nonlinear or time-varying frameworks such as Nonlinear Autoregressive Distributed Lag (NARDL) or Quantile NARDL models. These approaches could reveal the asymmetric or evolving relationships between economic development, institutional quality, and environmental performance over the time. Finally, additional investigation into mechanisms through which public management moderates the effects of DT, CE, and ENT on CO₂ emissions would enrich theoretical understanding. Mixed-methods research combining the quantitative models with qualitative case studies could provide deeper insights into how governance structures convert to the policy design into measurable sustainability outcomes. In conclusion, this study demonstrates that the digital technology, circular economy practices, and public management operate as complementary pillars of the sustainability in E-7 economies. Effective governance enhances capacity of technology and circular strategies to deliver the real reductions in carbon emissions. Although energy transition currently exhibits a temporary positive association with the emissions, this pattern has reflected transitional stages of infrastructure development rather than inherent inefficiency. Strengthening of institutional quality, accelerating digital innovation, and embedding circular principles into the national governance and business strategies constitute most viable pathways toward achieving the carbon neutrality and long-term sustainable development in the emerging economies.