The author group conducts this study to understand the relationship between renewable energy consumption, economic growth, foreign direct investment FDI, and innovation with CO2 emission
Reason for doing the topic
Climate change necessitates immediate action, prompting nations to establish ambitious net zero targets However, the path to achieving net zero is complex, involving the interplay of innovation, renewable energy, foreign investment, economic growth, and carbon emissions, which are not fully understood This research focuses on countries with legally mandated net zero targets, providing insights into the relationship between policy, economic development, and environmental impact Key questions explored include how innovation drives the transition to renewable energy, the role of foreign investment in this process, and whether economic growth can be separated from emissions The findings aim to offer valuable strategies and highlight challenges for other nations pursuing their net zero goals, ultimately contributing to a collaborative approach to climate action and a sustainable future.
This study explores the interconnectedness of innovation, renewable energy consumption, foreign direct investment (FDI) inflows, economic growth, and carbon dioxide emissions, building on previous research conducted in developed economies like Vietnam and Algeria With a focus on collaborative technology and the fifth industrial revolution, we aim to deepen our understanding of these relationships To align with the country's long-term goal of reducing CO2 emissions, we have adopted a Net Zero Target, which involves establishing clear policies or laws to achieve net zero emissions Our findings will provide evidence-based insights that can guide future policies and investments tailored to the specific needs and priorities of the nation.
Objectives of the study
This study explores the relationship between innovation, renewable energy consumption, foreign direct investment inflows, economic growth, and CO2 emissions in nations dedicated to achieving net-zero emissions By analyzing these factors, the research aims to uncover how they interact and influence each other in the context of sustainable development and environmental responsibility.
Research questions
Based on , this paper try to figure out 3 main questions
How does innovation in Industry 4.0 technologies, especially in the era of Artificial Intelligence, impact carbon dioxide emissions and adoption in Industry 5.0?
What effect do FDI inflows have on carbon dioxide emissions in Nations committed to achieving Net-zero emissions ?
Can government policies and regulations act as levers to accelerate innovation, shift towards renewable energy, attract foreign investment, and mitigate CO2 emissions ?
Policymakers in the era of Industry 5.0 can effectively harness the synergies between innovation, renewable energy, foreign investment, and economic growth by implementing strategic frameworks that promote sustainable development By fostering collaboration among stakeholders, they can drive technological advancements while ensuring that the benefits are distributed equitably across society It is crucial to address potential unintended consequences through comprehensive regulations and policies that prioritize environmental sustainability and social equity, ultimately leading to a balanced and inclusive economic growth.
2010) How can they achieve this delicate balance, taking into account potential negative externalities and promoting social equity?
Research subjects and scopes
Research subjects
The researchers focus on finding the relationship of innovation, renewable consumption, FD1 inflows, economic growth and carbon dioxide emissions.
Research scopes
To enhance the accuracy of this study, we focus on 40 countries striving for Net Zero Status, reflecting their long-term commitment to fulfilling obligations under the Paris Agreement These nations have legally established their Net Zero goals through statutes or administrative regulations, ensuring that their efforts are backed by legal authority.
The list of countries includes diverse nations such as the Maldives, Finland, Austria, and Iceland, along with Germany, Sweden, Italy, and Poland Other notable mentions are the Netherlands, Belgium, Romania, and several Central and Eastern European countries like the Czech Republic, Slovakia, and Bulgaria Additionally, Croatia, Lithuania, Slovenia, Latvia, Estonia, and Cyprus are featured, alongside Malta, the United States, Japan, and the United Kingdom France, South Korea, Canada, Spain, and Australia are also included, as well as Colombia, Switzerland, Ireland, Chile, Portugal, Hungary, Greece, New Zealand, Luxembourg, Fiji, and Nigeria.
To enhance the validity of our hypothesis and ensure reliable data results, we examined 42 nations that have incorporated net zero targets in their policy documents, such as Nationally Determined Contributions (NDCs) and Long-term Strategies (LTSs) submitted to the UNFCCC, as well as plans released by relevant ministries The nations included in this analysis are Barbados, Dominica, Antigua and Barbuda, Nepal, Brazil, Vietnam, Argentina, Malaysia, the United Arab Emirates, Singapore, Peru, Oman, Ethiopia, Ecuador, Panama, Tunisia, Costa Rica, Uruguay, Cambodia, Lebanon, Laos, Georgia, Papua New Guinea, Namibia, Liberia, The Gambia, Cape Verde, Andorra, Belize, Vanuatu, Tonga, the Marshall Islands, Tuvalu, Monaco, Turkiye, China, the Russian Federation, Saudi Arabia, Ukraine, Kazakhstan, Thailand, and India.
Overall research methodology
This study employs descriptive statistics, a correlation matrix, and a multicollinearity test to analyze relationships between variables We utilize Ordinary Least Squares (OLS), Fixed Effects Model (FEM), and Random Effects Model (REM) regression techniques, applying the Hausman test to determine the appropriate model Subsequently, we assess heteroscedasticity and autocorrelation, employing Feasible Generalized Least Squares (FGLS) to address any model deficiencies This straightforward approach effectively identifies the independent variables impacting CO2 emissions across 82 nations with Net Zero targets from 2000 to 2022.
Practical meanings of the topic
This research analyzes real-world data to reveal the effectiveness of various strategies used by net zero pioneers, providing valuable insights for other nations By leveraging proven successes and steering clear of potential pitfalls, countries can customize their approaches to achieve net zero goals more effectively.
Furthermore, identifying challenges faced by these countries allows for proactive measures to overcome roadblocks on the path to decarbonization.
The research highlights the complex relationship between economic and environmental objectives, demonstrating how net zero policies drive innovation in clean technologies, which in turn promotes green growth and job creation Furthermore, it explores the economic advantages of integrating renewable energy, providing essential insights for nations moving towards sustainable energy solutions.
This research conducts a detailed analysis of foreign direct investment (FDI) inflows, highlighting the potential synergies and challenges in attracting foreign investment that supports net zero goals The findings aim to inform policy-making and incentives that can enhance international collaboration for sustainable development.
This research goes beyond theory to offer practical insights for policymakers, businesses, and global stakeholders By presenting empirical evidence and showcasing effective strategies, it acts as a crucial guide for navigating the complexities of achieving a decarbonized future.
Research layout
This study is divided into five part:
In the introduction, the authors outline the rationale behind selecting the topic, define the study's objectives, and present key research questions and subjects They detail the scope of the research and the overall methodology employed, highlighting the practical implications of the study Additionally, a structured research layout is provided to enhance reader comprehension.
In Part 2 of the article, we explore the theoretical framework and conduct a literature review to examine previous research on financial performance and firm value By summarizing key arguments and viewpoints in a table, we identify the variables that will be utilized in our model and outline the corresponding research hypotheses.
Part 3 - Proposed models and research methods In this section, the authors present the data collection methodology, outline the research model, and describe the steps for conducting the regression analysis.
In Part 4 of the study, the authors detail the empirical results obtained from the regression process They begin with an initial dataset and employ a range of tests and regression models to ensure the reliability and trustworthiness of their findings.
In Part 5 of the study, the authors summarize the key conclusions derived from their research, acknowledging the limitations and challenges faced throughout the process They also offer insightful recommendations and suggest future directions for further advancements in the topic.
LITERATURE REVIEW AND RESEARCH HYPOTHESIS
Factors affecting CO2 Emission
2.1.1 Impact of Innovation on CO2 Emissions
Scholarly discourse categorizes the impact of income, growth, and trade on environmental well-being into three main effects: scale, technique, and composition Scale effects arise from economic activities like manufacturing and waste generation, which increase pollution as the economy expands In contrast, technique effects occur later, where technological advancements help reduce environmental pollutants Composition effects depend on the economic production mix, leading to varying pollution impacts based on a country's comparative advantage, influenced by factors such as resource endowment and pollution policies The factor endowment hypothesis suggests that wealthier nations with abundant capital tend to export goods with higher pollution levels, aligning with the Heckscher-Ohlin theory that promotes a manufacturing and export portfolio based on factor intensity.
Environmental scholars and policymakers are increasingly focused on the need for reforms to address climate challenges linked to fossil fuel use and nonrenewable energy dependence This study aims to analyze how export diversification impacts carbon emissions in both developing and developed countries Export diversification involves either expanding the existing range of exports (intensive margin) or increasing the number of different products exported (extensive margin) The rise in exports is associated with manufacturing and resource consumption, including energy and fossil fuels Therefore, examining the relationship between export diversification indicators can provide valuable insights for economies at different levels of development.
Over the past 30 years, greenhouse gas emissions have significantly increased due to various economic and non-economic activities, including trade, income growth, industrialization, urbanization, and deforestation Urbanization and the reliance on fossil fuels have become major environmental stressors, leading to increased waste and pollution The connection between income levels and pollution is supported by the Environmental Kuznets Curve Hypothesis, highlighting the urgent need for effective climate policies to combat global warming and reduce emissions Key international agreements, such as the United Nations Framework Convention on Climate Change (1992), the Kyoto Protocol (1997), and the Paris Agreement (2015), represent collaborative efforts aimed at environmental protection.
Addressing research gaps in understanding the factors influencing environmental pollution in Vietnam is crucial for promoting a sustainable economy This study utilizes a regression model to analyze the effects of innovation, renewable energy consumption, foreign direct investment (FDI) inflows, and economic growth on CO2 emissions in Vietnam from 2000 to 2022 The results are significant for Vietnamese policymakers striving to enhance GDP, economic vitality, and environmental sustainability in the context of Industry Revolution 5.0.
Innovation has the potential to enhance production and consumption efficiency while significantly lowering environmental impact However, insufficient investment and lack of political commitment may hinder this progress Therefore, it is essential for governments, businesses, and individuals to prioritize research and development, support policies that foster innovation, and create a supportive environment for creative advancements.
2.1.2 Influence of Renewable Energy Consumption on CO2 Emissions
Renewable energy consumption is crucial for reducing carbon dioxide emissions and achieving environmental sustainability As the world moves towards a low-carbon economy, the need to utilize renewable energy sources intensifies Governments, businesses, and individuals play vital roles in promoting the adoption of renewable energy by investing in necessary infrastructure, providing incentives for its development, and incorporating renewable technologies into daily life.
This article explores the complex interplay between per capita Carbon Dioxide (CO2) emissions, the Economic Complexity Index (ECI), renewable energy, and inward Foreign Direct Investment (FDI), based on extensive research The BRICS countries, characterized by varying dependencies on fossil fuels, trade patterns, renewable energy usage, and CO2 emissions, have been rigorously examined As CO2 emissions significantly impact the environment, they have become central to research, analyzed through various perspectives such as electricity consumption, trade openness, financial development, renewable energy adoption, and the Environmental Kuznets Curve (EKC) framework.
Research highlights a U-shaped relationship between ecological degradation and economic growth, indicating that environmental harm increases before decreasing after reaching a specific income level Per capita CO2 emissions are crucial to the Environmental Kuznets Curve (EKC) hypothesis, reflecting factors such as policy measures, income inequality, and energy intensity The heavy dependence on fossil fuels is a major contributor to environmental degradation, intensifying CO2 emissions through both consumption and production activities.
Industry 4.0 represents a revolutionary epoch in industrial evolution, marked by enhanced production efficiency facilitated by cutting-edge technologies including the Internet of Things (IoT), Big Data Analytics, Cloud Computing, 3D Printing, Augmented Reality, and Robotic Systems The adoption of Industry 4.0 aspires towards multifaceted improvements spanning economic efficiency, automation, safety, and critically, bolstered environmental preservation Nevertheless, the environmental repercussions of Industry 4.0, encompassing its broader impact on ecosystems, labor markets, and ancillary factors, remain enigmatic, necessitating further comprehensive exploration.
The BRICS nations encompass diverse economies, with Russia reliant on fossil fuels and South Africa focused on mining, metallurgy, and automobile production, which significantly impact environmental pollution A comprehensive understanding of the relationship between Industry 4.0 and CO2 emissions is essential, emphasizing the need for sustainable development grounded in environmental stewardship, social equity, and economic prosperity Achieving Society 5.0 requires aligning the advancements of Industry 4.0 with the UN's 17 Sustainable Development Goals.
2.1.3 Impact of FDI Inflows on CO2 Emissions
Foreign direct investment (FDI) inflows are crucial for fostering innovation and economic growth by introducing new technologies and expertise Multinational corporations are increasingly prioritizing environmental preservation and the reduction of CO2 emissions However, without sustainable investment practices, FDI could negatively impact the environment Thus, it is essential for governments to establish policies that promote sustainable practices, ensuring that FDI aligns with environmental and social sustainability objectives.
The Economic Complexity Index (ECI) offers valuable insights into a country's expected income levels, economic growth, and greenhouse gas emissions By analyzing data that links regions to their main economic activities, the ECI evaluates an economy's potential based on its production, exports, knowledge, and quality Research highlights the relationship between ECI and environmental degradation, showing that sustainable environmental progress is connected to cultural factors and GDP Lower ECI levels are associated with higher CO2 emissions, energy consumption, and greenhouse gas emissions, reflecting a decline in economic complexity.
Foreign Direct Investment (FDI) significantly impacts CO2 emissions, as demonstrated by the Pollution Haven Hypothesis (PHV), which suggests that countries with lenient environmental regulations attract more FDI, resulting in increased pollution Conversely, the Pollution Halo Hypothesis (PHH) indicates that nations with strict environmental standards benefit from FDI, leading to improved environmental conditions.
2.1.4 Impact of Economic Growth on CO2 Emissions:
Economic growth can drive prosperity but also risks environmental degradation due to increased energy use and carbon emissions Many developing countries focus on economic progress at the expense of environmental health, highlighting the need to separate growth from ecological damage To achieve this, it is essential to adopt sustainable production and consumption methods, invest in renewable energy, and implement circular economy principles Reducing carbon dioxide emissions is crucial for addressing climate change, requiring policy measures that support renewable energy, sustainability, public awareness, and global collaboration.
The intricate relationship between innovation, renewable energy use, foreign direct investment (FDI), economic growth, and carbon dioxide emissions highlights the necessity for a comprehensive strategy to promote sustainable development Developing nations frequently emphasize export strategies and attracting FDI to boost economic growth; however, achieving alignment with the Sustainable Development Goals (SDGs) requires a commitment to environmental protection and climate action.
Emergence of Smart Factories
Smart factories are a key component of the Fourth Industrial Revolution, signifying a transformative shift in manufacturing by utilizing advanced technologies to optimize production and improve efficiency These digitally integrated facilities harness the power of IoT, AI, machine learning, and robotics, fundamentally changing traditional manufacturing methods and providing unmatched flexibility and adaptability (Chen et al., 2023).
Key Features of Smail Factories:
1 Interconnectivity: Smart factories facilitate seamless communication between machines and systems through loT, enabling real-time data exchange and collaboration for enhanced decision-making (Kuo and Wu, 2023).
2 Data-driven decision-making: Leveraging big data analytics and AI, smart factories analyze vast datasets to optimize operations, predict issues, and enhance overall efficiency.
3 Automation and robotics: Advanced robotics minimize human intervention, boosting productivity, ensuring consistency, and reducing errors.
4 Adaptive manufacturing: Digital technologies enable swift adaptation to market changes and customer demands, facilitating the production of customized products with shorter lead times.
5 Enhanced supply chain management: Real-time data exchange throughout the supply chain enables better coordination, reduced lead times, and improved inventory management.
Advantages of Smart Factories in Industry 5.0
I Heightened productivity: Leveraging advanced automation and data analytics streamlines operations, eradicating inefficiencies and bolstering output.
2 Elevated product standards: Integration of AI, machine learning, and robotics ensures unwavering quality by swiftly identifying flaws and reducing inaccuracies.
3 Cost efficiency: Efficiency enhancements and flexibility contribute to reduced operational expenses, energy consumption, and waste production.
4 Augmented competitive edge: Embracing digital advancements positions enterprises to thrive amidst the rapid evolution of Industry 5.0.
5 Embracing sustainability: Intelligent factories champion eco-friendly practices by optimizing resource utilization and curbing environmental footprints (Noh et al., 2023).
The advent of smart factories marks a transformative epoch characterized by heightened efficiency, adaptability, and sustainability, propelling innovation and competitiveness within the global manufacturing arena.
CO2 Emissions and Climate Change
The link between CO2 emissions and climate change highlights the critical need to reduce greenhouse gas emissions to combat global warming Various sources contribute to CO2 emissions, making it essential to focus on their reduction for the sake of environmental sustainability.
Research indicates that fiscal decentralization significantly influences CO2 emissions, necessitating tailored environmental policy strategies (Oates, 1999; Sigman, 2003) Furthermore, investigations reveal a nonlinear relationship between fiscal decentralization and CO2 emissions, underscoring the critical role of regulatory factors in moderating this impact (Wang et al., 2018; Akhmetova et al., 2021).
Renewable energy consumption, foreign direct investment (FDI) inflows, and economic growth significantly impact CO2 emissions, highlighting the need for coordinated sustainability efforts Innovation and technological advancements are crucial in decreasing carbon dioxide emissions by improving efficiency and encouraging the adoption of renewable energy sources.
In summary, tackling CO2 emissions and climate change necessitates a comprehensive strategy that includes policy measures, technological advancements, and sustainable practices Through collaboration and the implementation of integrated approaches, countries can create a pathway toward a greener and more sustainable future.
Summary of past researches
Table 2.1 Summary of past researches
Author Title Space & Time Research
Nexus of innovation, renewable consumption, FDI, growth and CO2 emissions:
The research employs panel data methods to analyze the relationship between CO2 emissions and various factors including
The research findings indicate significant relationships between various factors and CO2 emissions in Vietnam from 2000 to
2022 Innovation, as measured by high- tech export value, innovation, renewable energy consumption,
FDI inflows, and economic growth in
Vietnam Panel data techniques address issues related to short time series data and allow for the consideration of regional heterogeneity
The study uses the Cobb-
Douglas function and ordinary least squares (OLS) regression to estimate the influence of independent variables on
CO2 emissions. shows a positive association with CO2 emissions, suggesting that an increase in high-technology exports leads to a rise in CO2 emissions
Conversely, renewable energy consumption exhibits a negative correlation with CO2 emissions, indicating that higher renewable energy consumption is associated with lower CO2 emissions.
Foreign Direct Investment (FDI) inflows and GDP are positively correlated with CO2 emissions, indicating that increases in these economic factors contribute to higher pollution levels This highlights the urgent need for Vietnam to adopt green energy consumption and sustainable development practices to reduce environmental impact The regression model indicates a complex relationship between innovation, renewable energy consumption, FDI inflows, GDP, and CO2 emissions, offering valuable insights for policymakers to focus on green growth strategies aimed at achieving long-term environmental sustainability.
Globalization, financial development, economic growth, environmental pollution, and renewable energy
This study uses the auto regressive distributed lag
Research indicates that globalization positively influences renewable energy consumption in Vietnam However, it also shows that financial development does not have a lasting effect on this consumption Overall, the findings highlight the significant role of globalization in enhancing renewable energy usage in the country.
CO2 emissions, nuclear energy, renewable energy and economic growth in the US
Regions in US from 1960 to 2007
The analysis reveals that nuclear energy consumption has the potential to significantly reduce CO2 emissions However, the current level of renewable energy consumption is insufficient to make a meaningful impact on emissions reduction.
Renewable and non renewable energy consumption growth nexus:
80 countries and spans from 1990 to
In summary, the findings imply that fostering the development of the renewable energy sector has the
Evidence from a panel error correction model
2007 Data for this analysis were sourced from the U.S
Energy Information Administration and World
Bank Development Indicators CD- ROM. potential to alleviate greenhouse gas emissions resulting from non-renewable energy consumption.
Nexus of innovation, foreign direct investment, economic growth and renewable energy: New insights from
The investigation compiled data from the
World Bank spanning the years 1990 to
The paper uses the ARDL and fixed and random panel data methods.
Research highlights a reciprocal relationship among innovation, foreign direct investment (FDI) inflows, economic growth, and the integration of renewable energy sources Empirical evidence emphasizes the crucial role of renewable energy in bridging the gap between economic development and innovative activities.
The results show that economic growth can significantly improve carbon dioxide
Evidence from 30 Provinces in China
The Durbin model highlights that China's economic growth serves as a significant driver of carbon dioxide emissions, yet reducing these emissions is unlikely to substantially impact economic growth There exists a reciprocal relationship between energy consumption and carbon dioxide emissions, indicating that these factors are interconnected Additionally, this relationship exhibits a negative spatial spillover effect, influencing carbon dioxide emissions in neighboring provinces and cities.
After reviewing prior studies, we established hypotheses to investigate the impact of four specific factors on CO2 emissions The following four hypotheses are formulated based on existing research articles.
Hl: Innovation is positively correlated with environmental pollution.
H2: Renewable energy consumption has a negative impact on environmental pollution. H3: FDI has a positive impact on environmental pollution.
H4: Economic growth positively influences carbon dioxide (CO2) emissions
PROPOSED MODELS AND RESEARCH METHODS
Analytical framework
Environmental pollution CO2 emissions World Development
Renewable energy consumption (% of total final energy consumption)
FDl inflows Foreign direct investment, net inflows (BoP, current us$)
Economic growth GDP (current us$) World Development
Indicator( Source : Compiled by authors )
Data collection methods, and data samples
In this research, data was collected from 90 countries aiming for net-zero emissions, as reported by The Energy and Climate Intelligence Unit The authors sourced information from the World Bank Data for the years 2000 to 2022 and used Microsoft Excel to exclude countries with incomplete data This process resulted in a refined dataset of 82 countries, yielding a total of 1,886 observations For data analysis, the authors utilized STATA17 software.
Symbols Variables Indicators Periods Sources
INNO Innovation High- technology exports (current us$)
Renewable energy consumption (% of total final energy consumption)
FDI FDI inflows Foreign direct investment, net inflows (BoP, current us$)
Development Indicator( Source : Compiled by authors )
Variable measurements
CO2 emissions, measured in millions of tons, are a significant indicator of environmental pollution, as highlighted by data from The World Development Indicators, the primary collection of development metrics by the World Bank This study posits that increased CO2 emissions correlate with heightened environmental pollution in Net Zero Target countries While various forms of pollution exist, including waste materials, industrialization, and water contamination, this article focuses solely on CO2 emissions as a representative measure of environmental pollution.
This study analyzes the impact of four independent variables—innovation, renewable energy usage, foreign direct investment (FDI) inflows, and economic growth—on CO2 emissions across 82 countries, utilizing data sourced from World Bank reports.
Research on the relationship between innovation and environmental pollution has produced varying perspectives, with some experts asserting that innovation reduces emissions, while others contend it introduces new challenges (Chatti & Khan, 2024) While certain innovations can increase CO2 emissions due to heightened energy consumption from advanced technologies like powerful computers and appliances, they can also lead to products with significant carbon footprints, such as air travel and SUVs Conversely, many innovations, particularly in renewable energy and energy-efficient appliances, contribute to lowering CO2 emissions This paper measures the impact of innovation on emissions through technology exports, quantified in US dollars.
To effectively reduce CO2 emissions and combat climate change, the consumption of renewable energy is essential (Menyah & Wolde-Rufael, 2010) Unlike fossil fuels that emit greenhouse gases during combustion, renewable energy sources such as solar, wind, and hydropower produce electricity without CO2 emissions This study focuses on the variable of renewable energy consumption, measured in Terawatt-hours, to analyze its impact on emissions reduction.
The relationship between foreign direct investment (FDI) inflows and carbon emissions is a significant global issue Historical data reveals that developing markets experience a positive correlation between FDI and carbon emissions, while established markets show a negative correlation (Kharb et al., 2024) This article measures foreign investment capital flows in the trillions of US dollars.
Economic growth can enhance both consumption demand and capacity, leading to higher carbon dioxide emissions driven by production (Zou and Zhang, 2020) This increase in growth often correlates with a rise in CO2 emissions, which we aim to quantify in this article using a metric of trillion dollars.
Table 3.3 Expectations of Explanatory Variables' Coefficients sign
GDP Economic growth + (Zou & Zhang,
Research models
The variables measured in the study were synthesized and selected by the authors according to previous research articles.
CO2it = a + piFDIit + P2GDPÌÍ + p3INNOit + p4RECit
Where the variables are defined as follows:
CO2: is the dependent variable measuring environmental pollution.
INNO, FDI, GDP REC: arc independent variables
Data processing methods
Descriptive statistics provide a comprehensive overview of a research sample by summarizing key characteristics The main types of measurements in this field include measures of central tendency, such as the mean, and measures of variability, which indicate dispersion To effectively analyze data, it's essential to evaluate dependent, independent, and control variables, allowing researchers to determine the mean, maximum and minimum values, and standard deviation of the dataset.
3.5.2 Pearson correlation coefficient matrix method
The correlation coefficient matrix analysis is utilized to investigate the relationships between the model's variables, helping to identify potential multicollinearity This analysis reveals the degree of correlation among the variables, indicated by the sign of the correlation coefficient, whether positive or negative The findings from this analysis also allow for the evaluation of the model's predictions As the strength of correlation between the variables increases, the likelihood of multicollinearity also escalates.
3.5.3 Testing to choose between Pooled OLS and FEM
Regression models like Random Effects Model (REM), Fixed Effects Model (FEM), and pooled Ordinary Least Squares (OLS) are commonly utilized for analyzing panel data The impact of independent variables on dependent variables is assessed through regression analysis, with the significance of this influence indicated by the p-value derived from the regression outcomes.
3.5.3.1 Pooled ordinary least squares regression (Pooled OLS)
The model maintains that assumptions regarding autocorrelation, heteroscedasticity, and the temporal and spatial differences between observed variables remain unaffected It presumes that the slopes of the coefficients and the y-axis are consistent over time and across variables In this context, only the ordinary least squares (OLS) regression model is utilized, ignoring the time and spatial dimensions of the panel data However, it is important to recognize that varying observation times and characteristics of different sample firms often lead to individual problems in empirical research.
The Pooled OLS regression model is favored for its simplicity and user-friendliness; however, it may overlook the relationships between variables.
Yit = a + (31 Xlit + (32 X2it + + (3k Xkit + pit
XI it, X2it, , Xkit: independent variables
(31, (32, , pk: parameters of the independent variables i: cross-sectional unit t: time-series unit pit: error term
The fixed effects model effectively addresses the limitations of the traditional Pooled OLS linear regression by accounting for the impacts of time series and cross-sectional units While it sacrifices the ability to estimate the effects of time-invariant variables on the dependent variable, it leverages panel data to provide highly reliable estimates of the influence of independent variables on the dependent variable.
Yit = a + p 1X1 it + p2 X2it + p3X3it + + pk Xkit + XID1 i + X2D2Ì + Ầ3D3Ì +••• + XNDNi + vit
XIit, X2it, , Xkit: independent variables pi, P2, P3, , pk: parameters of the independent variables i: cross-sectional unit t: time-series unit
XI, X2, X3, , XN: parameters of entity dummies vit: idiosyncratic error
To determine whether to reject the null hypothesis and select the appropriate model between Pooled OLS and FEM, we employ the p-value from the F-test The null hypothesis can be expressed as HO: XI = X2 = X3 = XN = 0.
We reject the null hypothesis, concluding that unobserved heterogeneity, represented by entity-fixed effects, leads to superior performance of the Fixed Effects Model (FEM) compared to Pooled Ordinary Least Squares (OLS) Conversely, if the p-value of the F-test is below the 5% significance level, the opposite holds true.
3.5.4 Testing to choose between FEM and REM by using Hausman Test
The fixed effects model competes closely with the random effects model, also referred to as the error components model Both approaches utilize distinct intercept terms for each entity, which remain constant over time It is anticipated that the relationships between independent and dependent variables are consistent across both cross-sectional and temporal analyses in both models.
The random effects model (REM) distinguishes itself from the fixed effects model (FEM) by assuming that each cross-sectional unit's intercepts are influenced by a common intercept, which remains consistent across all units and time periods Additionally, a random variable ci is introduced, reflecting the unique differences between each entity's intercept and the overall "global" intercept, while remaining stable over time This framework allows for a nuanced understanding of variations across cross-sectional units in panel data analysis.
Yit = a + pixiit + p2X2it + p3X3it + pkXkit + coit, coit = ci + vit
In the analysis of independent variables, denoted as XI, X2it, , Xkit, we consider parameters p1, p2, p3, , pk that correspond to these variables Each variable is defined across cross-sectional units (i) and time-series units (t) The comprehensive error term is represented as coil, while ei signifies the unobserved time-invariant effect, characterized by a zero mean and constant variance (σ²e), which follows a normal distribution Additionally, vit represents the idiosyncratic error associated with the model.
In order to help researchers choose between a FE model and a RE model, Hausman
(1978) created tests This test is referred to as a model error test, and it determines whether or not the entity fixed effect is uncoiTelated with all independent variables Using assumptions:
HO: appropriateness of the random-effects estimator
Hl: appropriateness of the fixed-effects estimator
If P-value < significance level: reject HO, FEM model is suitable If P-value > significance level: accept hypothesis HO REM model is suitable.
The Variance Inflation Factor (VIF) is a valuable tool for evaluating the correlation among independent variables in a regression model Multicollinearity occurs when these variables are correlated, which poses a challenge since they should ideally remain independent Strong correlations can complicate model fitting and hinder the interpretation of results.
The variance exaggeration factor (VIF) was utilized by Garrett Lane Cohee, Ronald F Piccolo, and Halil Kiymaz (2009) to examine the multicollinearity phenomenon Calculation\formula:
VIF = 1 / (1 - correlation coefficient between variables)
When the correlation coefficient approaches 1, the larger the VIF coefficient, and then multicollinearity occurs.
VIF coefficient < 10: The model has low multicollincarity.
VIF coefficient > 10: The model has high multicollinearity.
Autocorrelation refers to errors that are correlated over time, impacting the effectiveness of the OLS model estimates while maintaining their fairness and consistency with normal distribution Consequently, traditional statistical tests like the t-test and F-test may become unreliable To address this issue, employing more effective estimators such as Pooled GLS, FGLS, logging the variable, or using clustering methods can enhance the model's performance.
The Wooldridge test, Breusch-Pagan-Godfrey test, LM test, Pasaran CD lest, and Durbin-Waston test are autocorrelation tests with the following presumptions:
The hypothesis of the autocorrelation test is as follows:
HO: The model has no autocorrelation
If P-value < significance level: reject HO, the model has no autocorrelation.
If P-value > significance level: accept HO, the model has autocorrelation.
Heteroscedasticity occurs when error variances differ across observations, leading to inaccurate coefficients and t and F statistics in ordinary least squares (OLS) estimation Although the OLS estimates remain unbiased and stable, the presence of heteroscedasticity renders the OLS estimator less reliable To address this issue, more effective estimation methods can be employed, such as Robust estimation, logging the variable, Generalized Least Squares (GLS), Weighted Least Squares (WLS), or Feasible Generalized Least Squares (FGLS).
The White test, Wald test, Breusch-Pagan test, LM test, and other tests for heteroscedasticity are conducted under the following presumptions:
HO: The variance of errors does not vary over the observations.
Hl: the variance of errors varies over the observations.
If P-value < significance level: reject HO, the model has a heteroscedasticity problem.
If P-value > significance level: accept HO, the model is homoscedasticity.
3.5.8 Feasible generalized least squares (FGLS) to deal with heteroscedasticity and autocorrelation
The GLS model is defined by linear parameters, with the error component exhibiting a mean of zero, a nonspherical distribution, a normal distribution, and no correlation with the independent variables.
Yt = a + pl Xlt+ p2 X2t + + pk Xkt+ pt
With var(P)=(XTW-lX)-l, and var(pit)= Oit2
EMPIRICAL RESULTS
Descriptive statistics
Variable Obs Mean Std dev Min Max
Source: Authors compiled from Stata 17.0 software
The study examines the relationship between environmental pollution (CO2) and various economic factors, including foreign direct investment (FDI), economic growth (GDP), innovation (INNO), and renewable energy consumption (REC) Statistical significance is indicated at the 1%, 5%, and 10% levels, denoted by ***, **, and * respectively The model utilized is CO2it = α + PlFDlit + PiGDPu + PsINNOit + P4RECit, highlighting the impact of these variables on environmental outcomes.
Table 4.1 provides a comprehensive overview of the research data sample, comprising 1,886 observations from 82 countries committed to net zero emissions between 2000 and 2022, as reported by the Climate and Energy Intelligence Unit The descriptive statistics reveal an average CO2 emission of 5.79635 million tons per year, with values ranging from 0.0530643 to 28.13866 million tons GDP figures vary significantly, spanning from approximately $13.96 trillion to $25.44 trillion Notably, the Netherlands attracted the highest foreign direct investment, totaling $733.83 billion in 2007 Furthermore, the average renewable energy consumption accounts for 23.49% of total energy use, with Ethiopia leading at 95.55% In terms of innovation, China excelled in 2021, generating substantial revenue from high-quality goods exports.
Pearson correlation coefficient matrix method
Variable CO2 FDI GDP INNO REC
Source: Authors compiled from Stata 17.0 software
The study examines the relationship between environmental pollution (CO2) and various economic factors, including foreign direct investment (FDI), economic growth (GDP), innovation (INNO), and renewable energy consumption (REC) Statistical significance is denoted at the 1%, 5%, and 10% levels, highlighting the importance of these variables in the model CO2it = Ơ + plFDlit + P2GDP1Ị + P j INNO h + /hRECit The findings suggest that FDI inflows, economic growth, innovation, and renewable energy consumption play crucial roles in influencing environmental pollution levels.
The Pearson correlation coefficient matrix method will reveal the correlation between variables for the purpose of eliminating variables that can lead to multicollinearity before running the linear regression model.
Table 4.2 reveals that the model's variables exhibit correlations, with FDI, GDP, and INNO showing positive correlations with CO2, while REC demonstrates a negative correlation with CO2 The Pearson correlation coefficient is significant, as the observed significance levels for all variables are 0, which is below the 5% threshold Additionally, the correlations among the variables are not excessively high, suggesting that the model is unlikely to face multicollinearity issues (Gujarati, 2011).
Multicollinearity test
The authors assessed multicollinearity in their regression analysis using the Variance Inflation Factor (VIF) According to the findings presented in Table 4.3, all VIF coefficients for the model's variables are below 10, with the maximum VIF value recorded at 2.33 This indicates that multicollinearity is not a concern in the regression model (Trong & Ngoc, 2008).
Table 4.3 VIF coefficients of variables in the research model
Source: Authors compiled from Stata 17.0 software
This table presents the findings of a Multicollinearity test using annual data from 2000 to 2022, focusing on four key variables: Economic Growth (GDP), Innovation (INNO), Foreign Direct Investment (FDI) inflows, and Renewable Energy Consumption (REC).
Regression results of OLS, FEM, REM model
Table 4.4 Regression results of OLS, FEM, REM model
Breusch and Pagan LM test 0.0000
Source: Authors compiled from Stata 17.0 software
The study employs a regression analysis to examine the relationship between environmental pollution (CO2) and various factors, including foreign direct investment (FDI), economic growth (GDP), innovation (INNO), and renewable energy consumption (REC) Statistical significance is indicated by asterisks, with ***, **, and * representing significance levels at J %, 5%, and 10%, respectively The results are presented across three models: Column I utilizes Ordinary Least Squares (OLS), Column 2 applies Fixed Effects Model (FEM), and Column 3 employs Random Effects Model (REM).
Table 4.4 displays the regression results for a model analyzing CO2 emissions as the dependent variable, utilizing Pooled OLS, FEM, and REM regressions The findings reveal a positive correlation between Foreign Direct Investment (FDI) and CO2 emissions across all models, while renewable energy consumption exhibits a significant negative correlation with CO2 emissions at the 1% level The F-test comparison between the Pooled OLS and FEM models yields a p-value below the 5% significance threshold, indicating the FEM model's superiority In assessing the FEM versus REM models, the Hausman test returns a p-value of 0.0518, suggesting that the REM model is more appropriate Additionally, heteroscedasticity and autocorrelation tests, using the Breusch-Pagan Lagrangian and Wooldridge tests, produce p-values under 5%, confirming the presence of these issues Consequently, the authors opt to address these limitations through the FGLS method, with the summarized results provided in Table 4.4.
Regression results of FGLS model
Source: Authors compiled from Stata 17.0 software
at the 5% level, and * at the 10% level The regression equation is structured as CO2it = a + β1FDI + β2GDPit + β3ANNOit + β4RECit, highlighting the key variables influencing environmental pollution.
Table 4.5 is the results of the model using the FGLS method In the model, foreign direct investment shows a positive correlation with CO2 emissions at a statistical significance level of 10%.
Experimental findings indicate that foreign direct investment (FDI) flows are statistically significant at the 10% level and positively influence CO2 emissions Therefore, countries should focus on attracting more green FDI to mitigate CO2 emissions and safeguard the environment.
The analysis indicates that GDP is statistically significant at the 5% level, revealing a positive correlation between economic growth and environmental pollution This suggests that, over the long term, economic expansion significantly impacts environmental degradation Therefore, it is crucial for governments to prioritize green growth initiatives and sustainable development goals to mitigate these effects.
Innovation is significantly correlated with CO2 emissions at the 1% level, indicating a positive relationship likely due to the continued use of outdated, inexpensive technologies in many countries To address this issue, policymakers should prioritize the export of environmentally friendly high-tech goods and services By implementing this strategy over the long term, the relationship between CO2 emissions and high-tech exports is expected to become negative, suggesting that an increase in high-tech exports will lead to a reduction in environmental pollution.
Renewable energy consumption is statistically significant at the 1% level, indicating a strong negative relationship with CO2 emissions This data highlights a significant reduction in environmental pollution over time due to increased green energy use It is crucial for individuals and policymakers to prioritize environmental care and contribute to a cleaner future The findings of this article emphasize the need for governments to focus on renewable energy consumption to ensure long-term sustainable development To further reduce per capita CO2 emissions and uphold a clean environment, countries must continue to invest in and expand their renewable energy initiatives.
DISCUSSIONS AND CONCLUSIONS
Conclusions & recommendations
The research conducted by the author's team investigates the interplay between innovation, renewable energy, foreign direct investment (FDI), economic growth, and carbon dioxide emissions Findings reveal that, aside from renewable energy, all other factors positively contribute to CO2 emissions Renewable energy plays a vital role in diversifying energy sources, such as solar, wind, hydropower, and bioenergy, which is essential for reducing environmental degradation and CO2 emissions Additionally, the growth of the renewable energy sector generates new job opportunities and fosters economic growth (Balsalobre-Lorente et al., 2023) FDI enhances technology transfer and provides resources for developing renewable energy infrastructure While traditional economic growth is linked to increased energy consumption and emissions, sustainable development aims to decouple this relationship by improving renewable energy efficiency.
To effectively tackle CO2 emissions, the government must implement supportive policies that promote innovation and the adoption of renewable energy, such as tax incentives and carbon pricing mechanisms Upgrading infrastructure is essential for expanding renewable energy production and transmission, which aligns with sustainable growth goals (Can et al., 2020) Furthermore, enhancing international cooperation and facilitating technology transfer and knowledge sharing among countries are crucial for accelerating the transition to renewable energy.
Limitation & direction of topic development
The research findings align with numerous prior studies, but the authors recognize some limitations Notably, the study's model did not account for the time lag factor, leaving unexamined whether the relationships between the variables are influenced by time delays.
The study's sample area is restricted, and the authors did not conduct geographical or economic segmentation, limiting the understanding of the interplay between CO2 emissions, renewable energy, economic growth, innovation, and FDL Future research should explore the interactions among these variables while accounting for time lag effects Furthermore, implementing geographical or economic segmentation could yield more precise results, facilitating the development of targeted policy implications tailored to various cases and diverse economies.
Afif, R A., Ayed, Y., & Maaitah, o N (2023) Feasibility and optimal sizing analysis of hybrid renewable energy systems: A case study of Al-Karak, Jordan Renewable
Energy, 204, 229-249
Aghasafari, H., Aminizadeh, M., Karbasi, A R., & Calisti, R (2021) CO2 emissions, export and foreign direct investment: Empirical evidence from Middle East and North Africa Region Journal of International Trade & Economic Development, 30(7),
An, Y., Kim, J., Joo, H., Han, G., Kim, H., Lee, w., & Kim, M (2023) Retrofit of renewable energy systems in existing community for positive energy community
Energy Reports, 9, 3733-3744
Andersen, o., Gossling, " s., Simonsen, M., Walnum, H.J., Peeters, p., Neiberger, c.,
2010 CO2 emissions from the transport of China’s exported goods Energy Policy 38
Apergis, N., & Payne, J E (2012) Renewable and non-renewable energy consumption growth nexus: Evidence from a panel error correction model Energy Economics, 34(3),
Balsalobre-Lorente, D., Parente, c c D s., Leitao, N c., & Cantos, J M c (2023). The influence of economic complexity processes and renewable energy on CO2 emissions of BRICS What about industry 4.0? Resources Policy, 82, 103547.
Bergougui, B (2024, January) Moving toward environmental mitigation in Algeria: Asymmetric impact of fossil fuel energy, renewable energy and technological innovation on CO2 emissions Energy Strategy Reviews, 51, 101281
Chatti, w., & Khan, z (2024, February) Towards smart sustainable cities: Does technological innovation mitigate G7 CO2 emissions? Fresh evidence from CS-ARDL.
Science of the Total Environment, 913, 169723.
Chen, c., Pinar, M., & Stengos, T (2022) Renewable energy and CO2 emissions:
New evidence with the panel threshold model Renewable Energy, 194, 117-128.
Hoa, p X., Xuan, V N., & Phuong Thu, N T (2023, September) Nexus of innovation, renewable consumption, FDI, growth and CO2 emissions: The case of Vietnam Journal of Open Innovation: Technology Market, and Complexity, 9(3), 100100
Hoa, p X., Xuan, V N., Thu, N T p., & Huong, L M (2024) Nexus of innovation, foreign direct investment, economic growth and renewable energy: New insights from
Huang, Y., Kuldasheva, z., & Salahodjaev, R (2021) Renewable Energy and CO2
Emissions: Empirical Evidence from Major Energy-Consuming Countries Energies, 14(22), 7504
Kharb, R., Suneja, V., Aggarwal, s., Singh, p., Shahzad, u., Saini, N., & Kumar, D
(2024, February) The Relationship between investment determinants and environmental sustainability: Evidence through mcta-analysis The Quarterly Review of Economics and Finance,
Menyah, K., & Wolde-Rufael, Y (2010, June) CO2 emissions, nuclear energy, renewable energy and economic growth in the US Energy Policy, 38(6), 2911-2915
Ndem B E., James, H T., & Agala, F B (2022) Autoregressive Distributed Lag
Approach (ARDL) to corruption and economic growth nexus in Nigeria Journal of
Zou, s., & Zhang, T (2020, November 30) CO2 Emissions, Energy Consumption, andEconomic Growth Nexus: Evidence from 30 Provinces in China MathematicalProblems in Engineering, 2020, 1-10
sum C02 FDI GDP INNO REC
Variable Obs Mean std dev Min Max
C02 1,701 5.79635 4.964892 0530643 28.13866 FDI 1,836 1.906+10 5.58e+10 -3.30e+ll 7.34e+ll GDP 1,884 7.20e+ll 2.23e+12 1.40e+07 2.54e+13
pwcorr CO2 f-DI GDP INNO REC, sig
CO2 FDI GDP INNO REC
reg CO2 FDI GDP INNO REC
Total ss df MS Number of obs =
- F(4, 958) = 7392.13654 4 1848.03413 Prob > F = 13659.6842 958 14.258543 R-squared = - Adj R-squared = 21051.8207 962 21.8833895 Root MSE =
CO2 Coefficient std err t p>ltl [95% conf interval]
FDI 5.616-12 2.55e-12 2.20 0.028 6.086-13 1.066-11GDP 4.596-13 7.396-14 6.21 A AAA 3.146-13 6.04e-13INNO -7.846-12 2.076-12 -3.79 -1.196-11 -3.79e-12REC -11.63459 6035402 -19.28 -12.819 -10.45018_cons 8.496309 2027206 41.91 A AAA 8.098481 8.894136
xtreg C02 FDI GDP INNO REC, fe
Obs per group: min = 2 avg = 12.7 max = 14
C02 Coefficient std err t p>|t| [95% conf, interval]
7.80e-13 5.29e-13 1.47 0.141 -2.58e-13 1.82e-12 -2.54e-13 5.45e-14 -4.66 0.000 -3.61e-13 -1.47e-13 1.16e-ll 1.64e-12 7.09 0.000 8.406-12 1.48e-ll -15.76712 7056083 -22.35 0.000 -17.15198 -14.38225 _cons 9.664403 1724329 56.05 0.000 9.325977 10.00283 sigma_u sigma_e rho
4.2260978 67101865 97540896 (fraction of variance due to u_i)
xtreg C02 FDI GDP INNO REC, re
Number of obs = 963 Number of groups = 76
Obs per group: min = 2 avg = 12.7 max = 14 corr(u_i, X) = 0 (assumed)
C02 Coefficient std err z p>|z| [95% conf, interval]
8.64e-13 5.31e-13 1.63 0.104 -1.78e-13 1.91e-12 -2.216-13 5.36e-14 -4.13 0.000 -3.26e-13 -1.16e-13 1.08e-ll 1.62e-12 6.70 0.000 7.67e-12 1.406-11 -15.36556 6747198 -22.77 0.000 -16.68799 -14.04314 _cons 9.335635 4812652 19.40 0.000 8.392373 10.2789 sigma_u sigma_e rho
3.9251048 67101865 97160408 (fraction of variance due to u_i)
When analyzing the differenced variance matrix, it's important to note that its rank may not match the number of coefficients being tested If the rank is 1 while testing 4 coefficients, this discrepancy should be anticipated to avoid potential computation issues Additionally, reviewing the output of your estimators for any anomalies is crucial, and you might want to consider scaling your variables to ensure the coefficients are comparable.
(b) (B) (b-B) sqrt(diag(V_b-V_B)) fe re Difference std err.
-2.54e-13 -2.21e-13 -3.25e-14 9.83e-15 1.16e-ll 1.88e-ll 7.84e-13 2.76e-13 -15.76712 -15.36556 -.4015546 2864857 b = Consistent under H8 and Ha; obtained from xtreg.
B = Inconsistent under Ha, efficient under H0; obtained from xtreg.
Test of H8: Difference in coefficients not systematic
Breusch and Pagan Lagrangian multiplier test for random effects
xtserial CO2 FDI GDP INNO REC
Wooldridge test for autocorrelation in panel data
xtgls CO2 FDI GDP INNO REC, corr(arl) panels(h) force
Cross-sectional time-series FGLS regression
Correlation: common AR(1) coefficient for all panels (0.9441)
Number of obs = 963 Number of groups = 76 Obs per group: min = 2 avg = 12.67105 max = 14
C02 Coefficient std err z p>|z| [95% conf interval]
FDI 5.08e-13 2.79e-13 1.82 0.069 -3.89e-14 1.05e-12GDP 1.67e-13 6.76e-14 2.47 0.013 3.45e-14 2.99e-13INNO 3.69e-12 1.44e-12 2.57 0.010 8.73e-13 6.506-12REC -8.370505 3592152 -23.30 -9.074554 -7.666456_cons 6.695957 1567986 42.70 0.000 6.388637 7.003277