Investments, Economics, Renewables and Population Versus Carbon Emissions in ASEAN and Larger Asian Countries: China, India and Pakistan

Abstract

Our research explores the dynamic relationship between CO₂ emissions and four major influencing factors: foreign direct investment (FDI), economic growth (GDP), renewable energy consumption (REN) and population (POP) in the Association of Southeast Asian Nations (ASEAN) and three large Asian countries—China, India and Pakistan, collectively referred to as LACs (larger Asian countries), from 1990 to 2022. The study has three main objectives: (1) to assess the short-run and long-run effects of GDP, FDI, REN and POP on CO₂ emissions; (2) to compare the adjustment speeds and environmental policy responsiveness between ASEAN and LAC regions; and (3) to evaluate the role of renewable energy in mitigating environmental degradation. Against the backdrop of increasing environmental challenges and divergent development paths in Asia, this research contributes to the literature by applying a dynamic heterogeneous panel autoregressive distributed lag (panel ARDL) model. Unlike traditional static panel models, the panel ARDL model captures both long-run equilibrium relationships and short-run adjustments, allowing for country-specific dynamics. The results reveal a significant long-run cointegration among the variables. The error correction term (ECT) indicates a faster adjustment to equilibrium in LACs (−1.18) than ASEAN (−0.37), suggesting LACs respond more swiftly to long-run disequilibria in emissions-related dynamics. This may reflect more responsive policy mechanisms, stronger institutional capacities or more aggressive environmental interventions in LACs. In contrast, the slower adjustment in ASEAN highlights potential structural rigidities or delays in implementing effective policy responses, emphasizing the need for enhanced regulatory frameworks and targeted climate strategies to improve policy intervention efficiency. Results show that GDP and FDI increase emissions in both regions, while REN reduces them. POP is insignificant in ASEAN but increases emissions in LACs. These results provide insights into the relative effectiveness of policy instruments in accelerating the transition to a low-carbon economy, highlighting the need for differentiated strategies that align with each country’s institutional capacity, development stage and energy structure.

1. Introduction

The Association of Southeast Asian Nations (ASEAN) is a regional intergovernmental organization comprising ten countries in Southeast Asia: Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, the Philippines, Singapore, Thailand and Vietnam. ASEAN’s scope is to promote economic growth, regional stability and cultural development among its member states [1]. The dynamics between smaller countries, such as those in ASEAN, and larger Asian countries (LACs) like China, India and Pakistan are intriguing due to a variety of factors. ASEAN has made significant strides in economic integration, creating a single market and production base that enhances intra-regional trade and investment. The member states of ASEAN, such as Singapore, Indonesia, Malaysia and Vietnam, have diverse economic structures ranging from highly developed economies to emerging markets, providing a unique blend of opportunities and challenges [2]. China and India are among the world’s largest economies, contributing significantly to global economic growth. Their large consumer markets and manufacturing bases make them pivotal in global trade. Both China and India continue to show substantial economic growth rates, with China being a global manufacturing hub and India excelling in the services sector, particularly in IT and software services. China and India are major geopolitical players with significant influence in regional and global affairs. China’s Belt and Road Initiative (BRI) and India’s Act East policy are examples of their strategic outreach [3]. The India–China border disputes and the India–Pakistan conflict over Kashmir are sources of regional tension, influencing their foreign policies and defense strategies.

On the other hand, ASEAN countries experience varying levels of social development, with disparities in income, education and healthcare access. Some ASEAN countries face criticism over human rights issues, impacting their international relations and internal social dynamics. China and India are the two most populous countries in the world, leading to unique social challenges such as urbanization, poverty and healthcare. There are significant strides in social development, particularly in education and poverty reduction, in China and India, though challenges remain, such as gender inequality and rural–urban divides [4]. ASEAN countries often navigate between the strategic interests of major powers, notably the U.S. and China, striving to maintain a balance to preserve regional stability and autonomy. Issues like the South China Sea disputes highlight the strategic importance of maritime security and territorial sovereignty for ASEAN nations. China and India have substantial military capabilities and are nuclear powers, influencing regional security dynamics. Pakistan also possesses nuclear weapons, adding to the strategic complexity. These countries invest in defense, with China and India among the top military spenders globally, reflecting their strategic priorities and security concerns [5].

However, over the past few decades, ASEAN countries have experienced significant economic growth, transforming the region into one of the most dynamic parts of the global economy. Trade and investment have been key drivers of this growth. ASEAN has established itself as a critical hub for global trade, benefiting from strategic trade agreements and foreign direct investment [6]. The ASEAN Free Trade Area (AFTA) has enhanced trade between member countries by reducing tariffs and other trade barriers. Industrialization has also played a crucial role. Countries like Indonesia, Malaysia, Thailand and Vietnam have witnessed rapid industrialization, moving from agriculture-based economies to manufacturing and services. This shift has significantly boosted GDP growth rates. The tourism industry is another vital contributor to the economy in many ASEAN countries. Thailand, Indonesia and Malaysia are notable for their tourism sectors, attracting millions of visitors annually. Additionally, the region is embracing the digital economy, with a burgeoning tech sector in countries like Singapore, Indonesia and Vietnam. E-commerce, fintech and digital services are expanding rapidly, driving economic diversification [7].

As ASEAN countries continue to grow economically, there is an increasing focus on sustainable development and the integration of renewable energy sources (RESs). Many ASEAN countries have established national policies and targets to increase the share of RESs in their energy mix [8]. For example, Thailand’s Power Development Plan aims to have 30% of its energy from RESs by 2036, while Vietnam has set ambitious targets for solar and wind energy. The region has seen a surge in solar and wind power projects, with Vietnam becoming a regional leader in solar power capacity, driven by favorable government policies and incentives. Hydropower is a significant renewable energy source in ASEAN, especially in Laos and Myanmar. Laos, dubbed the “Battery of Southeast Asia”, exports a substantial amount of its hydropower to neighboring countries [9]. ASEAN countries are also collaborating on RESs through the ASEAN Plan of Action for Energy Cooperation (APAEC), which aims to enhance energy security and connectivity across the region, promoting the use of renewable energy sources.

Despite strides in RESs, ASEAN countries face challenges in managing CO₂ emissions due to ongoing industrialization and urbanization. Many ASEAN countries still rely heavily on coal, oil and natural gas for energy. Indonesia, Malaysia and Thailand, for instance, have substantial fossil fuel reserves, which continue to be significant energy sources [10]. Urbanization in ASEAN countries has led to increased energy demand and higher CO₂ emissions. Cities like Jakarta, Bangkok and Manila are expanding rapidly, contributing to rising emissions from transportation, industry and construction. Deforestation for agriculture and development in countries like Indonesia and Malaysia has contributed significantly to CO₂ emissions. Forests act as carbon sinks, and their loss exacerbates climate change. ASEAN countries are committed to reducing CO₂ emissions through various international agreements, such as the Paris Agreement. They are working on implementing measures to curb emissions, improve energy efficiency and enhance carbon capture and storage technologies [11].

China, India and Pakistan, three major countries in Asia, have experienced significant economic transformations over the past few decades. Each of these countries faces unique challenges and opportunities regarding economic growth, RES integration and CO₂ emissions [12]. China has emerged as the world’s second-largest economy, driven by a combination of industrialization, urbanization and trade. The country’s rapid economic growth is characterized by a shift from an agrarian economy to a manufacturing and service-oriented one. Massive infrastructural projects, foreign investments and an export-driven economy have been critical in this transformation. China’s BRI further underscores its role in global trade and investment [13]. India, with its diverse economy, is one of the fastest-growing major economies in the world. Economic reforms, technological advancements and a young workforce have fueled growth. The IT and service sectors, along with industrial development, play a significant role in India. Urbanization and the rise of the middle class have also contributed to increased domestic consumption and economic expansion. Pakistan’s economy, while smaller in comparison to China and India, has shown resilience and growth potential. Agriculture remains a significant sector, but industrialization and services are becoming increasingly important. Investments in infrastructure, such as the China–Pakistan Economic Corridor (CPEC), aim to boost economic growth and connectivity within the region.

In terms of RES integration, China has made substantial progress. It is the world’s largest producer of solar and wind energy. Government policies and investments have driven the growth of RES, with ambitious targets for reducing carbon intensity and increasing the share of RESs in the energy mix. China aims to achieve carbon neutrality by 2060, highlighting its commitment to sustainable development. India is also making strides in RES. The country has set ambitious targets to increase its RES capacity, particularly in solar and wind power. The National Solar Mission and various state initiatives aim to reduce dependence on fossil fuels and enhance energy security. India is committed to achieving 175 GW of RES capacity by 2022 and aims to reach 450 GW by 2030. Pakistan, while still reliant on fossil fuels, is gradually integrating RESs into its energy mix [14]. The government has set targets to increase the share of RESs, particularly through wind, solar and hydropower projects. By 2025, ASEAN aims to increase the share of renewable energy in its energy mix to 23%, with the ambition of further increasing installed renewable capacity to 35% by 2035 [15].

Despite progress in RESs, managing CO₂ emissions remains a significant challenge for all three countries [16]. China, the world’s largest emitter of CO₂, faces the daunting task of balancing economic growth with environmental sustainability. Industrial activities, coal dependence and urbanization contribute to high emission levels. However, China is implementing measures to reduce emissions, such as investing in clean energy technologies, enhancing energy efficiency and promoting electric vehicles. India is the third-largest emitter of CO₂ globally. Rapid urbanization, industrialization and a heavy reliance on coal for electricity generation are major contributors to its emissions. India is committed to reducing its carbon footprint through various initiatives, including increasing RES capacity and improving energy efficiency. Pakistan, while contributing a smaller share of global CO₂ emissions, faces its own challenges related to energy consumption and industrial activities. The reliance on fossil fuels for energy, combined with deforestation and agricultural practices, contributes to its emissions. Pakistan is working on measures to mitigate these impacts, including reforestation projects, enhancing RES capacity and improving energy efficiency. China, India, Pakistan and ASEAN are at critical junctures where they must balance economic growth, RES integration and CO₂ emissions management [17].

In this paper, the three LACs are compared with ASEAN countries on their CO₂, FDI, RES consumption (REN), economic growth (GDP) and population (POP). Even if several studies have analyzed CO₂ emissions within individual Asian countries or regional blocs, there is a notable lack of comparative research between ASEAN and LACs, leading to an insufficient understanding of how environmental drivers differ across diverse economic and institutional contexts.

The existing literature also tends to overlook country-specific heterogeneity, treating cross-national dynamics as homogeneous and thereby masking critical variations in policy effectiveness. In light of these gaps, our research contributes by (1) applying the panel ARDL model to capture both short- and long-run relationships in a dynamic framework; (2) conducting a regionally comparative analysis between ASEAN and LAC economies; (3) assessing the differential roles of FDI, GDP, renewable energy and population in influencing emissions; and (4) providing policy-relevant insights into the relative effectiveness of environmental interventions across diverse national contexts.

To address the identified gaps in the literature and guide the empirical analysis, this research seeks to answer the following questions:

RQ1. What are the short-run and long-run effects of FDI, economic growth (GDP), REN and POP on CO₂ emissions in ASEAN and LACs?

RQ2. How do the dynamics of CO₂ emissions differ between ASEAN and LAC economies in terms of policy responsiveness and adjustment speed?

RQ3. To what extent does renewable energy mitigate environmental degradation in rapidly developing economies?

The remainder of this paper is structured as follows. Section 2 presents a comprehensive review of the existing literature, focusing on regional characteristics and methodological limitations. Section 3 introduces the theoretical framework underpinning the relationships between economic development, renewable energy, FDI, population and CO₂ emissions. Section 4 details the econometric methodology. Section 5 discusses the empirical results, while Section 6 provides an in-depth discussion of the findings in light of the existing literature. Finally, Section 7 concludes with policy recommendations and outlines directions for future research.

2. Literature Review

2.1. Regional Characteristics

In this section, previous research papers that focused on ASEAN and its neighbors from the CO₂ emissions point of view are investigated. Researchers have aimed to examine the increase in energy-related CO₂ emissions within ASEAN, focusing on trends and drivers [18]. They have uncovered that while major emitters in the region have seen a slowdown in CO₂ emissions growth due to improved energy efficiency and a gradual shift towards lower-emission energy sources (like gas and RESs), these changes are insufficient for a substantial transformation. Most ASEAN countries still exhibited steady emissions growth over the studied period (up to 2016). Another research study explored the influence of FDI and transport-sector energy consumption on CO₂ emissions in five ASEAN countries (1980–2019), using the Environmental Kuznets Curve (EKC) and a nonlinear ARDL (NARDL) model [19]. The findings indicated a long-run equilibrium cointegrated relationship between carbon emissions and their determinants. Specifically, the EKC hypothesis was confirmed only for Singapore, while in Indonesia, Malaysia, the Philippines and Thailand, income growth correlates with increased CO₂ emissions. Transport energy consumption had a greater impact on CO₂ emissions than FDI, except in Singapore. The study recommended moderate growth policies, promoting clean FDI inflows and focusing on energy-efficient transport systems. Moreover, the researchers investigated the impact of population, GDP and energy consumption on CO₂ emissions in eight ASEAN countries from 1990 to 2017, using the Kaya identity and the Logarithmic Mean Division Index (LMDI) [20]. The results showed that population and economic activity significantly contribute to CO₂ emissions, while energy intensity reduces emissions. They pointed out that the effects of energy intensity and carbon intensity varied across income levels.

Another research study examined how hydropower consumption, FDI and manufacturing performance affected CO₂ emissions in four ASEAN countries over 1980–2015 using the ARDL bound test [21]. Results showed that hydropower consumption significantly reduced CO₂ emissions only in Malaysia. In contrast, manufacturing performance increased emissions in all ASEAN-4 countries, and FDI inflows further raised emissions in Malaysia and the Philippines. The study recommended enhancing hydropower use as a clean energy source to mitigate CO₂ emissions. In [22], the researchers investigated the effects of economic growth, globalization and financial development (FD) on CO₂ emissions in ASEAN countries using data from 2004 to 2018. Employing a panel fixed-effects model with Driscoll–Kraay standard errors, the findings revealed positive relationships between economic growth, globalization, FDI and CO₂ emissions. The study suggested more energy-efficient measures to manage the environmental impact of economic activities. The impact of transport-sector energy consumption and FDI on CO₂ emissions in ASEAN-5 countries, utilizing cointegration and Granger causality methods, was further analyzed [23]. The findings revealed that CO₂ emissions and their determinants are cointegrated in Indonesia, Malaysia and Thailand. Economic growth and transport energy consumption significantly influenced CO₂ emissions, while FDI was less impactful. The study suggested focusing on energy-efficient transportation policies to reduce emissions without hindering economic growth. In [24], the effects of RESs and fossil fuel energy consumption, natural resources and agriculture and political constraints on CO₂ emissions in Cambodia, Malaysia, Indonesia and Thailand were studied from 1990 to 2019, using common-correlated effect mean group and augmented mean group estimates. Results indicated that RESs reduce CO₂ emissions. This study also recommended promoting RESs and eco-friendly policies to improve environmental quality.

The relationship between sector-specific FDI and CO₂ emissions in five ASEAN countries from 1980 to 2018 was further investigated [25], employing panel Granger causality tests. Results indicated that FDI in polluting industries increased CO₂ emissions, while FDI in other sectors did not have a significant impact. The findings suggested regulatory measures to manage FDI in polluting sectors to mitigate environmental impacts. Another research study examined the determinants of CO₂ emissions in ASEAN+3 countries (including China, Japan and South Korea) from 1991 to 2010, using panel unit root tests, cointegration tests and Granger causality based on the Vector Error Correction Model (VECM) [26]. The results indicated that economic growth, energy consumption and trade openness were significant determinants of CO₂ emissions in the region. Additionally, the relationship between CO₂ emissions, economic growth, industrialization, population growth and RESs was analyzed in ASEAN-5 countries (Indonesia, Malaysia, Thailand, Philippines and Singapore) using the VECM [27]. Findings showed that population growth and RESs significantly affected CO₂ emissions in the long term, while industrialization and RESs had significant short-term effects. In [28], the relationship between CO₂ emissions, energy consumption and GDP in ASEAN-5 countries over 1980–2016 was studied, accounting for cross-sectional dependence. The study found heterogeneity in the causal relationships and supported the EKC hypothesis. Another research study investigated the dynamic relationship between energy consumption, CO₂ emissions and economic output in ASEAN from 1971 to 2015, using cointegration and causality models [29]. The results suggested a long-run relationship among these variables, indicating that policies aimed at reducing energy consumption can help lower CO₂ emissions without significantly affecting economic growth. The effects of population, GDP, oil consumption and FDI on CO₂ emissions in ASEAN-5 countries from 1985 to 2017 were examined, using a fixed-effects model [30]. Results showed that population, GDP and oil consumption significantly increased CO₂ emissions, while FDI had a negative effect, supporting the halo effect hypothesis. Furthermore, the impact of financing sources on CO₂ emissions in six ASEAN countries from 1986 to 2018 was studied, using pooled mean group estimation and dynamic fixed effects [31]. The results confirmed a long-run relationship among the variables, with government expenditure and FDI increasing carbon emissions, while foreign aid reduced emissions in both the short and long terms. A comparative analysis based on the previous references is provided in

Table 1. Comparative analysis based on previous research.

Ref.	Objective	Countries	Methods	Years	Variables	Main Findings
[18]	Analyze the growth of energy-related CO2 emissions in ASEAN	ASEAN countries	Decomposition analysis	1971–2016	CO2 emissions, energy efficiency, energy fuel mix	Growth of CO2 emissions slowed in major emitters due to energy efficiency and fuel mix changes, but not enough to counteract rising emissions overall.
[19]	Assess the impact of FDI and energy consumption in the transport sector on CO2 emissions	Indonesia, Malaysia, Philippines, Singapore, Thailand	NARDL, Environmental Kuznets Curve (EKC)	1980–2019	CO2 emissions, FDI, transport sector energy consumption, income	EKC holds only for Singapore; FDI and transport energy impact CO2 emissions, with transport energy being more impactful.
[20]	Examine the impact of population, GDP, energy intensity and carbon intensity on CO2 emissions	ASEAN countries	Logarithmic Mean Division Index (LMDI)	1990–2017	CO2 emissions, population, GDP, energy intensity, carbon intensity	Population and economic activity increase emissions in most countries; energy intensity reduces emissions in lower-middle-income countries but increases in higher-income ones.
[21]	Examine the impacts of hydropower consumption, FDI and manufacturing performance on CO2 emissions	ASEAN-4 countries	Autoregressive distributive lag bound test	1980–2015	CO2 emissions, hydropower consumption, FDI, manufacturing performance	Hydropower consumption negatively impacts CO2 emissions only in Malaysia; manufacturing impacts emissions in all countries; FDI impacts emissions in Malaysia and the Philippines.
[22]	Investigate the impact of economic growth, globalization and financial development on CO2 emissions	ASEAN countries	Fixed-effects model, Discroll–Kraay standard error	2004–2018	CO2 emissions, economic growth, globalization, financial development	Economic growth, globalization and FDI positively impact CO2 emissions; recommended policies for efficient energy use to control emissions.
[23]	Assess the impact of transport sector’s energy consumption and FDI on CO2 emissions	ASEAN-5 countries	Cointegration, Granger causality	1971–2008	CO2 emissions, transport energy consumption, FDI, income	EKC not applicable to ASEAN-5; bi-directional causality between GDP and CO2 emissions in Indonesia and Thailand; transport energy and FDI impact emissions in Malaysia and Thailand.
[24]	Examine the effects of energy, natural resources, agriculture, political constraint and regional integration on CO2 emissions	Cambodia, Malaysia, Indonesia, Thailand	CCEMG, AMG	1990–2019	CO2 emissions, renewable energy, fossil fuel energy, natural resources, agriculture, political constraint	RES reduces CO2 emissions, fossil fuels increase them; agriculture impacts negatively, political constraint induces emissions, regional integration impact is not significant.
[25]	Examine the sector-specific FDI and CO2 emissions	ASEAN countries	Panel Granger causality tests	1980–2018	CO2 emissions, sector-specific FDI	FDI in polluting industries increases CO2 emissions; no robust evidence that FDI in other sectors impacts emissions.
[26]	Investigate the determinants of CO2 emissions in ASEAN+3 countries	ASEAN+3 countries	Panel unit root test, cointegration test, VECM	1991–2010	CO2 emissions, energy consumption, GDP, urbanization, trade openness, transportation	Economic growth, energy consumption and trade openness are determinants of CO2 emissions.
[27]	Examine the dynamic relationship between CO2 emissions and GDP, industrialization, population growth and RES	ASEAN-5 countries	VECM	2007–2016	CO2 emissions, economic growth, industrialization, population growth, renewable energy	Long-term: population growth and RESs significantly affect CO2 emissions; short-term: industrialization and RESs affect emissions.
[28]	Re-examine the relationship between CO2 emissions, energy consumption and economic growth	ASEAN-5 countries	Panel test of Granger non-causality	1980–2016	CO2 emissions, energy consumption (EC), economic growth	Unidirectional causality from GDP to CO2 for Malaysia, Philippines, Singapore, Thailand; GDP to EC in Indonesia, Malaysia, Thailand; EC to GDP in Singapore; bi-directional causality in the Philippines; EKC hypothesis supported.
[29]	Examine the dynamic relationship between energy consumption, CO2 emissions and economic output	ASEAN countries	Cointegration and causality models	1971–2015	CO2 emissions, energy consumption, economic output	Long-run relationship and causality between energy consumption, economic output and CO2 emissions; policies to reduce energy consumption can reduce CO2 emissions without much impact on GDP.
[30]	Examine the effect of population, GDP, oil consumption and FDI on CO2 emissions	ASEAN-5 countries	Fixed-effects model	1985–2017	CO2 emissions, population, GDP, oil consumption, FDI	Population, GDP and oil consumption positively affect CO2 emissions; FDI negatively affects emissions.
[31]	Investigate the impact of financing sources on carbon emissions	Indonesia, Laos, Malaysia, Philippines, Thailand, Vietnam	Pooled mean group estimation, dynamic fixed effects	1986–2018	CO2 emissions, domestic credit, government expenditure, FDI, foreign aid	Long-run relationship among variables; government expenditure and FDI increase emissions, foreign aid reduces emissions in both short and long run.

.

This table summarizes the objectives, countries analyzed, methods used, years of analysis, variables and main findings of previous studies.

2.2. Methodological Limitations in the Literature

Even if the existing literature offers valuable empirical insights into the drivers of CO₂ emissions, several methodological limitations persist. First, many prior studies rely on static panel data models that overlook dynamic interdependencies and adjustment processes, which are essential in understanding the long-run versus short-run impacts of variables like GDP, FDI, REN and POP.

Second, some models fail to adequately address the order of integration among variables or ignore cross-sectional heterogeneity and dependence, potentially leading to biased estimations and misleading policy recommendations. Comparative studies across regions such as ASEAN and LACs are rare, despite their contrasting institutional frameworks, development stages and energy structures.

Existing applications of ARDL models often struggle with scalability in high-dimensional panel settings, potentially leading to biased estimations if CD or parameter heterogeneity is not properly accounted for.

Lastly, several papers include broad macro indicators without accounting for lag structures, adjustment speeds or asymmetric behavior in environmental responses. These gaps limit the explanatory power and practical applicability of empirical findings for policymakers aiming to design region-specific sustainability strategies.

2.3. Present Research Advancements

The current research addresses these gaps through both conceptual and methodological innovations. First, it employs the panel ARDL approach, which accommodates variables integrated of order I(0) and I(1) and allows for the simultaneous estimation of short-run dynamics and long-run relationships. This framework is suitable for macroeconomic panel data with mixed integration orders and heterogeneous country structures.

Second, this study explicitly compares two regional blocks, ASEAN and LACs, that have not been thoroughly examined in tandem within the CO₂ emissions literature. The comparative design provides novel insights into regional differences in adjustment speeds and environmental policy responsiveness, especially by means of FDI elasticity and REN effectiveness.

Third, the model includes lagged structures and error correction mechanisms, offering a more dynamic and policy-relevant interpretation of environmental behavior. The findings are interpreted through multiple theoretical approaches, such as the pollution haven/halo hypotheses and decoupling theory, thereby enriching the conceptual grounding of the analysis.

3. Theoretical Framework

This research draws on several well-established theoretical frameworks to contextualize the relationships between CO₂ emissions and their socioeconomic drivers. First, the Environmental Kuznets Curve (EKC) hypothesis suggests an inverted U-shaped relationship between environmental degradation and income per capita. According to the EKC hypothesis, emissions initially rise with economic growth but eventually decline after surpassing a certain income threshold due to structural transformation and environmental regulations. This concept is relevant for interpreting the nonlinear impacts of GDP on emissions in both ASEAN and LAC countries.

Second, it considers the opposing theories of a pollution haven and pollution halo effects in relation to FDI. The pollution haven hypothesis posits that multinational corporations may relocate pollution-intensive activities to countries with lax environmental regulations, increasing emissions. In contrast, the pollution halo effect argues that FDI can introduce cleaner technologies and higher environmental standards, potentially reducing emissions. These perspectives provide an interpretive context for the sign and significance of FDI coefficients in our model.

Third, the notion of decoupling economic growth from environmental impact underpins our analysis of renewable energy consumption. This framework posits that economies can grow while reducing emissions by shifting to cleaner energy sources and enhancing efficiency. Accordingly, the role of REN in mitigating emissions is assessed through this conceptual lens.

By integrating these theoretical perspectives, our research moves beyond descriptive correlations and provides a structured interpretation of the mechanisms behind the observed relationships.

4. Methodology

The dataset includes observations of FDI, GDP, REN, POP and CO₂ emissions from 1990 to 2022, collected from World Bank datasets. This study examines and compares ten ASEAN countries alongside three major Asian nations (China, India and Pakistan) to analyze the relationships between CO₂ emissions, foreign investment, economic growth, population and RES usage. The methodology of this paper consists of several data processing steps, as in Figure 1. We applied EVIEWS 12 for the econometric analyses.

4.1. Data Preprocessing

The dataset comprises annual macroeconomic indicators sourced from the World Bank and Our World in Data, which do not require seasonal adjustment. No significant outliers were identified based on visual inspection and preliminary diagnostics. Missing values in the dataset were addressed using linear interpolation to maintain data continuity across time. Furthermore, to address a potential scale effect and stabilize variance, all variables were transformed into natural logarithms.

The empirical analysis is based on an unbalanced panel comprising 13 countries (10 ASEAN and 3 LACs) over the period of 1990–2022. For the ASEAN group, the dataset includes 330 observations per variable. For the LAC countries, there are 99 observations per variable.

4.2. Cross-Sectional Dependence

Cross-sectional dependence (CD) is a concept in econometrics and statistical analysis that refers to the presence of correlation among cross-sectional units in a panel dataset. This can occur when the variables of different entities are influenced by common factors or exhibit mutual dependencies. Cross-sectional units might be affected by common global shocks or trends, such as economic crises, technological changes or policy shifts. Geographical proximity may lead to dependencies due to similar environmental conditions, regional policies or trade relations. Entities connected through trade, finance or social networks might influence each other, creating dependencies in their behaviors or outcomes. Ignoring cross-sectional dependence can lead to biased and inefficient estimates, incorrect standard errors and invalid hypothesis tests in panel data models. Tests like the Breusch–Pagan LM test [32], Pesaran’s CD test and the cross-sectional dependence test by Pesaran (2004) are commonly used to detect CD [33]. Originally developed to test for heteroskedasticity, the Breusch–Pagan LM test has been adapted to check for CD in panel data. Pesaran’s CD test computes pairwise correlation coefficients between panel cross-sections to determine if the average correlation is significantly different from zero. Pesaran’s scaled LM test, formally known as the scaled LM test for error CD in panel models, is widely used for detecting CD in large panel datasets [33]. It is particularly relevant when the number of units (N) is large, often exceeding the number of periods (T).

Data stationarity is tested using first-generation panel unit root tests (PURTs), which assume no CD between units in the panel, meaning each unit is independently affected by its unique shocks or characteristics. Notable tests in this category include the Levin, Lin and Chu (LLC) test [34] and the ADF-Fisher Chi-square (ADF) test [35].

4.3. First-Generation PURT

Next, we briefly examine the theoretical underpinnings of two first-generation PURTs which we apply.

4.3.1. LLC PURT

The LLC test assumes homogeneity in the autoregressive coefficients across cross-sections. It assumes that the coefficient of the lagged dependent variable is the same across all units in the panel. The model includes individual fixed effects to account for cross-sectional heterogeneity. Typically, the test does not include a time trend [36]. Considering a variable observed across N cross-sections and T periods, we specify a model with individual effects but no time trend:

$${∆y}_{it}={\alpha }_i+{\rho }_i{y}_{it-1}+{\sum }_{j=1}^k{\beta }_{ij}{∆y}_{it-\mathrm{j}}+{\epsilon }_{it}$$

(1)

where Δ is the first difference, $y_{it}$ is the variable of interest for unit i at time t and ${\epsilon }_{it}$ is the error term.

The LLC test operates under the null hypothesis that each individual time series in the panel contains a unit root (i.e., is non-stationary). The alternative hypothesis is that all individual series are stationary.

4.3.2. Fisher-Type PURT

The Fisher-type PURT is utilized to determine the presence of unit roots in panel data. Originally developed by Dickey and Fuller (1979) [37], these tests were adapted for panel data by Maddala and Wu (1999) [35] and Choi (2001) [38]. The individual ADF tests for each cross-section i in the panel are specified as in Equation (1). The p-values are obtained from each of these individual tests. The p-values are combined using Fisher’s method:

$${\chi }^2=-2{\sum }_{i=1}^N\mathrm{l}\mathrm{n}\left(p_i\right)$$

(2)

where N is the number of cross-sections and $p_i$ is the p-value from the individual ADF test from cross-section i. This statistic follows a chi-square distribution with 2N degrees of freedom when y tends to infinity and N is fixed.

The analysis also incorporates the second-generation unit root test to account for CD among countries, namely the Cross-sectional Augmented IPS test (CIPS). The CIPS test improves panel data analysis by including cross-sectional averages of the dependent variable and its lag(s), capturing unobserved common factors (e.g., global shocks or regional spillovers). This makes it particularly well-suited for macroeconomic panels where countries are interdependent.

4.4. Panel ARDL

The panel autoregressive distributed lag (ARDL) model [39] is a dynamic panel data approach that combines both short-term and long-term analyses within a single framework. This model is particularly useful for examining the relationship between variables over time and across different cross-sectional units. The panel ARDL methodology has a theoretical prerequisite that all variables must be integrated of order zero or one (I(0) or I(1)), but not of order two (I(2)), to ensure the validity of long-run cointegration estimation. It includes error correction terms to capture long-run equilibrium relationships, allowing for efficient estimation of both short-term dynamics and long-term relationships. This approach is widely used to investigate causal linkages and policy impacts. The ARDL (p, q) equation is

$$y_{it}={\sum }_{k=1}^p{\varphi }_{ik}{y}_{it-k}+{\sum }_{j=0}^q{\delta }_{ij}{x}_{it-j}+{\alpha }_i+{ϵ}_{it}$$

(3)

In Equation (3), $y_{it}$ is the dependent variable for unit i at time t, $x_{it-j}$ represents the independent variables for unit i at time t − j, ${\epsilon }_{it}$ is the error term, ${\alpha }_i$ is the individual-specific effect and p and q are lag orders. In the context of the ARDL model, the error correction term (ECT) represents the speed at which the dependent variable returns to equilibrium after a change in the independent variables. The ECT is derived from the long-term relationship and is included in the short-term dynamics of the model. The equation incorporating the ECT is typically written as

$$\mathsf{\Delta }{y}_{it}={\sum }_{k=1}^p{\varphi }_{ik}{\mathsf{\Delta }y}_{it-k}+{\sum }_{j=0}^q{\delta }_{ij}\mathsf{\Delta }{x}_{it-j}+{\alpha }_i+\mathsf{\lambda }{ECT}_{it-1}+{ϵ}_{it}$$

(4)

${ECT}_{it-1}$ is the lagged ECT, representing the deviation from the long-term equilibrium in the previous period. λ is the coefficient of the ECT, indicating the speed of adjustment back to equilibrium (it should be between −1 and 0 for stability). The ECT captures the extent to which the dependent variable adjusts to restore equilibrium following a shock. If λ is significantly negative, it suggests a meaningful error correction mechanism in place.

While the panel ARDL framework accounts for dynamic adjustments and cointegration among variables, it does not explicitly control for potential endogeneity arising from reverse causality—such as the possibility that higher CO₂ emissions may influence the level or sectoral composition of FDI, rather than the other way around. Given the bidirectional relationships documented in the literature between FDI, GDP and environmental indicators, this is a relevant concern. To partly address this issue, we complemented our analysis with Dumitrescu–Hurlin panel causality tests, which allow us to infer directional causality across heterogeneous panel units. However, future extensions could benefit from using instrumental variable techniques or system Generalized Method of Moments (GMM) approaches to more robustly correct for endogeneity bias and isolate the causal impact of FDI and other explanatory variables on CO₂ emissions.

4.5. Dumitrescu–Hurlin Panel Causality Test

The Dumitrescu–Hurlin (2012) (D-H) panel causality test is used to investigate the causal relationships between variables in a panel data context [40]. This test is particularly suited for heterogeneous panels, where the causality relationship might differ across cross-sectional units. The null hypothesis states that there is no Granger causality for all panel units, while the alternative hypothesis allows for Granger causality in at least one panel unit. For each cross-sectional unit, a pair of equations is estimated to test for Granger causality between the variables. The equations of the model are

$$y_{it}={\alpha }_i+{\sum }_{k=1}^p{\beta }_k{y}_{it-k}+{\sum }_{k=1}^q{\gamma }_k{x}_{it-k}+{ϵ}_{it}$$

(5)

$$x_{it}={\alpha }_i+{\sum }_{k=1}^p{\delta }_k{x}_{it-k}+{\sum }_{k=1}^q{\eta }_k{y}_{it-k}+{\upsilon }_{it}$$

(6)

For each cross-sectional unit, Wald statistics are calculated for the coefficients of the lagged independent variable. These individual test statistics are then averaged across all cross-sectional units. The combined test statistic is compared to a critical value from the standard normal distribution to determine whether to reject the null hypothesis of no causality.

The D-H panel causality test was selected due to its ability to account for heterogeneity across cross-sectional units, which is especially relevant given the structural, institutional and developmental diversity among ASEAN and LAC countries. Unlike traditional Granger causality tests that assume slope homogeneity, the D-H test allows each country to exhibit its own causal dynamics while still enabling inference at the panel level. This makes it well-suited for our analysis of heterogeneous economies.

5. Results

5.1. Input Data Analysis

The input data comprises observations of FDI, GDP, REN, POP and CO₂ emissions from 1990 to 2022, sourced from World Bank datasets. Missing values in the dataset were addressed using linear interpolation to maintain data continuity across time. This study investigates and compares the ten ASEAN countries and three larger Asian countries (China, India and Pakistan) to analyze the relationship between CO₂ emissions and foreign investment, economic growth, population and renewable energy usage. To understand the correlations and trends, scatter plots showing the relationships between CO₂ emissions and the four indicators for ASEAN countries are shown in Figure 2.

The relationship between CO₂ emissions and FDI appears varied among different ASEAN countries. Some countries show a positive correlation, where higher FDI is associated with higher CO₂ emissions. This could suggest that increased foreign investment is leading to industrial activities that contribute to CO₂ emissions. However, other countries show little to no correlation, indicating that the nature of foreign investments might be more environmentally friendly or that other factors are influencing CO₂ emissions more significantly. There is a general trend where higher GDP is associated with higher CO₂ emissions for most ASEAN countries. This is a common observation globally, as economic growth often leads to increased industrial activities, energy consumption and consequently higher CO₂ emissions. However, the trend lines for some countries might show variations, indicating different levels of efficiency in energy use or different stages of economic development and industrialization. The relationship between CO₂ emissions and the share of renewable energy in the energy mix shows a mixed pattern. Some countries might exhibit a negative correlation, where an increase in renewable energy usage leads to lower CO₂ emissions. This would indicate successful efforts in reducing the carbon footprint through renewable energy. Other countries might show a less clear relationship, suggesting that despite an increase in renewable energy, CO₂ emissions remain a high possibility due to continued reliance on fossil fuels or overall energy demand growth. Generally, there is a positive correlation between population size and CO₂ emissions. Larger populations tend to have higher energy demands, which can lead to increased CO₂ emissions, particularly if the energy sources are not environmentally friendly. However, the degree of correlation varies among countries, reflecting differences in energy efficiency, industrialization and policies aimed at reducing emissions.

Higher GDP levels often correlate with higher CO₂ emissions, highlighting the environmental cost of economic development. Countries with higher shares of renewable energy tend to have varying success in reducing CO₂ emissions, indicating that transitioning to renewable energy is complex and multifaceted. Larger populations generally contribute to higher CO₂ emissions, emphasizing the need for sustainable development practices in densely populated countries. These interpretations suggest that while economic growth and population increase are key drivers of CO₂ emissions, the adoption of RESs and energy efficiency measures can mitigate this impact.

Brunei shows a relatively stable trend with low levels of FDI and CO₂ emissions. This indicates that foreign investments in Brunei have not led to significant increases in CO₂ emissions. Brunei shows a high GDP with moderate CO₂ emissions, suggesting that its economic activities are not heavily carbon-intensive, possibly due to its focus on oil and gas industries, which are relatively less diversified. Data for REN is not available for Brunei, so no interpretation can be made here. The trend shows low CO₂ emissions despite a growing population, indicating efficient energy use and possibly strict environmental regulations.

Cambodia shows an increasing trend of CO₂ emissions with higher FDI, indicating that foreign investments are contributing to carbon-intensive industrial activities. There is a positive correlation between GDP and CO₂ emissions, suggesting economic growth is associated with increased CO₂ emissions, likely due to industrialization and energy use. Cambodia’s RES share is low, and its CO₂ emissions are high, indicating a heavy reliance on fossil fuels. An increasing population correlates with rising CO₂ emissions, suggesting that higher energy demands are not being met with clean energy sources.

Indonesia shows a varied relationship, indicating that the type of FDI might differ, with some investments being more carbon-intensive than others. A strong positive correlation suggests that economic growth in Indonesia is leading to higher CO₂ emissions, likely due to its large industrial sector and energy consumption. Despite increasing renewable energy capacity, CO₂ emissions remain high, suggesting that renewable energy is not yet sufficient to offset the carbon emissions from other sources. As one of the most populous countries, Indonesia’s CO₂ emissions rise with its population, highlighting the challenges of sustainable development in a densely populated country.

Laos shows a weak correlation between FDI and CO₂ emissions, indicating that foreign investments are either not large enough or not in carbon-intensive industries. There is a noticeable positive correlation, suggesting that economic activities are leading to higher CO₂ emissions. Laos shows some renewable energy capacity, but CO₂ emissions are still rising, indicating a partial reliance on fossil fuels. The correlation suggests that as the population grows, CO₂ emissions increase, pointing to higher energy demand and potential inefficiencies in energy use.

Malaysia shows a positive correlation, indicating that FDI is likely directed towards industries that contribute to CO₂ emissions. A clear positive trend suggests that economic growth is associated with higher CO₂ emissions, possibly due to industrial and manufacturing activities. RESs are increasing, but not enough to reduce CO₂ emissions significantly, suggesting the need for more aggressive renewable energy policies. A rising population correlates with higher CO₂ emissions, indicating that population growth is a significant factor in energy consumption and emissions.

There is a weak correlation, suggesting that FDI in Myanmar might not be heavily directed towards carbon-intensive sectors. A positive correlation indicates that economic development is leading to increased CO₂ emissions. Limited RES adoption is not significantly reducing CO₂ emissions, indicating a reliance on fossil fuels. An increasing population correlates with higher CO₂ emissions, highlighting the need for sustainable energy solutions.

The trend suggests a moderate correlation, indicating that foreign investments may be contributing to CO₂ emissions but not uniformly in the Philippines. The positive correlation between GDP and CO₂ emissions suggests that economic growth is associated with higher emissions. The increasing RES capacity shows some impact, but CO₂ emissions remain high, indicating the need for more RES adoption. As the population grows, CO₂ emissions increase, suggesting higher energy demands and a need for cleaner energy sources.

Singapore shows a varied trend, indicating that FDI might be diversified across different sectors with varying carbon intensities. There is a positive correlation, but Singapore has relatively lower CO₂ emissions despite its high GDP, indicating efficient energy use and possibly stringent environmental regulations. Singapore shows a limited renewable energy capacity, and CO₂ emissions remain stable, indicating a reliance on fossil fuels. The trend shows a correlation between population growth and CO₂ emissions, suggesting higher energy demands.

Thailand shows a positive correlation, suggesting that foreign investments are contributing to CO₂ emissions. A strong positive correlation indicates that economic growth is leading to higher CO₂ emissions, likely due to industrial and manufacturing activities. Increasing the renewable energy capacity is not significantly reducing CO₂ emissions, indicating a need for more aggressive renewable energy policies. The correlation indicates that as the population grows, CO₂ emissions increase, pointing to a higher energy demand and potential inefficiencies in energy use.

Vietnam shows a positive correlation, indicating that FDI is likely directed towards industries that contribute to CO₂ emissions. There is a strong positive correlation, suggesting that economic growth is associated with higher CO₂ emissions, likely due to industrialization and energy consumption. Despite some renewable energy capacity, CO₂ emissions remain high, suggesting that renewable energy is not yet sufficient to offset carbon emissions from other sources. An increasing population correlates with higher CO₂ emissions, highlighting the challenges of sustainable development in a densely populated country.

For most countries, economic growth is positively correlated with CO₂ emissions, indicating that industrial activities and energy consumption are key drivers. FDI shows varied impacts on CO₂ emissions across countries, suggesting that the nature of foreign investments and their environmental impact differ. The adoption of renewable energy is still insufficient in significantly reducing CO₂ emissions, indicating a need for more aggressive policies and investments. A higher population generally correlates with higher CO₂ emissions, emphasizing the need for sustainable energy solutions to meet the growing demand.

Scatter plots with trend lines for the LACs, showing the relationships between CO₂ emissions and the four indicators, are presented in Figure 3.

China shows a positive correlation, indicating that higher FDI is associated with higher CO₂ emissions. This suggests that foreign investments are contributing to industrial activities that increase CO₂ emissions.

The trend for the other two countries varies, with some showing a positive correlation and others showing little to no correlation, reflecting differences in how foreign investments impact CO₂ emissions. China displays a strong positive correlation, indicating that economic growth is leading to higher CO₂ emissions, likely due to industrial and economic activities. The other two countries generally show a positive correlation, suggesting that as GDP increases, CO₂ emissions also rise, highlighting the environmental cost of economic growth. China shows a complex relationship where increasing renewable energy does not significantly reduce CO₂ emissions, suggesting that despite renewable energy adoption, fossil fuel use remains substantial. For the other two countries, similar trends are observed, where renewable energy adoption is not yet sufficient to offset CO₂ emissions significantly. A positive correlation indicates that as the population grows in China, CO₂ emissions increase, reflecting higher energy demands. The other two LACs, generally show similar trends, suggesting that larger populations lead to higher CO₂ emissions, emphasizing the need for sustainable energy solutions. These interpretations suggest that LACs face similar challenges in balancing economic growth, foreign investments and population growth with the need to reduce CO₂ emissions. RES adoption, while progressing, has not yet achieved a scale sufficient to significantly impact CO₂ emissions.

A comparison of descriptive statistics for ASEAN countries and large Asian countries (LACs: China, India and Pakistan) is briefly provided. In terms of FDI, ASEAN countries have a higher mean FDI of 5.30 compared to 1.80 for LAC countries. The standard deviation for ASEAN is 5.75, indicating greater variability, with a minimum value of −2.76 and a maximum of 31.62. In contrast, LACs have a standard deviation of 1.43, a minimum of 0.03 and a maximum of 6.19. When looking at GDP, ASEAN countries have a significantly higher mean GDP of 9849.05 compared to 2409.36 for LAC countries. The standard deviation is also higher for ASEAN at 15183.36, with a range from 165.54 to 67948.89. For LACs, the standard deviation is 2631.34, with values ranging from 528.90 to 11560.24.

Regarding CO₂ emissions, ASEAN countries have a higher average of 4.59 compared to 2.31 for LAC countries. The standard deviation for ASEAN is 6.27, with emissions ranging from 0.10 to 26.72. LAC countries have a standard deviation of 2.26, with a range from 0.57 to 7.99. In terms of REN, both groups have similar mean values, with ASEAN at 36.77 and LACs at 37.01. However, ASEAN shows more variability, with a standard deviation of 30.10 and a maximum value of 91.10, compared to LAC’s standard deviation of 13.88 and maximum of 58.10. For the population variable, LAC countries have a substantially higher average population of 333.05 million compared to 24.01 million for ASEAN countries. The standard deviation for LACs is 246.17 million, with values ranging from 35.29 million to 897.58 million. ASEAN countries have a standard deviation of 32.64 million, with a range from 174,020 to 159.61 million.

Overall, ASEAN countries exhibit higher mean values and greater variability in FDI, GDP and CO₂ emissions compared to LAC countries. The RES adoption is similar in mean values between the two groups, but ASEAN shows higher variability. The population in LAC countries is significantly higher than in ASEAN, reflecting the inclusion of highly populous countries like China and India. The violin plots compare various metrics between ASEAN and LAC regions, as in Figure 4.

The violin plot for FDI shows the distribution of foreign direct investment for ASEAN and LAC countries. The ASEAN region exhibits a wider distribution with a long upper tail, indicating that some countries receive very high FDI. The median FDI for ASEAN is higher compared to LACs, which has a more compact distribution with lower FDI values overall. The GDP violin plot reveals that ASEAN countries have a more extensive range of GDP values, with some countries achieving a very high GDP. The plot shows a significant variation, with a long upper tail. In contrast, LAC countries have a more concentrated GDP distribution with lower values. This indicates that ASEAN includes some very economically strong countries. The CO₂ emissions plot illustrates that ASEAN countries have a wide range of emissions, with a higher peak and a broader spread. This indicates more variability in CO₂ emissions across ASEAN countries. LAC countries, on the other hand, show a more compact distribution with generally lower emissions. In the violin plot for REN, ASEAN again shows a wider range of values, indicating variability in how much different countries consume RESs. Some ASEAN countries consume a significant amount of RESs, as indicated by the long upper tail. LAC countries show a more consistent, but lower level of REN. The population plot shows that LAC countries generally have a higher population with a broader distribution. ASEAN countries have a lower and more consistent population range. This reflects the larger and more populous countries within the LAC region compared to the generally smaller populations in ASEAN countries. These violin plots illustrate that ASEAN countries exhibit greater variability and higher values in FDI, GDP, CO₂ emissions and REN compared to LAC countries. This variability suggests a mix of highly developed and developing countries within ASEAN, whereas LAC countries show more consistency.

5.2. Empirical Results

The dependence relationship is

$${\mathrm{C}\mathrm{O}}_2=\mathrm{f}(\mathrm{G}\mathrm{D}\mathrm{P},\mathrm{R}\mathrm{E}\mathrm{N},\mathrm{P}\mathrm{O}\mathrm{P},\mathrm{F}\mathrm{D}\mathrm{I})$$

(7)

In this paper, all variables are in log-linear form to linearize potential nonlinear relationships, reduce heteroscedasticity and allow for interpretation of coefficients as elasticities in both the short and long run.

$${{\mathrm{C}\mathrm{O}}_2}_{\mathrm{i}\mathrm{t}}=\mathsf{\alpha }+{\mathsf{\beta }}_1{\mathrm{G}\mathrm{D}\mathrm{P}}_{\mathrm{i}\mathrm{t}}+\mathsf{\beta }2{\mathrm{R}\mathrm{E}\mathrm{N}}_{\mathrm{i}\mathrm{t}}+{\mathsf{\beta }}_3{\mathrm{P}\mathrm{O}\mathrm{P}}_{\mathrm{i}\mathrm{t}}+{{\mathsf{\beta }}_4{\mathrm{F}\mathrm{D}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+\mathsf{\epsilon }}_{\mathrm{i}\mathrm{t}}$$

(8)

Next, we examine both the long-term and short-term impacts of GDP, REN, POP and FDI on CO₂ emissions for ASEAN and LACs. The results presented in

Table 2. CD test results for ASEAN.

Tests	Statistic	p-Value
Breusch–Pagan LM	262.60	0.000 ***
Pesaran-scaled LM	22.93	0.000 ***
Pesaran CD	6.40	0.000 **

**, *** significant at 5% and 1% levels.

and

Table 3. CD test results for LACs.

Tests	Statistic	p-Value
Breusch–Pagan LM	33.76	0.000 ***
Pesaran-scaled LM	12.56	0.000 ***
Pesaran CD	5.71	0.000 ***

*** significant at 1% level.

indicate that the hypothesis of “no CD” is rejected at the 1% significance level. Therefore, it is essential to employ tests and estimation methods that account for CD.

The data stationarity is checked by means of the first-generation PURT: LLC and ADF (

Table 4. First-generation PURT for ASEAN.

At levels
	CO2	GDP	REN	POP	FDI
Unit root (Common Unit Root Process)
LLC	−0.79 (0.212)	−2.63 (0.004) ***	1.22 (0.889)	−4.17 (0.000) ***	−3.87 (0.000) ***
Unit root (Individual Unit Root Process)
ADF-Fisher Chi-square	17.03 (0.650)	7.67 (0.993)	20.53 (0.424)	86.90 (0.000) ***	59.28 (0.000) ***
At first difference
Unit root (Common Unit Root Process)
LLC	−6.43 (0.000) ***	−7.04 (0.000) ***	−4.99 (0.000) ***	−3.81 (0.000) ***	−9.38 (0.000) ***
Unit root (Individual Unit Root Process)
ADF-Fisher Chi-square	128.55 (0.000) ***	105.01 (0.000) ***	86.18 (0.000) ***	31.16 (0.053) *	165.29 (0.000) ***

*, *** significant at 10% and 1% levels.

and

Table 5. First-generation PURT for LACs.

At levels
	CO2	GDP	REN	POP	FDI
Unit root (Common Unit Root Process)
LLC	−0.82 (0.203)	−0.86 (0.192)	−1.35 (0.088) *	−2.83 (0.002) ***	−1.87 (0.030) **
Unit root (Individual Unit Root Process)
ADF-Fisher Chi-square	2.27 (0.893)	1.38 (0.966)	3.26 (0.775)	9.11 (0.167)	13.73 (0.032) **
At first difference
Unit root (Common Unit Root Process)
LLC	−4.82 (0.000) ***	−378 (0.000) ***	−3.27 (0.000) ***	−2.39 (0.008) ***	−3.83 (0.000) **
Unit root (Individual Unit Root Process)
ADF-Fisher Chi-square	41.58 (0.000) ***	31.80 (0.000) ***	34.27 (0.000) ***	10.88 (0.092) *	44.94 (0.000) ***

*, **, *** significant at 10%, 5% and 1% levels.

).

Table 6. Second-generation PURT for ASEAN.

At levels
	CO2	GDP	REN	POP	FDI
CIPS	−3.06 (<0.01)	−2.50 (<0.05)	−1.64 (≥0.10)	−2.22 (<0.10)	−3.00 (<0.01)
At first difference
CIPS	−6.49 (<0.01)	−2.76 (<0.01)	−3.81 (<0.01)	−3.12 (<0.01)	−4.53 (<0.01)

and

Table 7. Second-generation PURT for LACs.

At levels
	CO2	GDP	REN	POP	FDI
CIPS	−2.22 (<0.10)	−1.13 (≥0.10)	−3.08 (<0.01)	−1.24 (≥0.10)	−2.82 (<0.01)
At first difference
CIPS	−3.68 (<0.01)	−2.31 (<0.10)	−4.95 (<0.01)	−2.42 (<0.05)	−5.00 (<0.01)

contain the second-generation PURT: CIPS.

Considering the aforementioned tests, it is obvious that all variables achieve stationarity after first differencing. Given that all variables are either I(0) or I(1), and none are integrated of order two, the panel ARDL model is appropriate for our analysis. Its flexibility in handling mixed integration orders eliminates the need for applying additional panel cointegration tests such as Pedroni or Kao.

Subsequently, we employ the panel ARDL model for ASEAN and LACs in

Table 8. Panel ARDL (4,3,3,3,3) for ASEAN.

Indicator	Coefficient	Marginal Effect	t-Statistic (Prob. *)
	Long-Run Equation	Long Run
GDP	0.46	${∆\mathrm{C}\mathrm{O}}_2/∆\mathrm{G}\mathrm{D}\mathrm{P}=0.46$	4.53 (0.000 ***)
REN	−0.08	${∆\mathrm{C}\mathrm{O}}_2/∆\mathrm{R}\mathrm{E}\mathrm{N}=-0.08$	−3.79 (0.000 ***)
POP	0.10	${∆\mathrm{C}\mathrm{O}}_2/∆\mathrm{P}\mathrm{O}\mathrm{P}=0.10$	0.68 (0.495)
FDI	0.06	${∆\mathrm{C}\mathrm{O}}_2/∆\mathrm{F}\mathrm{D}\mathrm{I}=0.06$	5.07 (0.000 ***)
	Short-Run Equation
COINTEQ01	−0.37		−2.35 (0.019 **)
D(CO2(-1))	−0.12		−0.89 (0.373)
D(CO2(-2))	−0.07		−0.52 (0.539)
D(CO2(-3))	−0.02		−0.18 (0.855)
D(GDP)	−0.45		−0.51 (0.606)
D(GDP(-1))	−0.63		−0.74 (0.454)
D(GDP(-2))	1.36		1.94 (0.053 *)
D(REN)	−1.06		−1.80 (0.072 *)
D(REN(-1))	−1.26		−1.87 (0.063 *)
D(REN(-2))	0.04		0.09 (0.926)
D(POP)	4.49		0.33 (0.739)
D(POP(-1))	−9.48		−0.46 (0.644)
D(POP(-2))	2.98		0.27 (0.787)
D(FDI)	−0.03		−1.15 (0.251)
D(FDI(-1))	−0.06		−1.73 (0.085 *)
D(FDI(-2))	−0.04		−1.26 (0.207)
C	−1.49		−2.12 (0.035 **)

*, **, *** significant at 10%, 5% and 1% levels.

and

Table 9. Panel ARDL (2,5,5,5,5) for LACs.

Indicator	Coefficient	Marginal Effect	t-Statistic (Prob. *)
	Long-Run Equation	Long Run
GDP	0.20	${∆\mathrm{C}\mathrm{O}}_2/∆\mathrm{G}\mathrm{D}\mathrm{P}=0.20$	41.50 (0.000 ***)
REN	−0.56	${∆\mathrm{C}\mathrm{O}}_2/∆\mathrm{R}\mathrm{E}\mathrm{N}=-0.56$	−143.07 (0.000 ***)
POP	0.17	${∆\mathrm{C}\mathrm{O}}_2/∆\mathrm{P}\mathrm{O}\mathrm{P}=0.17$	18.65 (0.000 ***)
FDI	0.01	${∆\mathrm{C}\mathrm{O}}_2/∆\mathrm{F}\mathrm{D}\mathrm{I}=0.01$	34.36 (0.000 ***)
	Short-Run Equation
COINTEQ01	−1.18		−2.51 (0.018 **)
D(CO2(-1))	−0.005		−0.03 (0.970)
D(GDP)	0.43		0.91 (0.370)
D(GDP(-1))	0.76		3.40 (0.002 ***)
D(GDP(-2))	0.15		0.49 (0.623)
D(GDP(-3))	1.06		1.69 (0.101)
D(GDP(-4))	−0.39		−1.59 (0.122)
D(REN)	0.09		0.25 (0.804)
D(REN(-1))	−0.28		−1.79 (0.084 *)
D(REN(-2))	−0.11		−0.73 (0.467)
D(REN(-3))	−0.12		−0.39 (0.693)
D(REN(-4))	−0.05		−0.17 (0.862)
D(POP)	−29.24		−1.00 (0.326)
D(POP(-1))	20.07		0.75 (0.456)
D(POP(-2))	−3.15		−0.14 (0.887)
D(POP(-3))	−11.62		−1.22 (0.232)
D(POP(-4))	−19.34		−0.73 (0.468)
D(FDI)	0.04		1.24 (0.225)
D(FDI(-1))	0.03		0.68 (0.500)
D(FDI(-2))	−0.002		−0.16 (0.867)
D(FDI(-3))	0.02		2.28 (0.030 **)
D(FDI(-4))	0.006		1.16 (0.254)
C	−2.03		−4.18 (0.000 ***)

*, **, *** significant at 10%, 5% and 1% levels.

. Table 8 and Table 9 present the long-run marginal effects (elasticities) derived from the panel ARDL model, quantifying the percentage change in CO₂ emissions associated with a 1% change in each explanatory variable, holding other factors constant.

For ASEAN, one can see from Table 8 that in the long run, a 1% increase in GDP implies a 0.46% increase in CO₂. For ASEAN countries, the long-term relationship between GDP and CO₂ emissions is quantified by the elasticity measure. This indicates that economic growth in ASEAN tends to lead to higher carbon emissions, though the growth rate of emissions is less than proportional to the GDP growth rate. This relationship suggests that while economic development is crucial, it brings about environmental challenges, highlighting the need for sustainable growth strategies that mitigate CO₂ emissions. The long-term analysis reveals that an increase in REN has a beneficial impact on CO₂ emissions. Specifically, a 1% increase in REN results in a 0.08% decrease in CO₂ emissions. This indicates that boosting the use of RESs effectively contributes to reducing carbon emissions. It underscores the importance of investing in and promoting RESs as a strategy for mitigating climate change and achieving more sustainable economic growth in the ASEAN region.

In the ASEAN panel, the coefficient for POP is statistically non-significant in the long run, as indicated by a t-value of 0.68 and a p-value of 0.495. This result suggests that population growth does not exert a measurable or systematic impact on CO₂ emissions within ASEAN over the long term. One possible explanation is that demographic changes in ASEAN economies are accompanied by structural shifts, such as urbanization or increased energy efficiency, that may offset the expected pressure of population growth on emissions. Alternatively, it may reflect that population-driven emissions are mediated by other factors such as economic activity, industrial composition or renewable energy adoption. The lack of statistical significance implies that POP alone is not a robust predictor of emissions in the ASEAN region, and it should be interpreted cautiously in the context of environmental policy formulation.

In the long run, a 1% increase in FDI exerts a 0.06% increase in CO₂. FDI often fuels industrial growth, leading to increased production and energy consumption. This industrial expansion typically relies on energy-intensive processes, contributing to higher CO₂ emissions. FDI usually supports infrastructural projects such as the construction of factories, roads and other facilities. These activities increase energy use and emissions, both during construction and subsequent operation. While FDI can bring advanced technologies that may be more efficient, the initial phases of technology transfer and setup might lead to higher emissions. In some cases, developing countries might also receive older, less efficient technologies from foreign investors. Increased FDI boosts overall economic activity, leading to greater demand for energy across various sectors, including transportation, manufacturing and services. This heightened energy demand typically results in higher CO₂ emissions. The environmental impact of FDI depends on the host country’s regulatory framework. In ASEAN countries, if environmental regulations are not stringent enough, increased economic activities from FDI can lead to higher emissions.

A statistically significant ECT of −0.37 suggests that if CO₂ emissions deviate from their equilibrium path due to changes in GDP, REN, POP or FDI, about 37% of this deviation will be corrected in the following period, ensuring that the variables converge back to their long-term relationship over time.

For LACs, one can see from Table 9 that a 1% increase in GDP leads to a 0.20% long-term increase in CO₂ emissions. This suggests that economic expansion in these nations, which are among the world’s most populous and rapidly developing, is associated with higher levels of CO₂ emissions. The positive correlation between GDP growth and CO₂ emissions underscores the environmental cost of economic development in these countries. As they continue to industrialize and urbanize, the demand for energy and resources increases, leading to higher emissions. A 1% increase in REN results in a 0.56% long-term decrease in CO₂ emissions. The strong negative correlation between REN and CO₂ emissions suggests that increasing the use of RESs can substantially lower carbon emissions. This indicates a viable path for these countries to reduce their carbon footprint. A 1% increase in POP results in a 0.17% long-term increase in CO₂ emissions. An increasing population leads to greater demand for energy, transportation, housing and other resources, which in turn increases CO₂ emissions. Population growth often drives urbanization and industrialization, further contributing to CO₂ emissions through increased construction and industrial activities. Managing the environmental impact of population growth poses significant challenges for sustainable development. A 1% increase in FDI leads to a 0.01% long-term increase in CO₂ emissions. While this increase is relatively small, it highlights a noteworthy relationship between economic activity driven by FDI and environmental impact. The positive correlation suggests that FDI, which often brings industrial development and infrastructural projects, can lead to higher CO₂ emissions due to increased production and energy use. This relationship underscores the importance of channeling FDI into sustainable and green technologies to minimize the environmental footprint of economic growth.

The ECT coefficient of −1.18 indicates a strong adjustment back to equilibrium in the case of short-term deviations. This strong error correction implies that any short-term shocks to CO₂ emissions are quickly corrected. LACs exhibit a much faster adjustment speed towards long-run equilibrium than ASEAN, as indicated by the more negative ECT. For LACs, policies to address deviations from long-run equilibrium can have quicker effects, while in ASEAN, policy effects will take longer to stabilize the system. The faster adjustment in LACs suggests higher economic stability in the face of shocks compared to ASEAN, where adjustments are more gradual.

The subsequent step involves performing a causality test to determine the direction of causality between variables. The D-H causality test is utilized to identify whether changes in one variable cause changes in another, if the relationship is reciprocal or if a bidirectional causality exists (

Table 10. D-H panel causality test for ASEAN.

Null Hypothesis (H0)	W-Stat.	Zbar-Stat.	Prob.	Conclusion
GDP n.c.i CO2	7.11	6.64	3 × 10−11 ***	GDP→CO2
CO2 n.c.i GDP	2.66	0.67	0.500
REN n.c.i CO2	6.31	5.57	3 × 10−8 ***	REN→CO2
CO2 n.c.i REN	3.23	1.43	0.150
POP n.c.i CO2	9.61	10.00	0.000 ***	POP→CO2
CO2 n.c.i POP	2.82	0.88	0.374
FDI n.c.i CO2	1.51	−0.87	0.383
CO2 n.c.i FDI	2.79	0.84	0.397
REN n.c.i GDP	2.34	0.24	0.810	REN→GDP
GDP n.c.i REN	6.72	6.12	9 × 10−10 ***	GDP→REN
POP n.c.i GDP	10.27	10.88	0.000 ***	POP→GDP
GDP n.c.i POP	10.08	10.63	0.000 ***	GDP→REN
FDI n.c.i GDP	2.14	−0.02	0.981
GDP n.c.i FDI	4.52	3.16	0.001 ***	GDP→FDI
POP n.c.i REN	5.02	3.83	0.000 ***	POP→REN
REN n.c.i POP	4.21	2.75	0.005 ***	REN→POP
FDI n.c.i REN	2.66	0.67	0.500
REN n.c.i FDI	3.76	2.14	0.031 **	REN→FDI
FDI n.c.i POP	3.64	1.99	0.046 **	FDI→POP
POP n.c.i FDI	6.12	5.30	1 × 10−7 ***	POP→FDI

**, *** significant at 5% and 1% levels; n.c.i stands for “no causal influence”.

and

Table 11. D-H panel causality test for LACs.

Null Hypothesis (H0)	W-Stat.	Zbar-Stat.	Prob.	Conclusion
GDP n.c.i CO2	2.68	0.38	0.703
CO2 n.c.i GDP	2.73	0.41	0.677
REN n.c.i CO2	3.99	1.34	0.178
CO2 n.c.i REN	2.31	0.10	0.913
POP n.c.i CO2	9.44	5.35	9 × 10−8 ***	POP→CO2
CO2 n.c.i POP	4.69	1.85	0.063 *	CO2→POP
FDI n.c.i CO2	1.21	−0.69	0.485
CO2 n.c.i FDI	7.70	4.07	5 × 10−5 ***	CO2→FDI
REN n.c.i GDP	5.22	2.25	0.024 **	REN→GDP
GDP n.c.i REN	3.44	0.93	0.349
POP n.c.i GDP	6.99	3.55	0.000 ***	POP→GDP
GDP n.c.i POP	2.65	0.35	0.721
FDI n.c.i GDP	1.65	−0.37	0.708
GDP n.c.i FDI	12.15	7.34	2 × 10−13 ***	GDP→FDI
POP n.c.i REN	6.90	3.48	0.000 ***	POP→REN
REN n.c.i POP	6.01	2.83	0.004 ***	REN→POP
FDI n.c.i REN	3.38	0.89	0.368
REN n.c.i FDI	5.04	2.11	0.034 **	REN→FDI
FDI n.c.i POP	3.73	1.14	0.250
POP n.c.i FDI	12.55	7.64	2 × 10−14 ***	POP→FDI

*, **, *** significant at 10%, 5% and 1% levels; n.c.i stands for “no causal influence”.

).

In Table 10, the significant causality from GDP to CO₂ emissions indicates that higher economic activity, which often involves industrialization and increased energy consumption, contributes to greater environmental pollution. The absence of significant causality from CO₂ emissions to GDP implies that current levels of pollution do not yet hinder economic performance, or the economy is resilient to such environmental impacts. The significant negative causality of REN to CO₂ emissions suggests that increasing the use of RESs helps reduce CO₂ emissions. This highlights the importance of investing in RESs as a strategy for mitigating climate change. No significant causality from CO₂ emissions to REN indicates that rising CO₂ levels do not drive an immediate switch to RESs, possibly due to economic, political or infrastructural barriers. The significant positive causality from population growth to CO₂ emissions suggests that as POP increases, CO₂ emissions rise, likely due to higher energy demand, increased transportation and greater consumption of goods and services. The lack of significant causality from CO₂ emissions to population growth suggests that higher CO₂ emissions do not immediately affect population trends, possibly because the adverse health impacts of pollution might not be severe enough yet to affect population size. The lack of significant causality from FDI to CO₂ emissions implies that foreign investments are not a direct driver of CO₂ emissions. This could be because FDI in ASEAN countries might be directed towards sectors with varying environmental impacts. Similarly, CO₂ emissions do not significantly influence FDI, indicating that environmental quality is not yet a critical factor for foreign investors in these countries. The lack of significant causality from REN to GDP suggests that increasing renewable energy use does not have an immediate noticeable effect on economic growth. This could be due to the initial costs and time required for RES investments to translate into economic benefits. Significant causality from GDP to REN indicates that economic growth enables greater investment in RESs. As economies grow, they can afford to allocate more resources towards sustainable energy projects.

From Table 11, one can see that there is no significant causality from GDP to CO₂ emissions, indicating that economic growth does not have a uniform impact on CO₂ emissions across LAC countries. This could be due to varying energy policies, industrial practices and levels of economic development. Similarly, CO₂ emissions do not significantly influence GDP, suggesting that environmental pollution does not have a consistent effect on economic growth across the LAC panel. There is no significant causality from REN to CO₂ emissions, implying that increasing RES usage does not uniformly reduce CO₂ emissions. This may be due to the small share of RESs in the overall energy mix. CO₂ emissions do not significantly affect REN, indicating that higher emissions do not drive a uniform shift towards renewable energy sources. There is significant causality from population growth to CO₂ emissions, suggesting that an increase in POP leads to higher CO₂ emissions due to greater energy demand and consumption. There is a weak causality from CO₂ emissions to population, implying that higher CO₂ emissions may slightly affect population growth, possibly through health and environmental impacts. There is a weak causality from CO₂ emissions to population, implying that higher CO₂ emissions may slightly affect population growth, possibly through health and environmental impacts. There is significant causality from CO₂ emissions to FDI, suggesting that higher CO₂ emissions may attract or deter foreign investments depending on environmental regulations and business practices. There is significant causality from REN to GDP, indicating that an increase in RES usage positively impacts economic growth. This could be due to the creation of green jobs and sustainable development. Economic growth does not significantly drive REN, suggesting that GDP growth alone does not ensure an increase in the use of RESs. There is significant causality from population growth to GDP, indicating that a growing population contributes to economic growth through an increased labor supply and consumption. Economic growth does not significantly impact population growth, implying that higher GDP does not uniformly lead to changes in population dynamics. FDI does not significantly drive economic growth across the panel, suggesting that foreign investments alone do not uniformly enhance GDP. There is significant causality from GDP to FDI, indicating that higher economic growth attracts more foreign investments, likely due to improved business opportunities and economic stability. There is significant causality from population growth to REN, suggesting that a larger population increases the demand for RESs as part of the overall energy mix. There is significant causality from REN to population growth, indicating that the use of RESs may positively influence population growth, possibly through improved environmental quality and health benefits. FDI does not significantly impact REN, suggesting that foreign investments do not uniformly promote the use of RESs. There is significant causality from REN to FDI, indicating that higher RES usage attracts more foreign investments, possibly due to a favorable investment climate and sustainability initiatives. FDI does not significantly affect population growth, suggesting that foreign investments do not uniformly influence demographic trends. There is significant causality from population growth to FDI, indicating that a growing population attracts more foreign investments due to the potential for a larger market and workforce.

6. Discussion

Our research provides new empirical evidence on the dynamic drivers of CO₂ emissions across ASEAN and LACs using the panel ARDL framework. The results reveal meaningful long-run relationships between economic growth, foreign direct investment, renewable energy consumption and emissions, with notable regional asymmetries.

The positive and statistically significant relationship between GDP and CO₂ emissions in both regions is consistent with previous findings (e.g., [41,42]), suggesting that economic expansion remains carbon-intensive. Although our model does not explicitly test the EKC due to the absence of a squared GDP term, the findings align with the early phase of the EKC hypothesis, where initial economic growth contributes to environmental degradation. Future research may investigate this relationship further by incorporating nonlinear terms or threshold models.

Differences in the environmental impact of FDI between regions deserve particular attention. The FDI elasticity in LACs (0.01) is much smaller than in ASEAN (0.06), implying that FDI inflows into LACs may be more aligned with low-emission sectors or accompanied by stricter environmental regulation, possibly reflecting stronger institutional oversight or higher Environmental Social Governance (ESG) compliance. In contrast, ASEAN economies may experience greater environmental pressure from FDI, consistent with the pollution haven hypothesis, where capital flows toward countries with laxer environmental standards [43]. This finding highlights the importance of screening FDI not only for economic value but also for environmental compatibility.

The consistently negative and significant relationship between REN and CO₂ emissions confirms the role of clean energy in environmental mitigation. These findings corroborate the work of [44,45], reinforcing the idea that investment in renewable infrastructure can decouple emissions from growth. However, the varying magnitudes between ASEAN and LACs also reflect differing levels of renewable adoption and institutional capacity to scale such technologies.

Population was statistically insignificant in the ASEAN sample, suggesting that demographic growth does not automatically translate into higher emissions. This insignificant relation between POP and CO₂ emissions may be due to factors like energy efficiency gains, shifts toward service-based economies or urban planning reforms. It also points to the idea that emissions intensity is not solely driven by headcount, but by how populations consume energy and resources.

This research acknowledges certain limitations. One such limitation is the role of technological progress, which can significantly influence carbon emissions by enhancing energy efficiency, supporting cleaner production processes and facilitating the adoption of renewable energy technologies. Future studies should aim to incorporate indicators of technological advancement, such as R&D expenditure, patent counts in environmental technologies or innovation indices, into the econometric framework.

Another limitation is the omission of governance-related control variables, which may differ substantially across countries in the sample. For instance, variations in institutional strength or environmental regulation, such as between Singapore and Myanmar, could impact carbon emissions. Governance disparities may influence emission outcomes and could contribute to unobserved heterogeneity across countries.

Additional limitation concerns the temporal coverage of the dataset; due to data unavailability for REN in the World Bank database, the current dataset ends in 2022.

Despite the robust panel ARDL framework, several methodological limitations must be acknowledged. First, the potential endogeneity among variables—such as the bidirectional causality between GDP and CO₂ emissions—may bias coefficient estimates. Second, differences in data quality, coverage and comparability across countries and over time may introduce measurement errors. Third, ARDL models are sensitive to lag structure selection, and mis-specification can affect the validity of the estimated long-run relationships and speed of adjustment parameters.

7. Conclusions and Policy Recommendations

Our research quantitatively examined the dynamic relationship between CO₂ emissions and some macroeconomic and energy indicators (GDP, REN, FDI, POP) across ASEAN and LACs using a panel ARDL framework. The results reveal significant long-run and short-run effects, with notable regional variation in adjustment speeds and variable responsiveness. These findings provide a quantitative foundation for designing regionally differentiated carbon reduction strategies. In particular, the faster equilibrium convergence in LAC economies suggests greater readiness for aggressive mitigation efforts, while ASEAN countries may benefit from capacity-building and incremental policy approaches.

The analysis reveals that REN significantly reduces emissions, justifying investment and regulatory support. While FDI has a moderate but varying impact, future research should explore sector-specific FDI flows, such as in manufacturing, energy and construction, to better understand their environmental externalities and policy implications.

The findings of our research suggest several targeted policy actions. First, in ASEAN countries with slower adjustment speeds, governments should enhance region-specific renewable energy subsidies. This can be achieved by implementing targeted tax incentives and financial support mechanisms for investments in solar, wind and hydro energy, particularly in rural and industrial areas where the energy transition is often more challenging.

Second, both ASEAN and LAC countries should facilitate technology transfer by incorporating environmental screening into FDI approval processes. Priority should be given to low-emission sectors and companies that prove an environmental commitment through adherence to environmental, social and governance criteria. This could encourage the inflow of cleaner technologies and more sustainable production practices.

Third, urban planning policies should be aligned with emission reduction goals through the implementation of carbon accounting frameworks. Local governments and municipalities are encouraged to integrate carbon budgeting into zoning regulations, focusing on the development of low-carbon transportation networks, green buildings and compact, mixed-use neighborhoods that reduce dependency on fossil fuels.

Fifth, regional cooperation should be strengthened to support knowledge sharing and joint innovation in clean energy solutions. ASEAN and LACs may benefit from forming regional platforms to share best practices, co-finance clean energy research and infrastructure and jointly access international climate finance mechanisms to support a just and inclusive transition.