Drivers of Environmental Sustainability, Economic Growth, and Inequality: A Study of Economic Complexity, FDI, and Human Development Role in BRICS+ Nations

Abstract

This study investigates the intricate relationships among CO₂ emissions, income inequality, the Economic Complexity Index (ECI), foreign direct investment (FDI), the Human Development Index (HDI), and the economic growth across countries. Three distinct models are developed: the first examines their effects on economic growth, the second analyzes their impact on income inequality, and the third explores their influence on CO₂ emissions. Advanced econometric methods, including Fully Modified Ordinary Least Squares (FMOLS) and Dynamic Ordinary Least Squares (DOLS), are employed to ensure robust and reliable results. The findings indicate that income inequality impedes economic growth, whereas economic growth and greater economic complexity help reduce inequality. While FDI significantly boosts GDP growth, it also widens the income disparities and intensifies environmental degradation, raising questions about the sustainability and quality of foreign investments. In contrast, human development emerges as a vital driver of economic growth and a critical factor in reducing CO₂ emissions, highlighting the value of investing in education, healthcare, and living standards to achieve sustainable development. These insights underscore the necessity for carefully designed policies that harmonize economic progress, social equity, and environmental sustainability.

1. Introduction

The global economy has experienced unprecedented growth since the industrialization phase started. Simultaneously, carbon dioxide (CO₂) emission has increased substantially over the previous years. The rate at which CO₂ emission has increased has shattered the existing high temperature records [1]. CO₂ emission is one of the major contributors to the rise in the Earth’s temperature and thereby primary reason for global climate change. Mitigating environmental degradation and economic stability are the primary concerns in the current times. However, economic growth can be boosted through industrial expansion which eventually increases the CO₂ emission. The policymakers are motivated to learn how economic growth can be increased without environmental degradation. One popular hypothesis widely applied in the environmental economics which quantify the economic growth and environmental degradation nexus is the Environmental Kuznets Curve (EKC). This hypothesis is an inverted U-shaped curve which implies that as the economic growth increases, environmental pollution rises initially and then begins to drop during the later period of growth [2]. Existing studies fail to find a consensus on the validity of the EKC. For instance [3] empirically provided support for the validity of the EKC hypothesis. Ref. [4] established an N-shaped curve, while ref. [5] lent their support to a monotonic rising curve. There is a continuous effort of pursuing sustainable development in the growing economies. There is a trade-off involved between a pollution-free environment and economic development in most instances. Increased economic activity raises energy consumption and CO₂ emission, which are the underlying causes of the greenhouse effect [6]. Therefore, establishing a link between environmental degradation, which is measured by CO₂ emission, and economic growth has been a central topic among the researchers in relation to sustainable growth. Foreign direct investment (FDI) plays a pivotal role in economic growth, alongside the key factors such as interest rates, exports, employment, and government expenditure. Over the past two decades, FDI has surged, attracting significant policy attention due to its potential benefits and associated costs. Particularly in emerging economies, FDI serves as a crucial driver of technological innovation, global market integration, and trade expansion [7]. Countries actively seek FDI for its advantages, including capital inflows, advanced technology transfers, and job creation [8]. Notably, FDI has outpaced foreign trade in recent years, with production-related inflows exceeding the goods traded internationally [9]. Consequently, extensive research has explored FDI’s role in stimulating economic growth [10,11,12]. The environmental impact of FDI remains a subject of debate. While increased economic activity driven by FDI raises energy consumption and carbon emissions, some argue that weak environmental regulations may attract polluting firms seeking cost advantages [13]. Conversely, FDI can facilitate the transfer of cleaner, more efficient technologies, helping host countries mitigate pollution [14]. As a result, the net effect of FDI on the environmental quality remains inconclusive, warranting further investigation. Economic complexity plays a crucial role in shaping the environmental quality by influencing a country’s productive structure and energy consumption patterns. As industrialization advances, the production of complex goods—such as cement, metal, and textiles—can increase pollution and CO₂ emission due to their energy-intensive nature [15]. However, economic complexity also fosters research and innovation, driving the development of cleaner technologies and alternative energy sources. Countries with a higher economic complexity have greater potential to optimize resource use and transition toward sustainable energy. Whether economic complexity ultimately mitigates or exacerbates CO₂ emission remains an open question.

Economic growth has traditionally been linked to rising income but often overlooks broader aspects of social well-being. A narrow focus on economic growth alone is insufficient for achieving sustainable development, leading to a shift toward human development indices. The Human Development Index (HDI) is a key measure of economic success, emphasizing overall well-being rather than just monetary wealth. Sustainable development advocates stress the need to reduce CO₂ emission to safeguard human welfare, as exposure to greenhouse gases increases health risks, hospitalizations, and mortality rates. While lower emissions are expected to enhance human development, this relationship is not always straightforward. Limiting carbon-intensive energy consumption may hinder economic expansion, particularly in developing countries reliant on non-renewable energy sources [16]. Ref. [17] argue that developing nations transitioning to advanced stages of development will require substantial carbon emissions, raising concerns that emission reduction efforts could inadvertently slow human progress. Thus, a thorough investigation of the link between CO₂ emission and human development is necessary to assess the broader implications of emission reduction policies. Rising income inequality has also emerged as a critical challenge [18] affecting both economic stability and environmental sustainability. Inequitable income distribution suppresses the aggregate demand, leading to economic stagnation [19] and has significant implications for climate change [20,21,22]. Several theories explain the relationship between income inequality and environmental quality. The political economy perspective suggests that as inequality declines, more individuals demand better environmental conditions [23]. The marginal propensity to emit theory argues that environmental degradation varies with income distribution [24], while the consumption competition hypothesis posits that rising inequality drives excessive energy consumption, as wealthier individuals spend on goods to signal their social status rather than functional necessity [25]. Despite extensive research, there is no consensus on how income inequality influences CO₂ emission, highlighting the need for further empirical investigation. The world leaders are making efforts to reduce the repercussions of climate change by putting considerable amounts of resources and funds into tackling the climate issues. The Paris Agreement on climate change, established under the United Nations Framework Convention on Climate Change (UNFCCC), was adopted by 196 countries. It aims to limit the global temperature rise to below 2 °C, with efforts to further restrict it to 1.5 °C, by reducing atmospheric CO₂ emission. At this stage, it has become imperative to develop such policies which effectively limit environmental degradation. In view of this, the aim of the current analysis is to understand the role of various economic variables in the environment quality referring to CO₂ emission in BRICS+ nations.

BRICS+ comprising Brazil, Russia, India, China, South Africa, Iran, Egypt, Ethiopia, and the United Arab Emirates (UAE) refers to a group of rapidly growing emerging economies. These economies have demonstrated impressive economic growth in the recent years. The BRICS+ economies account for 37.3% of the world’s GDP, reflecting their substantial economic power. They also play a significant role in global CO₂ emission, contributing approximately 45.8% of the total, with coal consumption being a major driving factor. Consequently, BRICS economies have taken several measures to improve the environmental conditions. For instance, in 2017, BRICS economies released a joint statement at the environmental ministers’ conference convened in Tianjin, China, and called for strengthening the cooperation in the area of sustainable development. Further, the Xiamen declarations of BRICS leaders in September 2017 declared that all the countries should thoroughly enforce the Paris Agreement following the principle of common and differentiated responsibilities. Thus, BRICS+ economies have started taking initiatives tobecome low-carbon economies and find ways to tackle the environmental problems. At this stage, it is important to analyze the determinants of CO₂ emission in BRICS+ economies in order to help the policymakers make more effective decisions towards sustainable development.

Figure 1, Figure 2 and Figure 3 highlight the trends in economic growth, CO₂ emission, and income inequality, respectively, during the research period from 1995 to 2022. Figure 1 shows a continuous increase in GDP for China, with a similar trend observed in Ethiopia. In most countries, the impact of COVID-19 is evident, as seen in the downward turn of economic growth during the pandemic. Figure 2 illustrates the income inequality over time, with the highest inequality levels observed in India and South Africa. The income inequality in China also increased, albeit with fluctuations. Regarding CO₂ emission, China, Ethiopia, and India displayed similar upward trends. In contrast, CO₂ emission in Iran and Saudi Arabia declined sharply in 2020 due to the lockdowns, as shown in Figure 3.

It is against this backdrop that this study aims to investigate the effect of these variables on the CO₂ emission, economic growth, and income inequality in BRICS+ nations. To address the objective of this study, a long annual data spanning from 1995 to 2022 was utilized and employed theFully Modified Ordinary Least Squares (FMOLS), Dynamic Ordinary Least Squares (DOLS), and Autoregressive distributed lag (ARDL) regression techniques. Thus, the findings of this study will offer a deeper perspective for BRICS+ nation’s policymakers to understand the impact of FDI, the ECI, and HDI on countries’ environmental sustainability, growth, and income disparity.

The present study contributes to the existing literature in the following ways: First, this is the first study which analyzed the dynamic role of economic growth, income inequality, FDI, the HDI and ECI on CO₂ emission in the BRICS+ nations. A previous strand of the literature lacked evidence for considering newly added nations of BRICS while analyzing the interaction between these economic variables. In addition, this study employed models which examine the effect of FDI, the HDI, and ECI on GDP, income inequality (GINI), and CO₂ emission simultaneously. The existing literature lacks this dynamic interaction between these variables. This investigation will be expedient in designing environmental policies, keeping in view the important role of economic growth and income inequality.

2. Literature Review

With the emergence of the concept of sustainable development, the existing literature has focused on analyzing the interrelationship between CO₂ emission and various economic variables to interrogate the role of each variable in the environment’s sustainability. The detail of such studies is discussed below.

2.1. ECI and CO₂ Emission

The recent research highlights the environmental implications of economic complexity as a key structural distinction among national economies. Using a sample of 25 EU countries, ref. [26] suggested that nations with a higher economic complexity experience significant reductions in pollution. This phenomenon can be attributed to the differences in energy composition and efficiency. Similarly, ref. [27] reported that while economic complexity fosters economic growth, it also elevates carbon emissions. However, the expansion of economic complexity enhances the energy efficiency, thereby reducing the energy intensity. Ref. [28] found that, initially, pollution levels rise with the increasing complexity of exported products. However, beyond a certain threshold, a higher economic complexity leads to a decline in pollutant emissions. Analyzing data from 28 OECD countries, ref. [29] examined the dynamics of economic complexity, renewable energy consumption, and carbon emissions. Their study supports the argument that economic complexity, coupled with technological innovation, significantly reduces CO₂ emission. Conversely, ref. [30] revealed that increasing economic complexity does not contribute to mitigating CO₂ emissions in MINT economies. Similarly, ref. [31] found that the economic complexity in China exacerbates the environmental pollution and increases the ecological footprints. Ref. [32] analyzing 102 countries categorized as low and high income, disclosed that the impact of the economic complexity varies across income groups. While the economic complexity reduces CO₂ emissions in wealthier nations, it deteriorates the environmental quality in lower-income countries. Regarding the nonlinear impact of economic complexity, ref. [33] using data from 86 countries, rejected the hypothesis of an inverted U-shaped relationship between economic complexity and CO₂ emission. However, they noted an exception for developed countries, where economic complexity expansion leads to a gradual decline in pollution levels. In contrast, ref. [34] presented somewhat contradictory findings for BRICS economies, demonstrating an inverted U-shaped relationship between economic complexity and CO₂ emissions. Their study suggests that a higher level of economic complexity benefits the environment, whereas a lower economic complexity contributes to environmental degradation. The study particularly emphasizes the role of technological innovation in enhancing the economic complexity and reducing CO₂ emission.

2.2. Economic Performance and CO₂ Emission

While analyzing the relationship between economic growth and CO₂ emission, previous research has primarily focused on the decoupling effect between CO₂ emissions and economic growth. The decoupling effect ensures that continued economic expansion is accompanied by improvements in environmental quality. Based on decoupling theory, ref. [35] conducted a quantitative evaluation of the nexus between economic performance and greenhouse gas emissions. Their findings suggested a strong decoupling pattern between economic growth and greenhouse gas emissions in V4 countries. Ref. [36] examined the decoupling relationship in 289 Chinese cities using nighttime light data and found that China has achieved economic growth without significantly compromising the environmental quality. Moreover, the decoupling effect was notably stronger in the northeastern and eastern regions compared to other parts of the country. This finding contrasts with [37] whose study on BRICS nations concluded that China’s economic output and CO₂ emissions exhibit an expansive coupling relationship. Additionally, their research indicated that in Russia and South Africa, CO₂ emissions have been completely decoupled from economic growth for an extended period. Ref. [38] analyzed the decoupling effect in 192 countries from 2000 to 2014. Their results showed that developed nations are transitioning toward a strong decoupling status, whereas developing countries do not display any consistent pattern of decoupling. Similarly, ref. [39] compared the decoupling performance of economic growth from CO₂ emissions in Japan and China. Their findings revealed that Japan has achieved a stable decoupled position, while China has recently shifted toward decoupling. However, these results contrast with [40] whose investigation of China and India found weak decoupling in China and a volatile decoupling pattern in India from 1980 to 2014.

2.3. Income Equality and CO₂ Emission

The rise in income inequality and the worsening environmental crisis have prompted researchers to examine the relationship between these factors. Ref. [41] argued that the link between income inequality and CO₂ emission depends on income levels. Higher income inequality may result in lower per capita carbon emissions in low- and middle-income economies, whereas it may increase per capita emissions in upper-middle- and high-income economies. Similarly, ref. [42] affirmed that the income inequality–carbon emissions relationship varies across different levels of development. Their findings indicated that income inequality and emissions are positively associated in wealthy nations, negatively related in middle-income countries, and exhibit no significant correlation in low-income countries. Ref. [43] analyzed the effect of income inequality on CO₂ emissions across different quantiles and found that a rising income inequality has worsened environmental conditions in middle- and high-emission countries. In contrast, ref. [44] in their study on China and India, found no statistically significant role of income inequality in environmental degradation in either country. Regarding the time-varying relationship between income inequality and carbon emissions, ref. [45] identified a negative relationship in Asian economies during 1975–1988 and 1997–2010, but a positive relationship between 1988 and 1997.

2.4. FDI and CO₂ Emission

The empirical relationship between FDI and CO₂ emissions has been extensively studied in recent years. Ref. [46] analyzed the impact of FDI on China’s industrial CO₂ emissions during the period 2000–2017. Their findings suggest that FDI plays a positive role in reducing CO₂ emissions by promoting technological advancements and improving energy efficiency. This result aligns with the findings of [47] who examined data from 188 countries and supported the existence of the “pollution halo” hypothesis, reinforcing the role of inward FDI in lowering the carbon intensity in host countries. Conversely, ref. [48] found a positive relationship between FDI inflows and pollution emissions in Pakistan, indicating that environmental degradation worsens with an increase in FDI. Analogously, ref. [49] explored the relationship between CO₂ emissions, foreign trade openness, and FDI in India’s industrial sectors using data from 2006 to 2021. Their results, based on a threshold regression approach, partially supported the halo hypothesis by demonstrating a negative impact of FDI on CO₂ emissions. They concluded that the effect of FDI varies depending on the intensity of CO₂ emissions. Using a dynamic spatial model, ref. [50] investigated the mechanism through which FDI influences CO₂ emissions across 30 provinces in China. Their findings revealed an inverted U-shaped relationship between FDI and emissions, implying that an increase in the FDI/GDP ratio initially raises the emission levels, but beyond a certain threshold, the emissions begin to decline. On the other hand, ref. [51] examined the individual effects of FDI inflows from 11 OECD countries on the environmental quality in BRICS nations. Their study concluded that a country’s FDI inflows significantly influence the environmental degradation in the recipient countries, with the impact varying by source country. Specifically, the FDI from Denmark and the UK contributed to rising CO₂ emissions, whereas the FDI from France, Italy, and Germany helped reduce the emissions in BRICS nations.

2.5. HDI and CO₂ Emission

The Human Development Index (HDI) has been a consistent measure of social and economic development within a country. The HDI is based on three key indicators: gross domestic per capita income, enrollment ratio, and life expectancy, which collectively represent the overall standard of living. Refs. [52,53] examined the relationship between the HDI and environmental quality in Bangladesh from 1990 to 2018. Using the Tapio decoupling model, they identified a weak dependence between the HDI and CO₂ emissions. Conversely, ref. [54] reported a negative unidirectional causality between the HDI and CO₂ emissions based on a dataset of 33 OECD countries spanning the years 2006 to 2016. In the same way, ref. [55] suggested that environmental degradation negatively impacts life expectancy and human health. In the same vein, refs. [56,57] found that rising CO₂ emissions hinder the inclusive human development in 44 sub-Saharan African countries. However, ref. [58] challenged these findings, demonstrating that fuel consumption did not impede the improvements in the HDI in emerging and developed countries between 1990 and 2014. Ref. [59] identified an inverted U-shaped relationship between the HDI and three components of CO₂ emissions—carbon intensity, carbon emissions per capita, and total carbon emissions—in the southwestern provinces of China from 2001 to 2015. Their study also highlighted a strong decoupling trend between 2013 and 2015, corresponding to the historical peak of CO₂ emissions in the country.

3. Methodology

In our analysis, we developed three distinct models. In the first model, the lnGDP is the dependent variable, while the GINI, CO₂, the ECI, FDI, and the HDI are the independent variables. In the second model, we consider the GINI as the dependent variable. In the third model, CO₂ emissions serve as the dependent variable, with the lnGDP, GINI, ECI, FDI, and the HDI as independent variables. We used data from the 10 BRICS member countries, covering the period from 1995 to 2022.

To specify our first econometric model, we adopted the following functional form:

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

(1)

In the equation provided, lnGDP represents the per capita GDP, while GINI, ECI, FDI, and HDI stand for the Gini index, Economic Complexity Index, foreign direct investment, and Human Development Index, respectively. These indicators were sourced from the World Development Indicators database, our world, Harvard data verse, and Human developments reports. The variables are detailed in

Table A1. Variables, measures, and source of data.

Symbols	Variables	Measure	Definition of Variable	Source
lnGDP	GDP	Log of GDP per capita (constant 2015 USD)	GDP per capita refers to the gross domestic product divided by the population at the midpoint of the year. The data is presented in constant 2015 U.S. dollars.	Worlddevelopmentindicators
GINI	GINI Index	GINI Index	The Gini index measures the extent to which the distribution of income or consumption among individuals or households within an economy deviates from a perfectly equal distribution. A Gini index of 0 represents perfect equality, while an index of 100 implies perfect inequality.	Worlddevelopmentindicators
CO2	CO2 emissions	Log of CO2 emissions (metric tons per capita)	CO2 emissions are generated through the combustion of fossil fuels and the production of cement. These emissions encompass the release of carbon dioxide during the utilization of solid, liquid, and gaseous fuels, as well as the process of gas flaring.	Worlddevelopmentindicators
ECI	Economic Complexity Index	ECI Index	The ECI takes data on exports and reduces a country’s economic system into two dimensions: (i) The “diversity” of products in the export basket, and (ii) the “ubiquity” of products in the export basket. Diversity is the number of products that a country can export competitively. And ubiquity is the number of countries that are able to export a product competitively.	https://ourworldindata.org/grapher/economic-complexity-rankings accessed on 10 April 2025 https://dataverse.harvard.edu/dataverse/atlas accessed on 10 April 2025
FDI	Foreign direct investment	Foreign direct investment, net inflows (% of GDP)	Foreign direct investment isthe net inflows of investment to acquire a lasting management interest (10 percent or more of voting stock) in an enterprise operating in an economy other than that of the investor. It is the sum of equity capital, reinvestment of earnings, other long-term capital, and short-term capital as shown in the balance of payments. This series shows net inflows (new investment inflows less disinvestment) in the reporting economy from foreign investors, and is divided by GDP.	Worlddevelopmentindicators
HDI	The Human Development Index	HDI index	The Human Development Index (HDI) is a summary measure of the average achievement in the key dimensions of human development: a long and healthy life, being knowledgeable and having a decent standard of living. The HDI is the geometric mean of normalized indices for each of the three dimensions	https://hdr.undp.org/data-center/human-development-index#/indicies/HDI accessed on 10 April 2025

Source: authors’ calculations, WDI.

of Appendix A. Our econometric methodologies are based on the prior literature [57,60,61,62]. To enable panel data analysis, we transformed our equation for Model 1 as follows:

$$\mathrm{I}\mathrm{n}{\mathrm{G}\mathrm{D}\mathrm{P}}_{\mathrm{i}\mathrm{t}}={\mathrm{\beta }}_0+{\mathrm{\beta }}_1{\mathrm{G}\mathrm{I}\mathrm{N}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_2{{\mathrm{C}\mathrm{O}}_2}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_3{\mathrm{E}\mathrm{C}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_4{\mathrm{F}\mathrm{D}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_5{\mathrm{H}\mathrm{D}\mathrm{I}}_{\mathrm{i}\mathrm{t}}++{\mathrm{\epsilon }}_{\mathrm{i}\mathrm{t}}$$

(2)

In Model 2, we replaced lnGDP with the GINI as the dependent variable.

$${\mathrm{G}\mathrm{I}\mathrm{N}\mathrm{I}}_{\mathrm{i}\mathrm{t}}={\mathrm{\beta }}_0+{\mathrm{\beta }}_1{\mathrm{l}\mathrm{n}\mathrm{G}\mathrm{D}\mathrm{P}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_2{{\mathrm{C}\mathrm{O}}_2}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_3\mathrm{E}{\mathrm{C}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_4{\mathrm{F}\mathrm{D}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_5{\mathrm{H}\mathrm{D}\mathrm{I}}_{\mathrm{i}\mathrm{t}}++{\mathrm{\epsilon }}_{\mathrm{i}\mathrm{t}}$$

(3)

In Model 3, we considered CO₂ emissions as the dependent variable, with the remaining factors as independent variables.

$${{\mathrm{C}\mathrm{O}}_2}_{\mathrm{i}\mathrm{t}}={\mathrm{\beta }}_0+{\mathrm{\beta }}_1{\mathrm{G}\mathrm{I}\mathrm{N}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_2{\mathrm{E}\mathrm{C}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_3{\mathrm{l}\mathrm{n}\mathrm{G}\mathrm{D}\mathrm{P}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_4{\mathrm{F}\mathrm{D}\mathrm{I}}_{\mathrm{i}\mathrm{t}}+{\mathrm{\beta }}_5{\mathrm{H}\mathrm{D}\mathrm{I}}_{\mathrm{i}\mathrm{t}}++{\mathrm{\epsilon }}_{\mathrm{i}\mathrm{t}}$$

(4)

The subscripts ‘i’ (where i = 1, N) and ‘t’ (where t = 1, T) correspond to the cross-sectional and time dimensions, respectively. The parameters ${\beta }_0$ to ${\beta }_5$ indicate the slope coefficients, while ε represents the error terms.

Econometric Methods

Before conducting the co-integration analysis, we employed interpolation to address the missing data. Specifically, we used linear interpolation, which estimates missing values based on known data points from adjacent years. This method was chosen because it preserves the overall trend in the data, minimizes bias, and improves the accuracy of the analysis. By filling in the missing years through interpolation, we were able to construct a more complete and coherent dataset, thereby enabling a more robust and reliable panel analysis.

After that, we assessed the stationarity of the variables using first-generation panel unit root tests, including the Levin-Lin and Chu (LLC) test [63], the Im-Pesaran-Shin (IPS) [64] test, and the Fisher-augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) tests [65]. These tests estimate an autoregressive model with a lagged dependent variable and the first difference of the series. If the test statistic is below the critical value at the chosen significance level, the null hypothesis of a unit root is rejected, indicating that the variable is stationary. After confirming the stationarity, we moved on to identifying the cross-sectional dependence using tests such as the Breusch–Pagan test, the Pesaran scaled LM test, and the PesaranCD test. The Pesaran [66,67] approach for testing cross-sectional dependence assesses whether the variables are correlated across cross-sections. The null hypothesis of the CD tests assumes no cross-sectional dependence among the variables. With cross-sectional dependence accounted for, we proceeded to the co-integration analysis to explore the long-term relationships among the variables. We employed the Johansen Fisher Panel co-integration test, an extension of the Johansen and Juselius [68] time-series co-integration test. This test is well regarded for its reliability and consistency, producing trace statistics and maximum eigenvalues that reveal co-integration relationships, as noted by [69].The Johansen test posits a null hypothesis of no co-integration and an alternative hypothesis of co-integration among the variables.

We then proceeded to estimate the relationships outlined in Equations (2)–(4) using two advanced econometric techniques: Fully Modified Ordinary Least Squares (FMOLS) and Dynamic Ordinary Least Squares (DOLS) regression. FMOLS is designed to correct for serial correlation and endogeneity in non-stationary panels. It modifies the OLS estimator to produce asymptotically unbiased estimates of the long-term equilibrium relationship between variables. FMOLS is particularly well suited for estimating the cointegration relationships in panel data, as it adjusts for potential biases arising from non-stationary time series and endogenous regressors. FMOLS modifies the standard OLS estimator to correct for endogeneity and serial correlation in the context of cointegration. The general form of the FMOLS estimator can be expressed as:

$${\mathrm{\beta }}^{^}\mathrm{F}\mathrm{M}\mathrm{O}\mathrm{L}\mathrm{S}=\left({\sum }_{\mathrm{i}=1}^{\mathrm{n}}{{X_i}^{\prime }{X}_i}^{-1}\right)({\sum }_{\mathrm{i}=1}^{\mathrm{n}}{X}_i^{\prime }(Y_i-{\mathrm{\Omega }}^{\mathrm{i}}))$$

(5)

Here, β^FMOLS represents the FMOLS estimator of the coefficient vector. X_i denotes the matrix of independent variables for the i-th cross-sectional unit, while Y_i is the dependent variable for the i-th cross-sectional unit. The term Ω_i stands for serial correlation and endogeneity in the error terms. N refers to the number of cross-sectional units. The correction term Ω_i is typically obtained from non-parametric estimations of the long-run covariance matrix on the residuals. However, panel DOLS is particularly advantageous for panel data analysis because it accounts for both within-group and between-group variations. This allows for the estimation of parameters that are consistent across all cross-sectional units in the panel. Additionally, panel DOLS serves as an effective tool for evaluating the validity of the strict exogeneity assumption in panel data models. By addressing the potential endogeneity issues, it improves the robustness of the estimates. The coefficients derived from this dynamic model provide valuable insights into the long-term relationship between the dependent and independent variables. DOLS involves augmenting the standard OLS regression by including the leads and lags of the first differences of the independent variables to correct for endogeneity and serial correlation. The general form of the DOLS estimator can be expressed as:

$$Y_i=a_i+\beta X_{it}+\sum _{k=-q}^q\gamma k{\Delta X}_{i,t+K}+{\epsilon }_{it}$$

(6)

In the DOLS model, Y_i represents the dependent variable for the i-th cross-sectional unit at time t. The term α_i, denotes the intercept for each cross-sectional unit. The coefficient vector β is associated with the independent variables X_it. The first differences of the independent variables, ΔX_i,t+k, account for the leads and lags in the model, where k ranges from −q to +q. The coefficients for these leads and lags are represented by γk. Finally, ε_it, denotes the error term. In this dynamic model, the inclusion of ΔX_i,t+k helps address the potential endogeneity and serial correlation, leading to more robust estimates. When dealing with panel data, addressing the heterogeneity across cross-sections is essential for accurate panel causality analysis. This heterogeneity may arise from the differences in the economic policies, cultural norms, or institutional frameworks across countries. To address this, Dumitrescu and Hurlin (DH) [70] introduced a panel causality model that allows for variation in coefficients across cross-sections.

The formal expression of the D-H causality test is as follows:

$$Y_{it}=a_i+\sum _{j=1}^p{\beta }_{ij}{Y}_i{,}_{ti-j}+\sum _{j=0}^p{\gamma }_{ij}{X}_i{,}_{ti-j}+{\epsilon }_{it}\dots$$

(7)

Here, $Y_{it}$ represents the dependent variable for observation i and time period t.

X_it is the explanatory variable for observation i and time period t. α_i is the individual-specific fixed effect (intercept) for each entity in the panel. β_it and γ_it are the coefficients to be estimated, representing the short-term and long-term effects, respectively, and p is the lag order, indicating how many past periods are considered in the model. $\epsilon$_it is the error term, capturing the unexplained variability. We employ the ARDL approach for time series analysis of BRICS+ countries individually. Its versatility allows application to I(0) or I(1) variables, unlike the conventional methods. ARDL is effective for small samples and simultaneously estimates short-run dynamics and long-run equilibrium relationships, making it a robust analytical tool [71]. The general form of an Autoregressive Distributed Lag (ARDL) model for a time series can be written as:

$$Y_i=a_i+\sum _{i=1}^p{\beta }_i{Y_{i-p}}_t+\sum _{j=0}^p{\gamma }_i{X}_i{,}_{i-j}+{\epsilon }_t\dots$$

(8)

For time series data analysis, we reformulated our equation for Model 1 as follows:

$$\begin{gathered} {lnGDP}_t=a_0+\sum _{p=1}^p{a}_1{lnGDP}_{t-p}+\sum _{q=0}^{Q_1}{\beta }_1{GINI}_{t-q}+\sum _{q=0}^{Q_2}{\beta }_2{CO}_{2t-q}+\sum _{q=0}^{Q_3}{\beta }_3{ECI}_{t-q} \\ +{\sum _{q=0}^{Q_4}{\beta }_4\mathrm{F}\mathrm{D}{\mathrm{I}}_{t-p}+\sum _{q=0}^{Q_5}{\beta }_5{HDI}_{t-q}}_{it}+{\epsilon }_t\dots \end{gathered}$$

(9)

The subscript t′ (where t = 1, T) represents the time dimensions, respectively. p is the maximum lag order for the dependent variable, lnGDP. The parameters ${\beta }_0$ to ${\beta }_5$ indicatetheslope coefficients, while ε represents the error terms. Q₁ to Q₅ represent the lag orders for the independent variables. For further analysis, we modified this equation by replacing lnGDP with GINI and introducing the lag of GINI to formulate Model 2.

$$\begin{gathered} {GINI}_t=a_0+\sum _{p=1}^p{a}_1{GINI}_{t-p}+\sum _{q=0}^{Q_1}{\beta }_1\mathrm{l}\mathrm{n}{GDP}_{t-q}+\sum _{q=0}^{Q_2}{\beta }_2{CO}_{2t-q}+\sum _{q=0}^{Q_3}{\beta }_3{ECI}_{t-q} \\ +{\sum _{q=0}^{Q_4}{\beta }_4\mathrm{F}\mathrm{D}{\mathrm{I}}_{t-p}+\sum _{q=0}^{Q_5}{\beta }_5{HDI}_{t-q}}_{it}+{\epsilon }_t\dots \end{gathered}$$

(10)

In Model 3, CO₂ emissions were designated as the dependent variable, while the remaining factors were retained as the independent variables.

$$\begin{gathered} {CO}_{2t}=a_0+\sum _{p=1}^p{a}_1{CO}_{2t-p}+\sum _{q=0}^{Q_1}{\beta }_1{GINI}_{t-q}+\sum _{q=0}^{Q_2}{\beta }_2\mathrm{l}\mathrm{n}\mathrm{G}\mathrm{D}{P}_{t-q}+\sum _{q=0}^{Q_3}{\beta }_3{ECI}_{t-q}+{\sum _{q=0}^{Q_4}{\beta }_4\mathrm{F}\mathrm{D}{\mathrm{I}}_{t-p}+\sum _{q=0}^{Q_5}{\beta }_5{HDI}_{t-q}}_{it} \\ +{\epsilon }_t\dots \end{gathered}$$

(11)

This formulation is suitable for estimating the short-run and long-run relationships in a time-series setting using the ARDL approach. This formulation captures both short-run and long-run dynamics in the ARDL framework. In the last stage, we used the canonical cointegration regression for a robustness check.

4. Results and Discussion

Table A2. Descriptive statistics.

Variables	lnGDP	GINI	CO2	ECI	FDI	HDI
Mean	8.50	0.611	7.27	0.005	2.14	0.68
Std.Dev.	1.29	0.060	7.17	0.60	1.83	0.134
Maximum	11.06	0.74	30.88	1.94	9.64	0.937
Minimum	5.48	0.47	0.044	−1.32	−1.78	0.28
Observations	280	280	280	280	280	280

Source: calculated by authors. Note: lnGDP represents the logarithm of Gross Domestic Product, GINI refers to the Gini index (a measure of income inequality), CO₂ denotes carbon dioxide emissions, ECI stands for the Economic Complexity Index, FDI indicates foreign direct investment, and HDI signifies the value of the Human Development Index.

in Appendix A presents the descriptive statistics for all the variables included in this study, while

Table 1. Correlation matrix.

Variables	lnGDP	GINI	CO2	ECI	FDI	HDI
lnGDP	1.00
GINI	0.526	1.000
CO2	0.836	0.305	1.000
ECI	0.334	0.08	0.007	1.000
FDI	0.004	−0.061	−0.037	0.046	1.000
HDI	0.74	0.404	0.708	0.364	−0.038	1.000
VIF		1.205	2.178	1.262	1.008	2.71
1/VIF		0.83	0.459	0.792	0.992	0.368

Source: authors’ calculations. Note: lnGDP represents the logarithm of Gross Domestic Product, GINI refers to the Gini index (a measure of income inequality), CO₂ denotes carbon dioxide emissions, ECI stands for the Economic Complexity Index, FDI indicates foreign direct investment, and HDI signifies the value of the Human Development Index.

highlights the key relationships among them. The lnGDP has strong positive correlations with CO₂ emissions and the HDI, indicating that a higher GDP is closely associated with increased CO₂ emission and improved human development. Moderate positive correlations are also observed between the lnGDP, GINI, and ECI, suggesting that economic growth tends to be accompanied by rising income inequality and increased economic complexity. In contrast, the correlation between the lnGDP and FDI is very weak, indicating little to no relationship. The GINI shows weak positive correlations with both CO₂ emissions and the HDI, implying that higher income inequality is somewhat related to increased emissions and human development.

The relationship between the GINI and FDI is weakly negative, suggesting a slight tendency for higher income inequality to be associated with lower foreign investment. CO₂ emissions exhibit a strong positive correlation with the HDI but show almost no relationship with the ECI or FDI. This implies that higher emissions are linked to improved human development but are not necessarily associated with economic complexity or foreign investment. The ECI has a moderate positive correlation with the HDI, indicating that more complex economies tend to achieve higher levels of human development. However, its correlation with FDI is very weak. Overall, FDI demonstrates minimal correlations with all the other variables, suggesting limited linear relationships with GDP, income inequality, emissions, economic complexity, and human development in this dataset. Additionally, the table includes Variance Inflation Factors (VIFs) and their reciprocals (1/VIF). The VIF quantifies the degree of multicollinearity among the independent variables, with higher values indicating greater multicollinearity [72,73]. The reciprocal of the VIF (1/VIF) represents the tolerance, where lower values suggest potential multicollinearity issues.

The unit root test results presented in

Table 2. Results of unit root test.

	LLC		IPS		ADF-Fisher		PP Fisher
Series	I(0)	I(1)	I(0)	I(1)	I(0)	I(1)	I(0)	I(1)
lnGDP	−1.73 (0.041)	−4.1 (0.000)	1.58 (0.944)	−4.53 (0.000)	10.88 (0.949)	57.97 (0.000)	10.37 (0.960)	86.63 (0.000)
GINI	−3.20 (0.001)	−3.5 (0.000)	−3.08 (0.944)	5.39 (0.000)	48.24 (0.000)	70.86 (0.000)	49.82 (0.000)	99.22 (0.000)
CO2	−0.520 (0.301)	−2.76 (0.000)	1.10 (0.864)	−3.23 (0.000)	12.32 (0.904)	59.50 (0.000)	12.16 (0.910)	132.68 (0.000)
ECI	0.96 (0.831)	−5.110 (0.000)	1.43 (0.924)	−10.23 (0.000)	9.29 (0.979)	130.79 (0.000)	17.15 (0.643)	215.01 (0.000)
FDI	−1.58 (0.055)	−6.888 (0.000)	−2.85 (0.002)	−8.78 (0.000)	40.33 (0.000)	109.6 (0.000)	46.59 (0.000)	179.5 (0.000)
HDI	−2.40 (0.002)	−0.646 (0.000)	1.83 (0.966)	−3.79 (0.000)	8.92 (0.983)	48.74 (0.000)	14.72 (0.792)	49.540 (0.000)

Source: authors’ calculations. Note: lnGDP represents the logarithm of Gross Domestic Product, GINI refers to the Gini index, CO₂ denotes carbon dioxide emissions, ECI stands for the Economic Complexity Index, FDI indicates foreign direct investment, and HDI signifies the value of the Human Development Index. The value in parentheses indicates the p-value.

evaluate the stationarity of the series using several established methods: Levin, Lin, and Chu (LLC), Im, Pesaran, and Shin (IPS), Augmented Dickey–Fuller (ADF-Fisher), and Phillips–Perron (PP-Fisher). These tests are performed at two levels: I(0), which assesses the series in its original form, and I(1), which examines the series after first differencing. For all these tests, the null hypothesis asserts that the series has a unit root, indicating non-stationarity. If the p-value (reported in parentheses) is below a conventional significance threshold (typically 0.05), the null hypothesis is rejected, signifying that the series is stationary at that level. The results indicate that most variables, including the lnGDP, CO₂, ECI, and HDI, are non-stationary in their original form but attain stationarity after first differencing, suggesting that they are integrated of order one, I(1). In contrast, the GINI coefficient is stationary at its original level, while FDI is nearly stationary at levels but achieves definitive stationarity following first differencing.

Cross-sectional dependence arises when cross-sectional units, such as different countries, are not independent of one another. If not properly accounted for, this dependence can introduce biases in panel data estimations, potentially leading to misleading inferences. The null hypothesis for cross-sectional dependence tests posits that no cross-sectional dependence exists within the panel data. The Breusch–Pagan LM test, with a p-value of 0.000, strongly rejects this null hypothesis, indicating the presence of cross-sectional dependence (

Table 3. Cross-section dependence test.

Test	Statistics	Prob.
Breusch–Pagan LM	400.54	0.000
Pesaran Scaled LM	54.42	0.000
Pesaran CD	9.28	0.000

Source: calculated by authors.

). Likewise, the Pesaran Scaled LM test and the Pesaran CD test yield consistent results, further corroborate this finding. Since all three tests reject the null hypothesis, it confirms the existence of cross-sectional dependence in the dataset.

Table 4. Findings of Johansen Fisher Panel Co-integration Test.

Hypothesized No. of CE(s)	Fisher Stat. from Trace Test	Fisher Stat. from Max Test
None	282.7 (0.000)	135.8 (0.000)
At Most 1	171.3 (0.000)	87.40 (0.000)
At Most 2	101 (0.000)	51.63 (0.000)
At Most 3	63.88 (0.000)	34.07 (0.000)
At Most 4	49.16 (0.000)	35.34 (0.000)
At Most 5	46.70 (0.000)	46.70 (0.000)
Kao residuals Co-integration test
ADF	T-statistics	p-Value
	−1.55	0.005
Residuals variance	0.0017
HAC variance	0.0024

Source: calculated by authors.

displays the results of the Johansen Fisher Panel Co-integration Test and the Kao Residuals Co-integration Test. The Johansen Fisher test examines the existence of co-integration among the variables across different hypothesized numbers of co-integrating equations (CEs). The test statistics, from both the Trace and Max tests, are accompanied by their respective p-values in parentheses. At all levels (None, At Most 1, At Most 2, etc.), the Fisher statistics are highly significant, as indicated by the p-values of 0.000. This strong significance indicates the presence of co-integration relationships among the variables, suggesting a long-term equilibrium connection between them. The Kao Residuals Co-integration Test reinforces these findings. The ADF T-statistic is negative and significant, with a p-value of 0.005, leading to the rejection of the null hypothesis of no co-integration. Additionally, the low residual variance and HAC variance further validate the stability of the model. Overall, the findings suggest that the variables in the panel maintain a stable long-run relationship, reinforcing the robustness of the model’s co-integration structure.

The results of the FMOLS and DOLS estimations, derived from Equation (2) using Equations (5) and (6) for BRICS countries, provide valuable insights into the long-term relationships, as presented in

Table 5. Result of regression model 1.

	FMOLS		DOLS
Variables	Coefficient	t-stat	Coefficient	t-stat
GINI	−1.07	−0.11	−3.67	−8.50 ***
CO2	0.50 ***	47.02	0.60	28.0 ***
ECI	−0.04 ***	−15.81	−0.16	−5.02 ***
FDI	0.001 ***	3.58	0.01	9.02 ***
HDI	3.19 ***	78.85	3.15	56.35 ***

Source: authors’ calculations. Note: lnGDP represents the logarithm of Gross Domestic Product, GINI refers to the Gini index, CO₂ denotes carbon dioxide emissions, ECI stands for the Economic Complexity Index, FDI indicates foreign direct investment, and HDI signifies the value of the Human Development Index. Dependent variable: log of GDP. ***: significant at the 1% level (t-statistic > 2.576 or <−2.576).

. In the FMOLS model, the GINI coefficient is negative (−1.07), but its t-statistic is not statistically significant, indicating that income inequality does not exert a strong impact on the GDP. Conversely, the coefficients for CO₂ (0.50), the ECI (−0.04), FDI (0.001), and the HDI (3.19) are all statistically significant at the 1% level, as evidenced by their high t-statistics. These results suggest that CO₂ emissions positively influence economic outcomes, whereas the ECI has a negative effect. FDI and the HDI also contribute positively to economic growth, with the HDI exhibiting a particularly strong impact. The DOLS model, which incorporates leads and lags of the first-differenced independent variables, accounts for the potential endogeneity and serial correlation that may not be fully captured in the FMOLS model. This approach enhances the flexibility and improves the estimation of the dynamic relationships in the data. The DOLS results largely align with those of FMOLS, albeit with some variations. Notably, the GINI coefficient (−3.67) is negative and statistically significant, with a t-statistic of −8.50, indicating that income inequality has an adverse effect on the economic growth in BRICS countries when additional lags and leads are considered. These findings are consistent with the studies of [74,75,76,77]. The CO₂ coefficient (0.60) remains positive and significant, further reinforcing the positive relationship between CO₂ emissions and economic performance. This corroborates the findings of [78,79,80,81,82]. Likewise, the ECI coefficient (−0.16) continues to exhibit a negative and significant effect, aligning with the FMOLS results. The coefficients for FDI (0.01) and the HDI (3.15) remain positive and highly significant, underscoring their strong contributions to economic growth, in line with the studies of [83,84,85,86].

Table 6. Findings from the ARDL model for BRICS countries.

Country	L1.LGDP	L1.GINI	GINI	L1.CO2	CO2	L1.ECI	ECI	L1.FDI	FDI	L1.HDI	HDI
Brazil
Coefficient	0.74 ***	−0.56	1.37	0.16 **	−0.19 ***	0.02	−0.09 *	0.00	0.00	0.89	−0.08
t-Value	(4.84)	(−0.81)	(1.94)	(2.77)	(−4.28)	(0.47)	(−2.11)	(0.14)	(−0.33)	(1.06)	(−0.07)
Russia
Coefficient	0.62 ***	−0.13	0.25	0.08 ***	0.001	0.05	0.06	0.01	0.01	2.63	−0.47
t-Value	(4.15)	(−0.31)	(0.72)	(3.39)	(0.12)	(0.82)	(1.31)	(1.67)	(0.98)	(1.90)	(−0.41)
India
Coefficient	0.84 ***	−1.76	0.56	0.30 ***	0.30 ***	−0.06	0.06	−0.01	0.01	4.76 ***	−2.87 **
t-Value	(5.30)	(−1.25)	(0.32)	(4.80)	(3.56)	(−0.95)	(1.09)	(−0.79)	(1.16)	(4.01)	(−2.19)
China
Coefficient	0.80 *	0.02	0.53	0.01	0.02	−0.13 *	−0.06	0.01	0.00	4.16 *	2.49
t-Value	(2.11)	(0.04)	(0.94)	(0.34)	(0.68)	(−2.22)	(−0.79)	(0.66)	(−0.28)	(2.27)	(0.70)
South Africa
Coefficient	0.59 **	0.96 *	−0.27	0.03 *	−0.01	−0.02	0.02 *	0.001	0.00	1.18	−1.29
t-Value	(2.73)	(2.16)	(−0.59)	(2.51)	(−0.55)	(−1.28)	(2.16)	(0.93)	(0.39)	(1.10)	(−1.19)
Egypt
Coefficient	0.91 ***	−0.67	0.41	−0.02	−0.02	0.06	0.07	0.01 *	0.00	−0.38	0.64
t-Value	(6.48)	(−0.87)	(0.69)	(−0.56)	(−0.60)	(0.88)	(1.27)	(2.14)	(0.81)	(−0.28)	(0.41)
Ethiopia
Coefficient	0.58 **	2.35	−1.60	2.01	−0.52	0.04	−0.05	−0.01	0.00	1.57	0.10
t-Value	(2.94)	(1.84)	(−1.93)	(1.34)	(−0.37)	(0.55)	(−0.72)	(−1.04)	(−0.39)	(1.58)	(0.07)
Iran
Coefficient	0.83 ***	−0.16	1.00	−0.01	0.01	0.02	0.02	0.02	0.01	3.00	−2.68 **
t-Value	(3.45)	(−0.10)	(0.58)	(−0.79)	(0.84)	(0.32)	(0.45)	(0.86)	(0.39)	(1.22)	(−1.26)
Saudi Arabia
Coefficient	0.79 ***	−5.64	6.80	0.00	−0.01	−0.02	0.01	0.00	−0.01	5.96 *	−5.09
t-Value	(3.12)	(−0.74)	(0.98)	(−0.24)	(−1.22)	(−0.92)	(0.27)	(−0.20)	(−0.86)	(2.32)	(−1.90)
United Arab Emirates
Coefficient	0.80 ***	−4.03	2.28	0.01 *	0.00	−0.10	−0.05	0.02 **	−0.01	1.83	−2.34 *
t-Value	(6.10)	(−1.38)	(0.74)	(2.17)	(−0.03)	(−1.26)	(−0.54)	(2.65)	(−1.20)	(1.14)	(−1.15)

Note: LGDP indicates log of GDP growth and L1.LGDPindicates its one-period lag, representing economic expansion and its past effects. GINI measures income inequality, while L1.GINI captures its lagged impact. CO₂ represents current carbon emissions, with L1.CO₂ accounts for past emissions’ persistence. ECI reflects a country’s economic complexity, and L1.ECI considers its delayed effects. FDI denotes foreign direct investment, with L1.FDI indicating its lagged influence. And HDI assesses human development, while L1.HDI captures its previous period’s effect on sustainability and emissions. *** Significant at the 1% level (t-statistic > 2.576 or <−2.576), **: significant at the 5% level (t-statistic > 1.96 or <−1.96), *: significant at the 10% level (t-statistic > 1.645 or <−1.645).

presents the results of the ARDL model for BRICS+ economies, incorporating lag variables from Model 1, estimated using Equation (9).

This reveals that in Brazil, the LGDP demonstrates strong persistence (0.74), reflecting economic inertia fueled by sustained investment and productivity growth. Current CO₂ emissions (0.16) boost GDP, reflecting the reliance on carbon-intensive industries, while lagged CO₂ emissions negatively impact growth, likely due to environmental or regulatory costs. The negative effect of lagged economic complexity suggests inefficiencies or adjustment costs. The results align with the research of [83,84,85,86]. Russia shows a similar momentum (0.62), with CO₂ emissions (0.08) supporting its GDP, highlighting the dependence on energy-intensive sectors. India’s strong LGDP persistence (0.84) reflects sustained economic activity, with current CO₂ emissions (0.30) fueling industrial expansion. The results validate the findings of [87,88]. HDI improvements (4.76) significantly enhance the GDP, while a lagged HDI (−2.87) suggests diminishing returns or resource constraints from past investments. In China, the LGDP (0.80) remains a key driver, but lagged economic complexity (−0.13) indicates inefficiencies from over-specialization, and the findings substantiate the work of [89]. A lagged HDI (4.16) supports growth, reinforcing the role of human development. South Africa exhibits economic persistence (0.59), with lagged income inequality (0.96) potentially driving the wealth concentration into productive investments. Lagged CO₂ emissions (0.03) and economic complexity (0.02) also contribute positively to GDP. Egypt’s lagged LGDP (0.91) signals strong momentum, while FDI (0.01) supports growth, emphasizing foreign capital’s role in infrastructure and exports. Ethiopia follows a simpler growth model, with a lagged LGDP (0.58) as its primary driver. In Iran, the LGDP (0.83) reflects the sustained potential, though a lagged HDI (−2.68) suggests that past human capital investments may have strained resources. Saudi Arabia’s LGDP (0.79) highlights oil-driven growth, while the lagged HDI (5.96) underscores the long-term benefits of education and healthcare investments. Finally, in the United Arab Emirates, the LGDP (0.80) remains a key growth driver, with current CO₂ emissions (0.01) reinforcing energy sector dominance. FDI (0.02) further boosts expansion, though the current HDI (−2.34) suggests inefficiencies or saturation from past human development gains.

Table 7. Result of regression model 2, dependent variable (GINI).

	FMOLS		DOLS
Variables	Coefficient	t-stat.	Coefficient	t-stat.
lnGDP	−0.01	−1.33	−0.09 ***	3.27
CO2	−0.02	−1.29	−0.08 ***	−14.46
ECI	−0.01 ***	−6.24	−0.03 ***	−2.95
FDI	0.001 ***	14.56	0.001 ***	6.56
HDI	0.33 ***	9.23	1.27 ***	6.55

Source: authors’ calculations. Note: lnGDP represents the logarithm of Gross Domestic Product, GINI refers to the Gini index, CO₂ denotes carbon dioxide emissions, ECI stands for the Economic Complexity Index, FDI indicates foreign direct investment, and HDI signifies the value of the Human Development Index. Dependent variable: GINI index. ***: significant at the 1% level (t-statistic > 2.576 or <−2.576).

presents the results of Model 2, based on Equation (3), utilizing Equations (5) and (6) to examine the impact of independent variables on the Gini index. The analysis reveals that economic growth is inversely related to income inequality, as indicated by the coefficient, aligning with the findings of [90]. CO₂ also exhibits a negative impact on income inequality, which contradicts the findings of [84,91,92]. Additionally, the ECI is negatively associated with income inequality, consistent with the finding of [93]. The results indicate that FDI is positively associated with income inequality, contradicting the finding of [94] but aligning with those of [52,95,96]. FDI through multinational corporations (MNCs) can increase the income inequality despite facilitating technology transfer and skill development [97]. Since MNCs typically offer higher wages, they primarily benefit skilled workers, widening the wage gap between skilled and unskilled labor. Additionally, MNCs may cluster in urban or more developed regions, leading to uneven economic development and regional disparities. As a result, while FDI boosts the overall productivity and innovation, it can also exacerbate inequality by disproportionately benefiting certain workers and regions over others. Lastly, the results suggest that the HDI exacerbates the income inequality, which contradicts the findings of [98].

Table 8. Findings from the ARDL model for BRICS countries.

Variables	L1.GINI	LGDP	L1.GDP	CO2	L1.CO2	ECI	L1.ECI	FDI	L1.FDI	HDI	L1.HDI
Brazil
Coefficient	−0.01	−0.08	0.05	−0.02	0.002	0.02	−0.01	0.00	0.00	−0.20	0.41
t-Value	(−0.04)	(−0.81)	(0.51)	(−0.74)	(0.06)	(1.24)	(−0.53)	(−0.36)	(0.47)	(−0.60)	(0.99)
Russia
Coefficient	−0.05	−0.05	0.05	0.02	0.03	0.03	0.02	0.00	0.00	−0.93	0.36
t-Value	(−0.22)	(−0.31)	(0.39)	(1.11)	(1.92)	(0.92)	(0.78)	(0.43)	(0.05)	(−1.02)	(0.52)
India
Coefficient	1.08 **	−0.05	0.08 *	0.01	−0.05 **	0.00	0.02	0.00	0.00	0.62 *	−0.65 **
t-Value	(7.88)	(−1.25)	(1.88)	(0.65)	(−3.29)	(−0.07)	(1.70)	(0.11)	(0.87)	(2.44)	(−3.21)
China
Coefficient	0.54 *	0.01	−0.23	0.01	−0.02	0.03	0.04	0.00	0.00	−0.12	2.25
t-Value	(1.90)	(0.04)	(−1.02)	(0.45)	(−1.28)	(0.72)	(0.96)	(0.72)	(0.12)	(−0.11)	(1.19)
South Africa
Coefficient	0.59 **	0.25 *	−0.02	−0.01	−0.01	0.003	0.009	0.001	0.002	−0.27	0.25
t-Value	(3.18)	(2.16)	(−0.12)	(−1.04)	(−0.80)	(0.55)	(−0.60)	(−1.01)	(−0.58)	(−0.48)	(0.44)
Egypt
Coefficient	0.71 **	−0.07	0.02	0.00	−0.01	0.05 **	0.01	0.00	0.00	−0.15	0.20
t-Value	(9.50)	(−0.87)	(0.26)	(0.10)	(−0.64)	(2.63)	(0.67)	(0.92)	(−0.78)	(−0.34)	(0.38)
Ethiopia
Coefficient	0.63 **	0.08*	−0.04	0.00	0.07	−0.01	0.00	0.00	0.00	0.23	−0.44
t-Value	(14.78)	(1.84)	(−0.79)	(0.01)	(0.28)	(−0.38)	(−0.01)	(−0.07)	(−1.40)	(1.22)	(−1.82)
Iran
Coefficient	0.88 **	0.00	0.01	0.001	0.002	0.01	0.01 **	0.005	0.003	−0.10	−0.08
t-Value	(6.18)	(−0.10)	(0.11)	(0.55)	(0.52)	(1.42)	(2.04)	(0.37)	(−0.13)	(−0.25)	(−0.22)
Saudi Arabia
Coefficient	0.52 *	−0.01	−0.01	0.002	0.004	0.007	0.003	0.004 **	0.091	−0.05	0.02
t-Value	(2.64)	(−0.74)	(−0.91)	(1.73)	(−1.61)	(−1.18)	(1.07)	(3.91)	(−0.80)	(−0.46)	(0.20)
United Arab Emirates
Coefficient	0.96 **	−0.03	0.04 **	0.002	0.003	−0.01	−0.02 **	0.002	0.004	−0.09	0.25
t-Value	(12.87)	(−1.38)	(2.60)	(0.76)	(0.67)	(−1.87)	(−2.78)	(0.00)	(0.43)	(−0.67)	(1.55)

Note: Gini index is treated as the dependent variable, LGDP indicates log of GDP growth, and L1.LGDPindicatesits one-period lag, representing economic expansion and its past effects. GINI measures income inequality, while L1.GINI captures its lagged impact. CO₂ represents current carbon emissions, with L1.CO₂ accounting for past emissions’ persistence. ECI reflects a country’s economic complexity, and L1.ECI considers its delayed effects. FDI denotes foreign direct investment, with L1.FDI indicating its lagged influence. And HDI assesses human development, while L1.HDI captures its previous period’s effect on sustainability and emissions. Significant at the 1% level (t-statistic > 2.576 or <−2.576), **: significant at the 5% level (t-statistic > 1.96 or <−1.96), *: significant at the 10% level (t-statistic > 1.645 or <−1.645).

illustrates the impact of various variables on the Gini index across the BRICS countries, based on Model 2 and Equation (10). In India, lagged inequality (1.08) exhibits strong persistence, while economic growth (0.08) weakly exacerbates the disparities. The coefficient of lagged CO₂ emissions (−0.05) contributes to reducing the inequality, whereas human development has mixed effects (0.62, −0.65). In China, past inequality (0.54) remains a primary determinant of the current disparities. In South Africa, income inequality persists (0.59), with economic growth (0.25) having a weak but significant impact. Egypt demonstrates strong inequality persistence (0.71), while rising economic complexity (0.05) appears to widen disparities. In Ethiopia, the coefficient of GINI (0.63) remains entrenched, and economic growth (0.08) has a marginal influence on the disparities. For Iran, the past coefficient of GINI (0.88) is a major driver of the current disparities, while economic complexity (0.01) further exacerbates the inequality. In Saudi Arabia, a lagged coefficient of GINI (0.52) continues to drive the income disparities, with foreign direct investment (0.004) also playing a role. The UAE exhibits deeply rooted inequality (0.96), with economic growth (0.04) intensifying the disparities, although lagged economic complexity (−0.02) contributes to reducing the inequality.

Table 9. Result of regression model 3, dependent variable (CO₂).

	FMOLS		DOLS
Variables	Coefficient	t-value	Coefficient	t-value
lnGDP	4.80 ***	46.10	4.85 ***	10.11
GINI	29.79	1.40	51.12 ***	4.49
ECI	−0.12 ***	−11.94	−0.11 ***	−14.73
FDI	0.02 ***	11.32	0.03 ***	9.19
HDI	−2.92 ***	−18.00	−6.12 ***	−5.01

Source: authors’ calculations. Note: lnGDP represents the logarithm of Gross Domestic Product, GINI refers to the Gini index (a measure of income inequality), CO₂ denotes carbon dioxide emissions, ECI stands for the Economic Complexity Index, FDI indicates foreign direct investment, and HDI signifies the value of the Human Development Index. Dependent variable: CO₂ emission, ***: significant at the 1% level (t-statistic > 2.576 or <−2.576).

presents the results of the regression for Equation (4) (Model 3) by utilizing Equations (5) and (6), where the dependent variable is CO₂ emissions, using both FMOLS and DOLS estimation techniques. The results indicate several significant relationships between the variables and CO₂. For GDP, both the FMOLS and DOLS models show a strong positive impact, with coefficients of 4.80 and 4.85, respectively. This suggests that economic growth is positively associated with higher CO₂, reflecting the environmental impact of increased economic activity. This finding supports the studies such as [99,100,101]. The GINI coefficient shows a positive relationship with CO₂, but it is only statistically significant in the DOLS model, suggesting that income inequality may contribute to environmental degradation. The findings support the study of [102] which concluded that as the disparity between rich and poor grows, wealthier people prioritize their economic gains over environmental stewardship. In the FMOLS model, the coefficient is negative (−0.12), indicating a reduction in CO₂ with a higher ECI. However, in the DOLS model, the ECI has a positive and significant effect on CO₂ suggesting that the ECI of an economy may sometimes lead to higher emissions. This result is inconsistent with the recent studies of [103,104], which linked the promotion of the role of economic complexity in tackling the climate crisis. Further, FDI also shows a positive relationship with CO₂ emissions in both models, with coefficients of 0.02 (FMOLS) and 0.03 (DOLS), indicating that foreign direct investment may contribute to environmental pollution, thus endorsing the results of [105,106]. The findings contradict the Halo Effect hypothesis which posits FDI as a mean to bring more energy-efficient, advanced, and cleaner technology into the country. The positive relationship can be explained by the insufficient environment regulations in the host country to protect the depletion of their environment. Finally, the HDI has a negative impact on CO₂ in both models, with coefficients of −2.92 (FMOLS) and −6.12 (DOLS), implying that higher levels of the HDI are associated with lower CO₂, possibly due to better environmental practices and awareness [107,108,109]. On the contrary, some prior empirical studies endorsed the viewpoint that HDI progress is responsible for deteriorating environmental performance due to increased consumption and industrial activities [53,60,110]. These results underscore the complex interplay between economic, social, and environmental factors in determining CO₂ emissions across different economies.

Table 10. Findings from the ARDL model for BRICS countries.

Country	L1.CO2	LGDP	L1.GDP	GINI	L1.GINI	ECI	L1.ECI	FDI	L1.FDI	HDI	L1.HDI
Brazil
Coefficient	0.77 ***	2.28 **	−1.29	−1.97	−4.66	0.12	0.39 **	0.02	0.00	−5.60 *	5.98
t-Value	(4.91)	(2.77)	(−1.43)	(−0.74)	(−1.68)	(0.85)	(2.72)	(1.06)	(−0.12)	(−1.86)	(1.49)
Russia
Coefficient	0.16	5.73 ***	−3.10 *	3.93	−3.47	−0.09	−0.14	−0.01	−0.05	0.27	−9.15
t-Value	(0.68)	(3.39)	(−1.79)	(1.11)	(−1.15)	(−0.17)	(−0.33)	(−0.18)	(−0.75)	(0.02)	(−0.93)
India
Coefficient	0.75 ***	2.01 ***	−1.68 ***	2.45	2.15	0.37 **	−0.20	−0.01	0.00	−10.69 ***	6.28 *
t-Value	(3.46)	(4.80)	(−3.11)	(0.65)	(0.48)	(2.44)	(−1.55)	(−0.52)	(−0.08)	(−3.12)	(1.78)
China
Coefficient	0.61 ***	0.82	7.53 **	2.02	11.28 **	−0.54	−0.16	0.18 ***	−0.12 *	3.83	−65.56 **
t-Value	(2.90)	(0.34)	(2.10)	(0.45)	(2.45)	(−0.86)	(−0.22)	(2.61)	(−1.83)	(0.19)	(−2.21)
South Africa
Coefficient	0.24	8.58 **	−3.45	−8.10	5.53	0.29 *	−0.38 **	0.02	−0.07 **	12.14	−18.14
t-Value	(0.92)	(2.51)	(−0.85)	(−1.04)	(0.76)	(1.56)	(−2.09)	(0.58)	(−2.20)	(0.70)	(−1.05)
Egypt
Coefficient	0.30	−1.01	3.11 *	0.58	−0.08	0.69 *	−0.11	0.04 **	−0.02 *	7.34	−13.91
t-Value	(1.13)	(−0.56)	(1.80)	(0.10)	(−0.02)	(1.40)	(−0.27)	(2.25)	(−1.53)	(0.80)	(−1.34)
Ethiopia
Coefficient	0.56 ***	0.05	0.01	0.00	0.00	−0.01	0.00	0.01 *	0.00	−0.08	−0.07
t-Value	(3.17)	(1.34)	(0.31)	(0.01)	(−0.02)	(−0.87)	(0.04)	(1.58)	(0.27)	(−0.43)	(−0.31)
Iran
Coefficient	0.71 **	−4.05	2.01	18.24	−24.44	−0.92	−0.50	0.10	0.33	43.52	−31.01
t-Value	(2.39)	(−0.79)	(0.31)	(0.55)	(−0.71)	(−0.97)	(−0.53)	(0.27)	(0.86)	(0.85)	(−0.70)
Saudi Arabia
Coefficient	0.79 ***	−2.41	21.68 *	47.87 *	122.18	0.63	−0.94	−0.63 ***	0.01	72.07	−40.77
t-Value	(4.01)	(−0.24)	(1.91)	(1.73)	(0.44)	(0.63)	(−1.00)	(−2.64)	(0.06)	(0.63)	(−0.35)
United Arab Emirates
Coefficient	0.69 **	17.33 **	−15.55 **	82.42	−91.61	3.54 *	0.81	−0.20	0.18	−88.35 *	66.58
t-Value	(2.31)	(2.17)	(−2.02)	(0.76)	(−0.85)	(1.32)	(0.28)	(−0.68)	(0.73)	(−1.61)	(0.91)

Note: CO₂ emission treated as the dependent variable, LGDP indicates log of GDP growth and L1.LGDPindicates its one-period lag, representing economic expansion and its past effects. GINI measures income inequality, while L1.GINI captures its lagged impact. CO₂ represents current carbon emissions, with L1.CO₂ accounting for past emissions’ persistence. ECI reflects a country’s economic complexity, and L1.ECI considers its delayed effects. FDI denotes foreign direct investment, with L1.FDI indicating its lagged influence. And HDI assesses human development, while L1.HDI captures its previous period’s effect on sustainability and emissions. *** Significant at the 1% level (t-statistic > 2.576 or <−2.576), **: significant at the 5% level (t-statistic > 1.96 or <−1.96), *: significant at the 10% level (t-statistic > 1.645 or <−1.645).

presents the ARDL results, highlighting the determinants of CO₂ emissions across different countries using Model 3 and Equation (11). In Brazil, past CO₂ emissions (0.77) exhibit strong persistence, while GDP growth (2.28) significantly increases emissions, reinforcing the link between economic expansion and carbon-intensive activities. Economic complexity (0.39) has a delayed positive effect, indicating the long-term environmental costs of industrial structures. However, improvements in human development (−5.60) mitigate emissions, reflecting the role of education and healthcare in sustainability. For Russia, past emissions are statistically insignificant, but the current GDP growth (5.73) drives emissions, while lagged GDP (−3.10) reduces them, possibly due to structural adjustments. In India, past emissions (0.75) remain a major determinant, with economic complexity (0.37) increases emissions, suggesting the environmental costs of industrial diversification. Human development (−10.69) strongly reduces emissions, though its lagged effect (6.28) suggests an initial rise before sustainability measures take effect. In China, past emissions (0.61) significantly influence the current levels. While GDP has a minor positive effect, lagged GDP (7.53) increases emissions, reflecting long-term industrial impacts. Past income inequality (11.28) contributes to higher emissions, likely due to consumption patterns. FDI (0.18) raises emissions, while its lagged effect (−0.12) indicates a shift toward cleaner technologies. The lagged HDI (−65.56) strongly mitigates emissions, emphasizing the role of social investments. For South Africa, GDP growth (8.58) significantly raises emissions, highlighting the environmental costs of rapid expansion. However, lagged economic complexity (−0.38) suggests earlier industrial efficiency gains. Lagged FDI (−0.07) also indicates a gradual transition toward sustainability. Egypt’s emissions are not strongly driven by past levels, but lagged GDP (3.11) increases emissions, while economic complexity (0.69) and FDI (0.04) contribute positively. Lagged FDI (−0.02) suggests a delayed shift toward cleaner technologies. In Ethiopia, past CO₂ emissions (0.56) persist as a key driver, while other economic factors have a minimal influence, indicating that emissions are shaped more by historical trends than current dynamics. Iran’s emissions (0.71) also show strong persistence, with rising human development initially increasing energy consumption before transitioning toward sustainability. For Saudi Arabia, past emissions (0.79) and lagged GDP growth (21.68) drive emissions, reflecting a carbon-intensive economy. High income inequality (47.87) further exacerbates emissions, possibly due to the unequal consumption patterns. However, lagged FDI (−0.63) suggests that earlier foreign investment fosters cleaner technologies over time. In the United Arab Emirates, past emissions (0.69) remain a key driver. GDP growth (17.33) increases emissions, while lagged GDP (−15.55) suggests efficiency improvements over time. Economic complexity (3.54) reflects industrial development’s environmental costs, but human development (−88.35) significantly reduces emissions, reinforcing the role of social progress in sustainability.

Table 11. Results of Dumitrescu-Hurlin Panel Causality Tests.

Null Hypothesis	W-Stat.	Z.-Stat	Prob.	Causality Direction
GINI⇏lnGDP	5.99	9.37	0.000	↔
lnGDP⇏GINI	7.64	12.52	0.000
CO2⇏lnGDP	1.96	0.060	0.000	↔
lnGDPS⇏lnCO2	2.11	7.45	0.000
ECI⇏lnGDP	1.40	3.0	0.54	→
lnGDP⇏ECI	4.99	−0.61	9 × 10−14
FDI⇏lnGDP	2.66	5.91	0.000	↔
lnGDP⇏FDI	0.77	10.20	0.005
CO2⇏GINI	6.31	9.99	0.000	↔
GINI⇏lnCO2	6.53	10.40	0.000
ECI⇏GINI	2.89	3.44	0.000	↔
GINI⇏ECI	2.39	2.49	0.000	↔
FDI⇏GINI	1.09	0.00	0.996
GINI⇏FDI	0.53	−1.06	0.288
HDI⇏GINI	4.39	6.32	3 × 10−10	↔
GINI⇏HDI	5.46	8.36	0.000

Source: authors’ calculations. Note: ⇏ shows “does not Granger cause” → and ↔ denote uni-directional and bi-directional causality.

presents the results of the Dumitrescu-Hurlin Panel Causality Tests by utilizing Equation (7), which examine the causal relationships between various variables. The table indicates whether there is bidirectional (↔) or unidirectional (→) causality between the variables based on the W-statistic, Z-statistic, and p-values. The results show bidirectional causality (↔) between the GINI and lnGDP, CO₂ and lnGDP, and FDI and lnGDP, suggesting that these variables influence each other. For example, income inequality and GDP, as well as CO₂ emission and GDP, have a mutual impact. Unidirectional causality (→) is observed from the ECI to lnGDP, indicating that economic complexity influences GDP but not vice versa. Similarly, the ECI and GINI have a bidirectional relationship, meaning that they both impact each other. For other variables like FDI and the GINI, the tests show no significant causal relationship. In summary, the table highlights the complex interactions between economic growth, inequality, emissions, investment, and human development, with several variables showing both direct and reciprocal influences.

Robustness Check

The present study utilized the CCR estimators to evaluate the consistency of the DOLS and FMOLS long-run estimations. The outcomes of the CCR analyses are presented in

Table 12. Result of canonical co-integration regression.

	Model 1		Model 2		Model 3
Variables	Coefficient	T	Coefficient	Standard Err.	Coefficient	t
lnGDP	-		−0.01	−0.46	3.71 ***	31.57
GINI	−1.12	1.58	-		27.71	0.83
CO2	0.49 ***	31.94	−0.02	−0.82	-
ECI	−0.4 ***	−7.88	−0.01 ***	−3.72	−0.29 ***	6.99
FDI	0.001 ***	2.87	0.001 ***	10.66	0.01 ***	7.69
HDI	3.22 ***	65.79	0.30 ***	−7.41	−4.56 ***	−14.52

Source: authors’ calculations. Note: ***: significant at the 1% level (t-statistic > 2.576 or <−2.576).

for all three models.

Table 12 presents the outcomes of three different models (Model 1, Model 2, and Model 3) estimated using the CCR method. These models examine the relationships between the various independent variables—the lnGDP, GINI, CO₂, ECI, FDI, and the HDI—and the dependent variable. In Model 1, significant relationships are observed with CO₂, the ECI, FDI, and the HDI. Both CO₂ and the HDI show a positive association with economic growth. Conversely, the ECI has a negative impact, indicating an inverse relationship. Model 2 also reveals significant associations with the ECI, FDI, and the HDI. The coefficients of the ECI and CO₂ remain negatively correlated with inequality. However, in this model, the lnGDP and CO₂ do not exhibit significant effects, suggesting that these variables may not be strong predictors in this context. In Model 3, the focus shifts to the lnGDP and FDI, both of which demonstrate strong positive relationships with CO₂, implying that increases in these variables are associated with higher CO₂ levels. On the other hand, the HDI exerts a strong negative influence, while the ECI continues to show a significant negative relationship. Our findings are consistent with the results obtained from the previously applied DOLS and FMOLS methods.

5. Conclusions and Policy Implications

This study examines the complex relationships between CO₂ emissions, income inequality, the Economic Complexity Index (ECI), foreign direct investment (FDI), the Human Development Index (HDI), and economic growth across countries. The findings underscore the interconnectedness of economic growth, social equity, and environmental sustainability, highlighting the need for balanced policy interventions. The results indicate that income inequality negatively affects economic growth, particularly in BRICS countries, by limiting the access to resources and opportunities. While CO₂ emissions, FDI, and the HDI positively influence GDP, the impact of FDI on inequality and environmental sustainability is more nuanced. In regions with weak regulatory frameworks, FDI can exacerbate the inequality and contribute to environmental degradation, emphasizing the need for policies that ensure sustainable and inclusive investment. Additionally, GDP, CO₂ emissions, and the ECI are negatively associated with income inequality, while FDI and the HDI tend to increase it. However, the HDI remains a key driver of economic growth, reinforcing the importance of investments in education, healthcare, and living standards in fostering innovation and productivity. Moreover, GDP, the Gini index, and FDI are positively linked to CO₂ emissions, whereas the HDI plays a crucial role in mitigating environmental harm by promoting sustainable practices and technological advancements. The Dumitrescu-Hurlin Panel Causality Tests further reveal bidirectional causality between GDP and CO₂ emissions, as well as between GDP and income inequality. This indicates a feedback loop where economic growth drives higher emissions, and vice versa, while inequality both influences and is influenced by GDP. Interestingly, the ECI exhibits a unidirectional causal relationship with GDP, suggesting that greater economic complexity fosters economic growth, but not the other way round. This highlights the role of sophisticated and diversified economic structures in driving long-term development. These findings emphasize the need for comprehensive policy frameworks that balance economic expansion, social equity, and environmental responsibility. While economic growth can lead to environmental challenges, improvements in human development mitigate these effects by promoting cleaner technologies and sustainability awareness. However, persistent income inequality remains a barrier to inclusive growth, necessitating targeted policies to reduce the disparities while fostering innovation-driven development. The findings suggest several important policy implications. First, promoting green industrialization and innovation by investing in renewable energy sectors, offering R&D tax credits, and fostering smart infrastructure can drive economic growth while minimizing the reliance on fossil fuels. In addition, encouraging infrastructure development through public–private partnerships and promoting trade diversification through multilateral trade agreements can enhance resilience and stimulate long-term growth. Addressing income inequality requires the implementation of progressive tax systems to fund social safety nets, expanding the conditional cash transfers to low-income households, and investing in inclusive education policies with a focus on skill development in green and high-tech industries. Furthermore, supporting small and medium enterprises through microfinance, tax incentives, and green subsidies will further foster equitable growth. To reduce emissions, BRICs should adopt carbon pricing mechanisms, including carbon taxes and emissions trading systems, and invest heavily in renewable energy infrastructure through subsidies and incentives. Strengthening environmental regulations for high-emission industries and promoting sustainable agricultural practices will contribute to lowering carbon footprints. Governments should carefully manage FDI inflows to ensure that they contribute to long-term sustainable development. While FDI can boost economic activity, policymakers need to enforce environmental regulations and promote investments in cleaner technologies to avoid the adverse environmental consequences associated with some forms of foreign investment. Strengthening the regulatory frameworks can ensure that FDI supports economic growth without exacerbating the inequality or damaging the environment. By aligning economic growth with social equity and environmental sustainability, BRICs can achieve a balanced and inclusive development model.

The study of the drivers of environmental sustainability, economic growth, and inequality in BRICS+ nations has certain limitations that need to be acknowledged. First, the availability of consistent data, particularly for some variables, posed a constraint. Data gaps for earlier years and specific countries limited the scope of the analysis, impacting the time frame and the inclusion of the key indicators. Additionally, important variables such as technological innovation, institutional quality, and environmental policies were omitted due to data limitations or their high correlation with other variables, potentially leading to multicollinearity issues. Stationarity issues were also encountered, with some variables exhibiting non-stationarity even after transformations like differencing, affecting the robustness of certain results. Furthermore, the study primarily focused on associations rather than fully exploring the causal relationships, which could have provided deeper insights. Geopolitical factors, policy shifts, and global events, such as the COVID-19 pandemic, were not fully accounted for, potentially influencing the findings related to economic growth, inequality, and environmental sustainability.

Looking ahead, there are several areas for future research to address these limitations. Future studies could incorporate additional variables, such as innovation, institutional quality, energy consumption patterns, and environmental regulations, to provide a more comprehensive analysis. Extending the time frame and gathering more granular data for underrepresented countries could further enhance the robustness of the findings, offering a deeper exploration of historical trends and long-term impacts. Employing advanced methodologies like panel vector autoregression (PVAR), instrumental variables (IV), or difference-in-differences (DID) could help identify causal relationships and provide clearer policy recommendations. Additionally, sector-specific analyses could shed light on the varying impacts of different industries—such as manufacturing, agriculture, and services—on growth, inequality, and sustainability. Incorporating policy simulations or scenario-based models could also offer valuable insights into the effects of specific interventions, such as carbon taxes or green finance. Finally, expanding the analysis to include non-BRICS+ nations or conducting inter-regional comparisons would provide a broader perspective, highlighting how BRICS+ countries perform relative to other emerging or developing economies. By addressing these limitations and exploring these areas for future research, scholars can contribute to a more comprehensive understanding of the complex dynamics within BRICS+ nations.