Diverging Paths in Environmental Performance: A Comparative Analysis of Innovation, Growth and Renewable Energy in OECD and BRICS Countries

Volume : 6 Issue : 2 Year : 2025 Pages : 59-75 e-ISSN : 2717-9230 59 DIVERGING PATHS IN ENVIRONMENTAL PERFORMANCE: A COMPARATIVE ANALYSIS OF INNOVATION, GROWTH AND RENEWABLE ENERGY IN OECD AND BRICS COUNTRIES Osman SUNTURa, Bekir Sami OĞUZTÜRKb a Corresponding Author, Süleyman Demirel University, Graduate School of Social Sciences, Department of Economics, [email protected], hAps://orcid.org/0000-0003-3585-1959. b Prof. Dr., Suleyman Demirel University, Faculty of Economics and AdministraMve Sciences, Department of Economics, [email protected], hAps://orcid.org/0000-0003-3076-9470. ABSTRACT: Comba&ng global climate change requires urgent and differen&ated strategies to reduce carbon dioxide (CO₂) emissions. The environmental performance trajectories of developed (OECD) and emerging (BRICS) economies represent a cri&cal area of research, as they account for a significant por&on of global emissions. This study aims to compara&vely analyze the key determinants of CO₂ emissions in OECD and BRICS countries, focusing on 2021, which reflects the unique condi&ons brought about by the post-COVID-19 economic recovery. Using 2021 data obtained from the World Bank, a cross-sec&onal “snapshot” analysis was conducted using mul&ple regression methods. In the model, the dependent variable is total CO₂ emissions (kt); the independent variables are defined as ‘renewable energy consump&on’, ‘GDP’, ‘urbaniza&on’, and ‘total patent applica&ons’ (innova&on proxy). Empirical findings confirm that renewable energy consump&on has a sta&s&cally significant and nega&ve effect on CO₂ emissions. In contrast, GDP and urbaniza&on were found to have a posi&ve effect on emissions. It is noteworthy that innova&on, measured by ‘total patent applica&ons’, shows a weak or sta&s&cally insignificant effect on emission reduc&on. The study contributes to the literature by presen&ng an analysis of a cri&cal period such as 2021 and highligh&ng the structural differences between the OECD and BRICS blocs. The results indicate that emission reduc&on policies should be designed according to countries' levels of development and the specificity of their innova&on policies (specifically targe&ng green technologies), rather than a “one-size-fits-all” approach. Keywords: CO₂ Emissions, Green Innova:on, Renewable Energy, Environmental Performance, Mul:ple Regression, OECD and BRICS. RECEIVED: 22 September 2025 ACCEPTED: 10 November 2025 DOI: hPps://doi.org/10.5281/zenodo.18057002 CITE Suntur, O., Oğuztürk, B. S., (2025). Diverging Paths in Environmental Performance: A Compara:ve Analysis of Innova:on, Growth and Renewable Energy in OECD and BRICS Countries. European Journal of Digital Economy Research, 6(2), 59-75. hPps://doi.org/10.5281/zenodo.18057002 Research Paper Suntur, Oğuztürk 60 1. INTRODUCTION In recent years, the accelera:on of globaliza:on, industrializa:on, and urbaniza:on has led to a sharp increase in global energy demand and carbon dioxide (CO₂) emissions. A par:cularly large por:on of these emissions originate from a limited number of economies that combine high income, produc:on, and trade integra:on with significant energy consump:on. OECD and BRICS countries account for a large share of global GDP, popula:on, and primary energy use, and therefore occupy a central posi:on in the global carbon budget. At the same :me, these countries exhibit heterogeneous development paths. OECD members are generally characterized by mature industrial structures, high urbaniza:on rates, and stricter environmental regula:ons, while BRICS economies have experienced rapid economic and demographic growth, o\en accompanied by energy-intensive industrializa:on and expanding urban agglomera:ons. This structure makes joint analysis of the OECD and BRICS crucial for understanding the drivers of global CO₂ emissions and the possibili:es for low-carbon transi:ons. Sustaining economic growth alongside industrializa:on has become a priority at the global level; however, this process has also brought significant externali:es, such as environmental degrada:on and, in par:cular, increased CO₂ emissions. OECD countries have long accounted for a large share of global emissions due to their high energy consump:on and produc:on volumes. On the other hand, BRICS countries have also aPracted aPen:on in recent years with their rapidly growing economies, increasing popula:ons, and energy demand, making them central actors in global climate change discussions. Indeed, although per capita carbon emissions in BRICS countries are rela:vely low, the environmental impact of these countries is increasing in terms of total volume. Therefore, examining the effects of economic growth, energy structure, and technological developments on CO₂ emissions in OECD and BRICS countries is cri:cally important in terms of both environmental sustainability and policy-making. In this context, the ques:on of whether variables such as the shi\ towards renewable energy sources and innova:on capacity, measured by the number of patents, have a carbon emission-reducing effect is current and important (Setyadharma et al., 2024; Van & Sadradin, 2021). The complex rela:onship between environmental performance, economic growth, urbaniza:on, and energy consump:on has been extensively examined in the econometric literature. In this field, renewable energy consump:on and technological innova:on are o\en highlighted as fundamental strategies for reducing CO₂ emissions. Empirical findings consistently support the role of renewable energy in reducing emissions, while the net effect of technological innova:on shows significant uncertainty, largely depending on how innova:on is measured (Dialchiev et al., 2023; Rainville et al., 2025; Sahoo et al., 2022; Van & Sadradin, 2021). For example, a panel cointegra:on analysis covering 37 OECD countries (1990–2019) reached the paradoxical conclusion that technological developments significantly increased CO₂ emissions when using “total number of patents” as an innova:on indicator (Van & Sadradin, 2021). Similarly, another study on 14 developing Asian countries, despite using the more specific “environmental technology patents,” found that these innova:ons played only a “modest role” in reducing emissions and that their effects were condi:onal on being supported by economic growth (Sahoo et al., 2022). These conflic:ng findings highlight the cri:cal methodological weaknesses of patent indicators used in empirical analyses. Indeed, recent methodological studies comparing patent classifica:on systems have shown that “green patent” systems that specifically label technologies comba:ng climate change (such as the Coopera:ve Patent Classifica:onCPC Y02 class) provide a superior measure that is more comprehensive, detailed, and carries a lower risk of misclassifica:on compared to general inventories (Rainville et al., 2025). Climate change is an urgent global issue due to the con:nued increase in greenhouse gas emissions. The Intergovernmental Panel on Climate Change (IPCC) emphasizes that rapid and deep cuts in CO₂ emissions are necessary to achieve interna:onal targets. For example, roadmaps limi:ng warming to 1.5°C project a 35-51% reduc:on in CO₂ emissions from the energy system by 2030 and an 87-97% reduc:on by 2050. Achieving these targets requires a rapid transi:on to low-carbon technologies. Scenarios limi:ng warming to 2°C Diverging Paths in Environmental Performance: A Compara:ve Analysis of Innova:on, Growth and Renewable Energy in OECD and BRICS Countries 61 project that low-carbon sources will provide approximately 93-97% of global electricity by 2050 (Clarke et al., 2022). Technological innova:on and renewable energy use are considered cri:cal drivers of this transi:on. Indeed, while the IPCC notes that breakthroughs such as the widespread adop:on of solar photovoltaics and LEDs would not be possible without focused innova:on efforts, it also warns that innova:on alone, if not guided by robust policies, could create undesirable “backlash” effects (Blanco et al., 2022). In this context, OECD and BRICS countries require special aPen:on. The five BRICS countries (Brazil, Russia, India, China, South Africa) currently account for approximately 42% of global CO₂ emissions due to rapid industrializa:on and intensive fossil fuel use (Erkılıç et al., 2025). OECD members, represen:ng advanced economies, have historically contributed significantly to emissions but are also pioneers in clean technology research and applica:ons. Analyzing these two blocs together allows us to understand their different levels of economic development and policy environments. For example, Iranmanesh (2025) found that OECD and BRICS financial markets are largely unintegrated and that there is no overall convergence due to differences in infrastructure, economic size, and regulatory factors (Iranmanesh, 2025). This implies that innova:on systems and investment models are similarly differen:ated. Nevertheless, most empirical research treats the OECD and BRICS separately. Studies either examine innova:on– emission dynamics in OECD samples (e.g., Saqib et al., 2023) or individual BRICS countries; the number of studies directly comparing these two groups is quite limited. This study aims to fill this gap with a cross-sec:onal analysis focusing on 2021, which coincides with the post-COVID-19 recovery process. This year provides a unique context for understanding energy use and the innova:on-emissions rela:onship. Analyzing data from 2021, this study provides a global “snapshot” of the condi:ons that emerged a\er the ini:al shocks of COVID-19. 2021 was the first full year of recovery from the pandemic and was characterized by unique policy changes and challenges. Cross-sec:onal analyses provide “snapshot insights” into the composi:on of the popula:on during this period, contribu:ng to an understanding of the period's specific condi:ons (Wang & Cheng, 2020). For instance, the global distribu:on of vaccines in 2021 marked a significant turning point; however, Klobucista and Merrow (2021) noted that "persistent pressures on health systems" con:nued to prevail. Notably, the global economy was projected to grow by 5.9% in 2021 (following the sharp contrac:on in 2020), revealing that 2021 was a year of uneven recovery (IMF, 2021). This situa:on has also led to a no:ceable rebound in global CO₂ emissions. Following the historic decline in 2020, 2021 is a cri:cal turning point in understanding how the structural rela:onship between growth (ln_GDP) and emissions (ln_CO₂) is being reestablished. Furthermore, 2021 is the year when many countries began implemen:ng their “Green Recovery” policies. In this context, concentra:ng on a single year allows for the evalua:on, within the most current framework, of whether the new policy commitments have a measurable effect on "green sustainability" (Renewable) and "innova:on" (ln_Patent). The single-year analysis approach ensures that the findings reflect these urgent phenomena (and are not overshadowed by preor post-pandemic changes) (Wang & Cheng, 2020; Yacoubian et al., 2025). Although this approach cannot directly track changes over :me, it provides policymakers with a clear picture of the condi:ons in 2021. 2. THEORETICAL FRAMEWORK Fundamental debates in environmental economics offer various theore:cal frameworks aimed at explaining the complex rela:onship between economic ac:vi:es and environmental degrada:on. The Environmental Kuznets Curve (EKC) assumes an inverted U-shaped rela:onship between per capita income and environmental degrada:on. Inspired by Kuznets' income inequality curve, Grossman and Krueger (1991) and Panayotou (1997) argued that as economies grow, pollu:on first increases (scale effect) and then decreases a\er a certain income threshold is crossed (due to composi:on and technology effects). In other words, low-income growth increases pressure on the environment, but higher income eventually reverses this trend by encouraging cleaner technologies and demand for environmental quality (Bousnina et al., 2025; Ertaş & Uysal, 2014). The inverse U-shaped rela:onship between per capita income and environmental pollu:on is generally explained through three fundamental mechanisms: economies of scale, structural Suntur, Oğuztürk 62 change, and technological development. When income levels are low, economic growth intensifies environmental pressures by increasing produc:on volume; this process is defined as economies of scale and explains the rising part of the curve. However, once income levels exceed a certain threshold, the economy's structure shi\s toward less pollu:ng sectors (structural effect), and cleaner produc:on techniques and environmentally friendly technologies become more widespread (technology effect). This transforma:on leads to a decrease in environmental degrada:on and forms the descending part of the curve (Ertaş & Uysal, 2014, p. 7). In recent literature, this approach has been frequently tested in the context of CO₂ emissions in both developed and developing countries. For example, (Mirziyoyeva & Salahojaev, 2023) provide a detailed theore:cal defini:on of the EKC in one study, while (Bousnina et al., 2025) examine the rela:onship between economic growth and carbon emissions from an EKC perspec:ve in another study. Similarly, another study (Gieraltowska et al., 2022) analyzes different dimensions of this rela:onship through the variables of industrializa:on and energy consump:on. However, empirical findings vary from country to country and depending on the methods used, which makes the universal validity of the ECF hypothesis debatable. The Porter Hypothesis (PH) argues that welldesigned, stringent environmental regula:ons can encourage innova:ons that offset compliance costs and even enhance compe::ve advantage. Michael Porter (1991) and Porter and van der Linde (1995) argued that strict but flexible regula:ons encourage firms to discover costsaving clean technologies (an innova:on trade-off) and thus the overall effect can be win-win. Porter and van der Linde describe this as a proac:ve environmental strategy that leads to increased profits in the long term despite short-term costs (Akdemir Ömür, 2021). According to Porter and van der Linde, there is a mul:faceted mechanism that explains how environmental regula:ons encourage innova:on. According to this view, regula:ons primarily serve as a signal that reveals exis:ng resource inefficiencies and poten:al technological opportuni:es for companies. This increased ins:tu:onal awareness, driven by knowledgebased regula:ons, combines with a guarantee that reduces uncertainty about the future value of investments. The pressure created by regula:ons mo:vates firms to innovate while also triggering an industry-wide transforma:on on a level playing field by crea:ng equal condi:ons for all compe:tors. However, the authors also acknowledge that these innova:ons may not always fully offset compliance costs in the short term (especially before learning curve costs are reduced) (Ambec et al., 2010). There are different approaches in the literature that classify the Porter Hypothesis (PH). In par:cular, the widely accepted typology developed by Jaffe and Palmer (1997) divides this theory into three main versions: weak, strong, and narrow. The weak Porter Hypothesis argues that environmental regula:ons encourage firms (especially green ones) to innovate, but it does not concern itself with whether this innova:on results in a net economic gain. The focus is on the trigger effect of regula:on on innova:on. The Strong Porter Hypothesis, on the other hand, makes a more asser:ve claim. The innova:on triggered by regula:ons more than offsets compliance costs, providing the firm with a net compe::ve advantage and profitability; this is a true “win-win” scenario. Finally, the Narrow Porter Hypothesis focuses on the type of regula:on. According to this view, flexible, market-based policies (e.g., pollu:on taxes or emissions trading permits) are much more powerful and effec:ve in encouraging innova:on than rigid command-and-control rules. These theore:cal dis:nc:ons are cri:cal in determining which hypothesis (innova:on incen:ves or net profitability) empirical tests target (Zhang et al., 2024). Despite the op:mis:c “win-win” expecta:on of the Porter Hypothesis, the assump:on that innova:on will always be ‘green’ is theore:cally debatable. At this point, the Directed Technical Change (DTC) theory offers a cri:cal perspec:ve. This theory, developed by Acemoglu, formalizes that innova:on is not random but is driven by economic factors. According to the model, the ‘price effect’ directs R&D towards inputs such as scarce resources that become more expensive, while the market size effect favors sectors with larger outputs (such as the s:ll dominant fossil fuel sector). In short, the DTC theory implies that policies (taxes, subsidies) and rela:ve factor Diverging Paths in Environmental Performance: A Compara:ve Analysis of Innova:on, Growth and Renewable Energy in OECD and BRICS Countries 63 supply can ac:vely steer innova:on toward dirty or clean direc:ons (Acemoglu, 2002). In an environmental context, Acemoglu et al. (2012) extend the model to dirty (fossil fuel-based) and clean (renewable) technologies, showing that when le\ to its own devices (if dirty technology is more profitable), the market can lock innova:on into the dirty direc:on. They argue that ac:ve policy interven:ons, such as carbon taxes and green R&D subsidies, are necessary to break this lock and redirect innova:on toward clean inputs. This theore:cal framework is directly related to the empirical findings of this study. If dirty technologies are s:ll more profitable than clean technologies in the sample of OECD and BRICS countries, firms will direct their innova:on efforts (ln_Patent) towards the “gray” area, as predicted by DTC theory. This situa:on provides a theore:cally strong explana:on for why a general innova:on indicator such as total patents could have a posi:ve (+) effect on CO₂ emissions (ln_CO₂). Gray innova:on refers to innova:ons that increase the efficiency of exis:ng (dirty) technologies rather than developing en:rely new clean technologies. In low-carbon innova:on research, academics dis:nguish between ‘clean’ innova:ons (new renewable or zero-carbon technologies) and ‘gray’ innova:ons, which are incremental improvements in the energy efficiency or pollu:on reduc:on of exis:ng industrial processes. For example, Yan et al. (2017), by categorizing patented low-carbon technologies into clean and gray categories, found that clean innova:ons significantly reduce CO₂ emissions, while gray innova:ons have an uncertain effect. They explain this with rebound effects: increases in energy efficiency (gray innova:on) lower the effec:ve price of energy and may encourage greater use, par:ally offse€ng the direct savings (Yan et al., 2017). The rebound effect (or “recovery” effect) describes the phenomenon where improvements in energy efficiency typically result in less-than-expected reduc:ons in energy use. This is because the saved resources are par:ally recovered through increased consump:on. In other words, more efficient devices or vehicles reduce the cost of energy services, so consumers use them more (like driving longer distances when gasoline is cheaper). This idea dates back to Jevons (1865), who warned that increases in coal efficiency could backfire and increase coal use. Khazzoom (1980) and Brookes (1990) later applied this idea to modern economies. They argued that overall efficiency gains could increase total energy consump:on at the macro level (the Khazzoom–Brookes paradox) (Gillingham et al., 2015). The modern literature classifies the rebound effect as follows: Direct rebound: A sudden increase in usage due to the cheapening of energy services (e.g., increased car usage a\er purchasing a fuel-efficient model). Efficiency increases demand by lowering the implicit price of energy services. Indirect rebound: Addi:onal consump:on of other goods thanks to the income saved (e.g., taking a plane trip with the money saved on fuel). Increases in efficiency raise real income, and part of this income is spent again on goods whose produc:on requires energy. Macroeconomic feedback: The sum of direct and indirect effects across the economy. This includes macroeconomic price adjustments and growth effects. In the overall balance, efficiency can increase overall economic growth and energy demand, making macroeconomic feedback significant (Kavaz, 2023). Finally, the Energy Subs:tu:on Hypothesis proposes that the widespread adop:on of renewable energy will replace fossil fuel use and thus reduce carbon emissions. In other words, as economies shi\ from non-renewable to renewable sources, the carbon intensity of energy decreases. Replacing fossil fuels with renewable energy sources reduces humanity's ecological footprint (Kılınç, 2023). In prac:ce, most econometric studies empirically demonstrate that renewable energy consump:on has a reducing effect on CO₂ emissions, while fossil fuel consump:on increases emissions (Shafiei & Salim, 2014; Bölük & Mert, 2014). This subs:tu:on effect is the ‘technology effect’ mechanism of the Environmental Kuznets phenomenon (Panayotou, 1997). At higher stages of development, economies invest in clean energy and reduce emissions by elimina:ng dirty inputs (Jie and Khan, 2024). 3. PREVIOUS EMPIRICAL STUDIES The effects of economic growth (GDP), technological innova:on (ln_Patent), and energy consump:on (Renewable) on CO₂ emissions have Suntur, Oğuztürk 64 been examined in considerable detail and comprehensively in the sustainability literature within the context of the Environmental Kuznets Curve (EKC), Porter's Hypothesis, or Energy Subs:tu:on theories. In this context, the main objec:ve of this study is to analyze the combined effect of renewable energy consump:on as a ‘green sustainability’ indicator and patent applica:ons as a ‘green innova:on’ indicator on environmental performance (ln_CO2) under the control variables of economic growth and urbaniza:on (Urban). Despite the intense interest in these topics in the literature and the numerous empirical studies on the determinants of CO₂ emissions, there is a lack of research that simultaneously addresses the dynamics of both developed (OECD) and emerging (BRICS) economies, which represent a very large por:on of the global economy and emissions, and examine the simultaneous effect of this specific set of variables (innova:on, renewable energy, and growth) in this mixed group of countries (OECD+BRICS). This study aims to fill this gap using 2021 cross-sec:onal data. Alam (2024) iden:fied cointegra:on between CO₂, GDP per capita, GDP per capita squared, and energy consump:on per capita using panel data for 24 OECD countries for the period 1971–2016 and found an inverse U-shaped rela:onship between income per capita and CO₂ based on Fully Modified Least Squares model es:mates. The study confirmed the EKC hypothesis. Muratoğlu et al. (2024) tested the EKC for 38 OECD countries for the period 1990–2022 in four sectors—agriculture, industry, manufacturing, and services—using a panel nonlinear ARDL (PNARDL) model. The analysis revealed that the validity of the EKC varies across sectors, with the EKC being proven correct in all sectors except the industrial sector. Akar et al. (2025) examined the impact of economic growth and energy consump:on on carbon emissions in OECD countries and some lateindustrialized Asian economies between 1990 and 2020. The analysis, conducted using advanced panel methods, found a significant inverse-U rela:onship between per capita income and CO₂ for OECD countries, concluding that the EKC is valid for OECD countries but not for other latedeveloping Asian countries. Again, this study found a unidirec:onal causality from economic growth to emissions and energy consump:on. Furthermore, Kasperowicz (2015), in his study covering the years 1995-2012 and performing panel data analysis on 18 EU member countries, confirmed that while there is a nega:ve rela:onship between GDP and CO₂ in the long term, there is a posi:ve rela:onship in the short term. Alam et al. (2016), in their study covering the period 1970-2012 and the countries of India, Indonesia, China, and Brazil, found that CO₂ emissions increased with rising income and energy consump:on in these four countries and that there was a posi:ve rela:onship. Naimoğlu and Özbek (2022), in their study covering the period 1990-2019 for Turkey, confirmed that the EKC is valid for Turkey in both the short and long term. Konya (2022) tested the existence of the EKC for 10 developing country economies using panel data models for the period 1992-2014. No clear result was obtained regarding the rela:onship between carbon emissions and economic growth in terms of the EKC. Ridzuan et al. (2022) conducted research using the ARDL approach with a data set covering the years 1971-2019 to test the CEC for Malaysia, a BRICS member country. The analysis found that in the short term, there is an inverted U-shaped CEC, while in the long term, there is a Ushaped CEC. Nica et al. (2025) examined the dynamic rela:onship between economic growth, technological innova:on, and carbon emissions in BRICS countries during the period 1991-2023 within the scope of the EKC. The results confirmed the EKC hypothesis, showing that there is an Nshaped rela:onship between GDP and carbon emissions, i.e., there are two different turning points. Cheng et al. (2021) used panel quan:le regression for 35 OECD countries for the period 1996–2015 and found that innova:on directly reduces CO₂ emissions, but the effect is heterogeneous across countries and upper quan:les of the distribu:on. Similarly, Saqib et al. (2023) reported that patent development reduces CO₂ using panel quan:le analysis for 32 OECD countries (1996–2020). They also found that the effect varied across quan:les. In contrast, Van and Sadradin (2022) applied panel regression (including unit root tests such as CADF and ADF) in the OECD and concluded that patent development increased CO₂ emissions. Haq et al. (2024) found that an increase in intellectual capital significantly reduced CO₂ emissions in the OECD, whereas no similar reduc:on effect was observed in the BRICS countries. Studies conducted for BRICS countries (Brazil, Russia, India, China, South Africa) similarly use patent counts, R&D, and “green innova:on” indicators. The findings are Diverging Paths in Environmental Performance: A Compara:ve Analysis of Innova:on, Growth and Renewable Energy in OECD and BRICS Countries 65 heterogeneous from the OECD. Haq et al. (2024) could not detect a significant CO₂ reducing effect of intellectual capital for BRICS. Xiaoyang et al. (2022), however, using a two-stage least squares and GMM (Panel Generalized Method of Moments) analysis on a panel of 36 OECD + 5 BRICS countries (2005–2018), found that innova:on and R&D expenditures are posi:vely related to CO₂ emissions, meaning that as innova:on increases, so do emissions. On the other hand, Qamruzzaman et al. (2025) reported that technological and environmental innova:ons significantly reduced carbon emissions for BRICS countries (1995–2023) using panel ARDL analysis. Similarly, Mehta et al. (2025) used annual panel data for BRICS countries from 2000 to 2024 and found that an increase in technological innova:on and renewable energy integra:on would lead to a reduc:on in emissions. Empirical studies conducted for OECD countries generally emphasize the reducing effect of renewable energy consump:on on CO₂ emissions. For example, Işık et al. (2024), using panel data from 27 OECD countries (2001–2020) in their quan:le regression analysis, found a nega:ve rela:onship between renewable energy consump:on and CO₂ emissions; that is, they showed that renewable energy use has a reduc:on effect on emissions. Mirziyoyeva and Salahodjaev (2023), in their study examining the 50 most global countries, also found that an increase in renewable energy use reduces carbon emissions. The study found that a 1-point increase in the share of renewable energy in total energy consump:on led to a 0.26 reduc:on in per capita carbon emissions. Setyadharma et al. (2024), using data from 1992-2020, found that a 1% increase in renewable energy use in BRICS countries resulted in a 0.029% reduc:on in CO₂. Sahoo et al. (2022), in an analysis using data from 14 developing countries in Asia between 1990 and 2018, found that renewable energy consump:on and globaliza:on play an important role in reducing carbon emissions. Similarly, Van and Sadradin (2021), using data from 37 OECD countries between 1990 and 2019, found that renewable energy reduces carbon emissions. Gieraltowska et al. (2022), in their study using data from 163 countries between 2000 and 2016, concluded that renewable energy consump:on reduces carbon emissions and that there is an inverse U-shaped rela:onship between urbaniza:on and carbon emissions. There are conflic:ng findings in the literature regarding the effect of urbaniza:on on carbon emissions. While there are findings that urbaniza:on increases or decreases carbon emissions, there are also findings that its effect is limited. Voumik and Sultana (2022) showed in their CS-ARDL panel analysis that the urbaniza:on rate increases environmental degrada:on (and therefore CO₂ emissions) in BRICS countries. On the other hand, Ma and Ogata (2024) concluded that urbaniza:on reduces emissions in a group of developing countries including BRICS members. Furthermore, Vo et al. (2022) found that urbaniza:on has a limited effect on carbon emissions in OECD countries. In summary, this empirical literature review reveals a lively and complex academic debate surrounding the key variables of our study (growth, innova:on, renewable energy, and urbaniza:on). In par:cular, theore:cal and empirical uncertain:es regarding the effects of innova:on (green or gray) and growth (does the EKC hold?) increase the importance of the empirical test this study will conduct for the OECD and BRICS mixed group. 4. METHODOLOGY In this study, the following Multiple Linear Regression model was established to test the relationship between the factors discussed in the Theoretical Framework (Growth, Innovation, Renewable Energy, and Urbanization) and environmental performance (CO₂ ). ln(CO2)i=β0+β1ln(GDP)i+β2ln(Patent)i+β3 Renewablei+β4Urbani+εi Here, i represents each country, β0 is the constant term, β1….β4 are the coefficients, and ε is the error term. The natural logarithm (ln) transformation was applied to the CO2, GDP, and Patent variables included in the model. This transformation has two main purposes (Gujarati & Porter, 2009; Wooldridge, 2015). Statistical Reason (Data Normalization): These variables (GDP, emissions, and patent counts) exhibit very large-scale differences (outliers) across countries and tend to have a right-skewed distribution. The logarithmic transformation ‘normalizes’ the distribution of the data by softening the disproportionate effect of these Suntur, Oğuztürk 66 extreme values on the regression results and ensures better compliance with OLS regression assumptions (particularly the normality of the error terms). Reason for Interpretation (Flexibility): More importantly, the logarithmic transformation allows us to interpret the relationship between variables in ‘percentages’ rather than ‘units’ (e.g., dollars, tons). Thus, the coefficient of (for example) ln_GDP on ln_CO2 can be interpreted as an elasticity coefficient showing how much a 1% increase in GDP changes CO2 emissions. This provides a much more meaningful and powerful result for economic analysis. In this study, 2021 was selected as the crosssectional data year for analysis. Two main factors played a role in selecting this date. First, 2021 is the most complete and up-to-date data set (covering OECD and BRICS countries) for all five variables used in the study. Second, and more importantly, 2020, the peak year of the COVID-19 pandemic, was considered an outlier year due to artificial declines in GDP and CO2 emissions caused by global lockdowns. In order to analyze the structural relationship between the variables more accurately, data from 2021, when economic recovery began and the ‘new normal’ emerged, was preferred. In the 2021 data, missing data was found for only one country (Ethiopia), and this country was excluded from the analysis, resulting in a final sample size of n=47. All data required for the variables examined in the study were obtained from the World Bank (World Bank World Development Indicators) database. The sample for the study consists of OECD and BRICS countries. These two country groups were selected for the study because they play a key role in the global economy and emissions and also have different positions and characteristics. In this study, which examines the effects of country GDP, innovation, renewable energy, and urbanization on carbon emissions in OECD and BRICS countries, carbon emissions (ln_CO2) are the dependent variable. GDP (ln_GDP), established patent applications (ln_Patent), which we define as innovation, renewable energy consumption (Renewable), and urbanization (Urban) are the independent variables. The expected results of the study are as follows: As economic size (GDP) increases, production and consumption increase, which in turn increases carbon emissions (ln_CO2). H1: There is a statistically significant and positive relationship between economic growth (ln_GDP) and carbon emissions (ln_CO2). H2: As discussed in detail in the “Theoretical Framework” section of the study, the effect of innovation on CO2 emissions constitutes one of the most important theoretical conflicts in the literature. Therefore, two competing hypotheses have been developed for this variable: Hypothesis 2a: Porter Hypothesis (Green Innovation Effect) According to this optimistic view (Porter & van der Linde, 1995), well-designed policies trigger innovation. This “eco-innovation” (green innovation) increases firms' resource efficiency, cleans up production processes, and reduces net CO2 emissions. H2a: An increase in innovation capacity (ln_Patent) has a statistically significant and negative (-) (reducing) effect on CO2 emissions (ln_CO2). Hypothesis 2b: Gray Innovation / Backlash Effect According to an alternative (and more realistic) view, a general indicator such as ‘total patents’ may reflect ‘gray’ innovation (i.e., innovations that increase efficiency or consumption in dirty technologies) rather than ‘green’ innovation (Directed Technological Change theory; Acemoglu et al., 2012). Furthermore, efficiency gains can increase total consumption and thus emissions by creating a “Rebound Effect” (Sorrell, 2007). H2b: An increase in innovation capacity (ln_Patent) has a statistically significant and positive (+) (incremental) effect on CO2 emissions (ln_CO2). H3: An increase in the renewable energy consumption rate (Renewable) has a statistically significant and negative effect on carbon emissions (ln_CO2). The effect of urbanization on carbon emissions has different findings, as discussed earlier. There are two opposing views on the impact of urbanization on carbon emissions: the efficiency effect and the agglomeration effect (Voumik & Sultana, 2022; Ma & Ogata, 2024). Accordingly, two different hypotheses have been developed for the urbanization variable. According to the Concentration Effect, urbanization leads to the geographical concentration of industrial activities, energy consumption, transportation networks, and consumption patterns. This “scale effect” Diverging Paths in Environmental Performance: A Compara:ve Analysis of Innova:on, Growth and Renewable Energy in OECD and BRICS Countries 67 increases environmental pressure (Voumik & Sultana, 2022). H4a: There is a statistically significant and positive relationship between the urbanization rate (Urban) and carbon emissions (ln_CO2). According to the Efficiency Effect, urbanization is a factor that increases resource efficiency. High population density facilitates the development of lower-carbon infrastructure, such as public transportation, more efficient heating/cooling systems (apartments), and an economy based on the service sector (Ma & Ogata, 2024). H4b: There is a statistically significant and negative relationship between the urbanization rate (Urban) and carbon emissions (ln_CO2). To estimate the parameters of the model established in this study (β₁,β₂, β₃, β₄), a “Multiple Linear Regression” analysis based on the “Least Squares” (LS) method was performed. The ability of this method to produce reliable and unbiased results depends on the model meeting basic econometric assumptions. Therefore, along with the main regression analysis, a series of diagnostic tests were applied to test the statistical validity of the model. First, the issue of “multicollinearity” was examined to measure the risk of high correlation between the independent variables. The correlation analysis, as expected, showed a high positive relationship (r = .894) between the ln_GDP and ln_Patent variables. To test whether this situation caused a bias in the OLS estimates, the VIF (Variance Inflation Factor) values, which are a more reliable indicator, were examined. As reported in the “Findings” section, the VIF values of all variables (e.g., ln_GDP=5.466, ln_Patent=5.109) were found to be below the critical threshold of 10 accepted in the literature. This confirms that there is no serious multicollinearity problem in the model. Second, to test the assumption of constant variance (homoscedasticity), one of the OLS assumptions, the “Heteroscedasticity” (varying variance) situation was examined. This test was visually examined using a scatterplot plotting the standardized residuals (*ZRESID) against the standardized estimated values (*ZPRED). The graph presented in the “Findings” section shows that there is no distinct ‘funnel’ structure and that the residuals are distributed in a random “cloud” shape, indicating that this assumption is met. Finally, the “Normality of Residuals” assump:on was tested using the Histogram and Normal P-P Plot of the regression residuals. As will be shown in the “Findings” sec:on, the close clustering of observa:ons around the 45-degree diagonal line in the P-P plot and the reasonable fit of the histogram to a bell curve confirm that the normality assump:on is also sa:sfied. 5. FINDINGS This section presents the results of the econometric analyses conducted for the research model and their interpretation in light of the theoretical framework. Table 1. Descriptive Statistics Variable N Minimum Maximum Mean Std. Deviation Renewable 47 0.10 82.40 22.94 17.26 Urban 47 35.39 98.12 76.89 13.34 ln_CO2 47 1.17 9.44 4.89 1.80 ln_Patent 47 2.71 14.17 7.38 2.49 ln_GDP 47 23.97 30.80 27.08 1.51 Table 1 confirms that the analysis covers the entire final sample of n=47 countries and that there is no missing data in the variables. According to 2021 data, the average renewable energy consumption rate (Renewable) of the sample consisting of OECD and BRICS countries is 22.94%, while the average Suntur, Oğuztürk 74 Grossman, G. M., & Krueger, A. B. (1991). Environmental impacts of a North American Free Trade Agreement (NBER Working Paper No. 3914). National Bureau of Economic Research. https://doi.org/10.3386/w3914 Gujarati, D. N., & Porter, D. C. (2009). Basic econometrics (5th ed.). McGraw Hill. Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Mul\variate data analysis (7th ed.). Pearson. Haq, W., Ahmad, I., Arshad, A., & Alvi, S. (2025). The role of human capital and innovation in reducing carbon emissions in OECD and BRICS countries. Sustainable Futures, 10, Article 101475. hAps://doi.org/10.1016/j.s{r.2025.101475 InternaMonal Monetary Fund. (2021, Ekim). World Economic Outlook, October 2021: Recovery during a pandemic. hAps://www.imf.org/en/PublicaMons/WEO/Issues/ 2021/10/12/world-economic-outlook-october2021. Iranmanesh, N. (2025). Analysis of financial convergence between the BRICS and OECD countries. PLOS ONE, 20(2), Article e0316410. https://doi.org/10.1371/journal.pone.0310950 Işık, C., Bulut, U., Ongan, S., Islam, H., & Irfan, M. (2024). Exploring how economic growth, renewable energy, internet usage, and mineral rents influence CO2 emissions: A panel quantile regression analysis for 27 OECD countries. Resources Policy, 92, Article 105025. https://doi.org/10.1016/j.resourpol.2024.105025 Jaffe, A.B. & Palmer, K. (1997). Environmental regulation and innovation: A panel data study. The Review of Economics and Statistics, 79(4), 610-619. hAps://doi.org/10.1162/003465397557196 Jie, W., & Khan, R. (2024). Breaking the CO2 gridlock: Can renewables lead the way for the OECD? Energies, 17, Article 4511. h<ps://doi.org/10.3390/en17174511 Kasperowicz, R. (2015). Economic growth and CO2 emissions: The ECM analysis. Journal of International Studies, 8(3), 91-98. https://doi.org/10.14254/20718330.2015/8-3/7 Kavaz, İ. (2023). Gelişmekte olan ülkelerde enerji verimliliği ve geri tepme etkisinin analizi. Anadolu İktisat ve İşletme Dergisi, 7(2), 49-73. hAps://doi.org/10.59293/anadoluiid.1363571 Kılınç, E. C. (2023). Yenilenebilir enerji ve fosil yakıt tüketiminin ekolojik ayak izi üzerindeki etkisi: Türkiye örneği. Finans Ekonomi ve Sosyal Araştırmalar Dergisi, 8(3), 731-749. hAps://doi.org/10.29106/fesa.1307807 Klobucista, C., & Merrow, J. (2021, 22 Aralık). Seven charts that explain the COVID-19 pandemic in 2021. Council on Foreign RelaMons. hAps://www.cfr.org/arMcle/seven-charts-explaincovid-19-pandemic-2021 Konya, S. (2022). Panel estimation of the Environmental Kuznets Curve for CO2 emissions and ecological footprint: Environmental sustainability in developing countries. Folia Oeconomica Stetinensia, 22(2), 123-145. DOI: 10.2478/foli-20220022 Ma, B., & Ogeta, S. (2024). Impact of urbanization on carbon dioxide emissions: Evidence from 136 countries and regions. Sustainability, 16(18), Article 7578. h<ps://doi.org/10.3390/su16187878 Mehta, D., Patel, N., & Abdi, A. H. (2025). Nexus between carbon emissions, economic growth, renewable energy, and innovation of BRICS: Kuramoto dynamic synchronization approach. Discover Sustainability, 6, Article 987. DOI:10.1007/s43621025-01799-1 Mirziyoyeva, Z., & Salahodjaev, R. (2022). Renewable energy, GDP and CO2 emissions in high-globalized countries. Frontiers in Energy Research, 11, 1-8. hAps://doi.org/10.3389/fenrg.2023.1123269 Muratoğlu, Y., Songur, M., Uğurlu, E., & Şanlı, D. (2024). Testing the Environmental Kuznets Curve hypothesis at the sector level: Evidence from PNARDL for OECD countries. Frontiers in Energy Research, 12, Article 1452906. https://doi.org/10.3389/fenrg.2024.1452906 Naimoğlu, M., & Özbenk, S. (2022). Testing the Environmental Kuznets Curve hypothesis under structural breaks: The case of Turkey. Alanya Akademik Bakış Dergisi, 6(3), 3069-3084. hAps://doi.org/10.29023/alanyaakademik.992079 Nica, I., Georgescu, I., & Kinnunen, J. (2025). Economic growth, innovation, and CO2 emissions: Analyzing the Environmental Kuznets Curve and the Innovation Claudia Curve in BRICS countries. Sustainability, 17, Article 3507. h<ps://doi.org/10.3390/su17083507 Panayotou, T. (1997). Demystifying the environmental Kuznets curve: Turning a black box into a policy tool. Environment and Development Economics, 2(4), 465-484. DOI:10.1017/S1355770X97000259 Porter, M.E. (1991). America’s green strategy. Scientific American, 264(168). hAp://dx.doi.org/10.1038/scienMficamerican0491168 Porter, M. & van der Linde, C. (1995). Toward a new conception of the environment-competitiveness relationship. Journal of Economic Perspectives, 9(4), 97-118. DOI: 10.1257/jep.9.4.97 Qamruzzaman, M., Almulhim, A. A., & Aljughaiman, A. A. (2025). Nexus between uncertainty, innovation, and environmental sustainability in BRICS: An analysis under the Environmental Kuznets Curve (EKC) framework. Frontiers in Environmental Science, 13, Article 1570150. https://doi.org/10.3389/fenvs.2025.1570150 Rainvelle, A., Dikker, I., & Buggenhagen, M. (2025). Tracking innovation via green patent classification systems: Are we truly capturing circular economy progress? Diverging Paths in Environmental Performance: A Compara:ve Analysis of Innova:on, Growth and Renewable Energy in OECD and BRICS Countries 75 Journal of Cleaner Production, 486, Article 144385. https://doi.org/10.1016/j.jclepro.2024.144385 Ridzuan, N. H. A. M., Abdullah, A., & Ibrahim, M. A. (2022). RevisiMng the environmental Kuznets curve hypothesis in Malaysia: An ARDL approach. Environmental Science and Pollu\on Research, 29(44), 66597–66611. hAps://doi.org/10.1007/s11356-022-20516-x Sahoo, B., Behera, D., & Rahut, D. (2022). Decarbonization: Examining the role of environmental innovation versus renewable energy use. Environmental Science and Pollution Research, 29, 48704-48719. DOI:10.1007/s11356-022-18686-1 Saqib, N., Usman, M., Mahmood, H., Abbas, S., Ahmad, F., Mihai, D., & Mallek, R. S. (2023). The moderating role of technological innovation and renewable energy on CO2 emission in OECD countries: Evidence from panel quantile regression approach. Economic Research, 36(3), 1-22. https://doi.org/10.1080/1331677X.2023.2168720 Setyadharma, A., Anawati, T., Sa’adah, N., & Sholikhatun Jamil, N. (2024). Analysis of the determinants of carbon dioxide emissions in BRICS countries: A panel dynamic ordinary least square (DOLS) estimation approach. E3S Web of Conferences, 557, Article 04010. DOI:10.1051/e3sconf/202455704005 Sorrell, S. (2007). The rebound effect: An assessment of the evidence for economy-wide energy savings from improved energy efficiency. UK Energy Research Centre (UKERC). Van, M. H., & Sadradin, H. F. (2021). The effect of technological innovation on CO2 emissions in OECD countries: Using panel regression. International Journal of Contemporary Economics and Administrative Sciences, 11(2), 438-453. https://doi.org/10.5281/zenodo.6582440 Vo, D. H., Ho, C. M., Thuong Le, Q. T., & The Vo, A. (2022). Revisiting the energy-growth-environment nexus in the OECD countries: An application of the CS-ARDL approach. Energy, Sustainability and Society, 12, Article 47. hAps://doi.org/10.1186/s13705-02200375-z Voumik, L. C., & Sultana, T. (2022). Impact of urbanization, industrialization, electrification and renewable energy on the environment in BRICS: Fresh evidence from novel CS-ARDL model. Heliyon, 8(11), Article e11340. hAps://doi.org/10.1016/j.heliyon.2022.e11457 Wang, X., & Cheng, Z. (2020). Cross-sectional studies strengths, weaknesses, and recommendations. Chest, 158(1), S65–S71. 10.1016/j.chest.2020.03.012 Wooldridge, J. M. (2015). Introductory econometrics: A modern approach (6th ed.). Nelson Education. Xiaoyang, X., Balibae, M., & Epadile, M. (2022). The impact of technological innovation, research and development, and energy intensity on carbon emissions: An experience from BRICS and OECD countries. International Journal of Sustainable Development & World Policy, 11(1), 1-17. DOI:10.18488/26.v11i1.2898 Yacoubian, L. J., Garcia, D. S., Silva, B. N., Ferreira de Carvalho, G. H., & Ketenchian, G. (2025). Peace and prosperity: An empirical analysis using GDP and GDP per capita. Revista JRG de Estudos Academicos, 8(18), 1-15. DOI:10.55892/jrg.v8i19.2372 Yan, Z., Yi, L., Du, K., & Yang, Z. (2017). Impacts of low-carbon innovation and its heterogeneous components on CO2 emissions. Sustainability, 9(4), Article 548. h<ps://doi.org/10.3390/su9040548 Zhang, W., Zhu, B., Li, Y., & Yan, D. (2024). Revisiting the Porter Hypothesis: A multi-country meta-analysis of the relationship between environmental regulation and green innovation. Humanities and Social Sciences Communications, 11, Article 232. DOI: 10.1057/s41599-024-02671-9 76