Evaluating climate-related financial policies’ impact on decarbonization with machine learning methods

Evaluating climate-related financial policies’ impact on decarbonization with machine learning methods

Mapping climate-related financial policy evolution: sequences, transitions, and policy sequencing scores

In this study, a policy sequence refers to the chronological progression of climate-related financial policies (CRFPs), including Green Bonds (GB), Green Credit Allocation Policies (GCA), Green Prudential Policies (GPP), Green Financial Guidelines (GFG), and Other Disclosure Requirements (OGD). By analyzing these sequences, we identify patterns and transitions that reveal their implications for decarbonization and financial stability.

As shown in Fig. 1, countries with longer sequences demonstrate a structured and diverse approach to policy implementation. For instance, Brazil, China, Indonesia, and Vietnam prioritize policies like GCA and GPP early in their sequences, establishing a stable financial system before adopting market-based instruments such as green bonds and guidelines. This strategic progression highlights their focus on balancing stability with green investment. Advanced economies like the United Kingdom and France exhibit extensive sequences, starting with OGD to enhance transparency and investor confidence, followed by green guidelines and market-oriented measures like green bonds. In contrast, countries with shorter sequences usually focus on one or two policy types, reflecting targeted strategies often constrained by limited resources or specific priorities.

Fig. 1
figure 1

Policy implementation sequences across countries. The order of policy implementation for various countries is categorized into five policy types: Green Bonds (GB), promoting green lending through bonds; Green Credit Allocation Policies (GCA), encouraging green lending and investments; Green Prudential Regulations (GPP), identifying and protecting against climate-related financial risks; Green Financial Principles (GFG), supporting the development of green financial markets; and Other Disclosure Requirements (OGD), encouraging public disclosure of climate-related financial risks.

To further understand the evolution of CRFPs, we analyze the likelihood of transitioning from one policy type to another, as depicted in the Transition Probability Matrix (TPM) in Panel A of Fig. 2. The matrix highlights the probabilities of moving from one policy type (rows) to another (columns). The transitions with the highest probabilities include from GPP (Green Prudential Policies) to GFG (Green Financial Guidelines), as indicated by the brightest cell, followed by notable transitions such as from GCA (Green Credit Allocation) to GFG and from OGD (Other Green Disclosure) to GFG. These patterns suggest that GFG often serves as a common follow-up policy type, reflecting its likely role in complementing or reinforcing other policy measures within the CRFP framework.

Fig. 2
figure 2

(A) Policy Transition Matrix (left panel): probabilities of transitioning from one policy type (From Policy) to another (To Policy). The highest transition probabilities are indicated by the brightest colors (green and yellow) in the heatmap, representing the highest transition probability among all combinations of policy types. (B) Policy Sequencing Scores for Top 20 Countries (right panel): evolution of Policy Sequencing Scores (PSS) over time for the top 20 countries (y-axis) based on their average scores. Darker shades of blue represent higher PSS values, revealing differences in policy intensity and trends across countries and years.

Panel B in Fig. 2 illustrates the evolution of Policy Sequencing Scores (PSS) for the top 20 countries from 2000 to 2023. The PSS quantifies the systematic adoption and structuring of CRFPs, revealing significant disparities across countries. Advanced economies like Australia, Japan, France, and the United Kingdom maintain consistently high PSS values, reflecting early leadership and sustained efforts to integrate green finance into policy frameworks. EMDEs like Brazil, China, and Indonesia also show early and notable growth of policy sequences. China, in particular, has experienced rapid acceleration since 2015, signaling a concentrated push toward comprehensive policy frameworks. In contrast, countries like Canada, Italy, and Singapore display late and slower progress.

How policy sequencing and bindingness-weighted adoption drive climate transitions

To study how the sequencing, duration, and binding nature of CRFPs influence key decarbonization indicators, such as CO2 emissions and renewable energy production (REP), we rely on the Policy Sequencing Score (PSS) and Cumulative Policy Index \(\times\) Bindingness (CumBind). They are complementary metrics that capture different aspects of climate-related financial policy (CRFP) adoption. The PSS evaluates how closely a country’s policy adoption aligns with typical global sequences. It is calculated by summing the conditional frequencies of policies adopted up to a given year, where conditional frequencies represent the likelihood of one policy preceding another globally. A higher PSS indicates that a country’s policies are implemented in a structured and globally consistent order, reflecting strategic alignment with best practices. The CumBind accounts for both the quantity and enforceability of policies. It is computed by summing the bindingness-weighted policies adopted over time, where Bindingness reflects a policy’s enforceability (e.g., mandatory, voluntary, or non-binding). This index highlights the intensity and effectiveness of a country’s policy framework by incorporating the strength of each policy.

Through SHAP summary plots and fitted curves, in this section, we assess the contributions of these policy metrics across four country groups: Emerging Markets and Developing Economies (EMDEs), Advanced Economies (ADV), G20, and OECD countries. The results reveal how variations in economic structures, institutional capacity, and policy frameworks influence the impact of policy measures in achieving decarbonization.

How policy sequencing and cumulative bindingness shape \(\textbf CO_2\) emissions: evidence across institutional contexts We propose that countries with longer sequences (PSS) of climate-related financial policies (CRFPs) and higher implementation stringency—reflected in elevated CumBind scores—are associated with significant increases in renewable energy production (REP) and reductions in CO2 emissions. This proposition is supported by evidence from other policy domains41,42, which demonstrates that extended policy sequences foster stable business environments and stimulate greater investment in clean energy technologies, both of which are critical for transitioning to renewable energy systems and mitigating emissions.

To explore whether countries with longer policy sequences (high PSS) and higher implementation stringency (high CumBind) align with these environmental outcomes, we utilize machine learning models to uncover patterns and relationships between CRFPs and renewable energy production or CO2 emissions reductions. By leveraging SHAP values, we gain interpretable insights into the relative contributions of policy characteristics to these outcomes.

Figure 3 presents SHAP summary plots illustrating the impact of PSS and CumBind on \(\hbox CO_2\) emissions across the four country groups. The SHAP values measure the direction and magnitude of each feature’s contribution to \(\hbox CO_2\) emissions, ranking them by importance.

Fig. 3
figure 3

Impact of Policy Sequencing Score (Panels A–D) and Cumulative Adoption with Bindingness (Panels E–H) on \(\textbf CO_2\) Emissions across country groups, measured using SHAP summary plots. Panel A (EMDEs) highlights GDP, vulnerability, and openness as dominant factors, with PSS contributing modestly. Panel B (Advanced Economies) shows GDP, openness, and readiness as the most influential factors, with PSS playing a significant role in reducing emissions. Panel C (G20 countries) emphasizes GDP, openness, and domestic credit, with PSS aiding consistency in diverse economies. Panel D (OECD countries) identifies GDP, openness, and readiness as key drivers, with PSS providing steady contributions. In Panels E-H, cumulative bindingness gains prominence: in Panel E (EMDEs), GDP, vulnerability, and openness dominate, but cumulative bindingness adds meaningful impact; Panel F (Advanced Economies) highlights its strong role alongside GDP and readiness; Panel G (G20 countries) shows its influence tapering after initial contributions; and Panel H (OECD countries) shows cumulative bindingness providing modest but consistent support in advanced institutional contexts.

In EMDEs (Panel A), GDP, vulnerability, and openness are the primary drivers of CO2 emissions, with PSS playing a secondary role by stabilizing policy environments. Vulnerability highlights the significant influence of external risks on emissions outcomes. When considering CumBind (Panel E), GDP, vulnerability, and openness remain dominant factors, while CumBind has a modest but meaningful impact, underscoring the importance of policy enforcement despite institutional constraints. In Advanced Economies (Panel B), GDP, openness, and readiness emerge as the main determinants of emissions, with PSS significantly enhancing emissions reductions by fostering stable conditions for long-term investments. Similarly, in the context of CumBind (Panel F), GDP, openness, and readiness lead the way, with CumBind playing a critical role in transforming policies into effective emissions reductions. In G20 countries (Panel C), GDP, openness, and domestic credit are key contributors to CO2 emissions, while PSS ensures consistent policy frameworks across these diverse economies. For CumBind (Panel G), GDP, openness, and domestic credit remain dominant, with CumBind providing secondary but relevant support, especially in countries with varying institutional capacities. For OECD countries (Panel D), GDP, openness, and readiness are the leading factors, with PSS offering stable support for emissions reductions in institutionalized settings. When examining CumBind (Panel H), GDP, openness, and readiness continue to play a central role, while CumBind facilitates emissions reductions by strengthening policy enforcement within mature institutional frameworks.

We also analyze the SHAP values, which quantify the contributions of each feature discussed above to emissions outcomes, revealing distinct trends for each group. Results are shown in Fig. 4.

Fig. 4
figure 4

Impact of Policy Sequencing Score (PSS) and Cumulative Adoption with Bindingness on \(\textbf CO_2\) Emissions across country groups. Panels (A to D) (PSS) illustrate the diminishing returns of sequencing: in EMDEs (Panel A), the relationship is logarithmic with SHAP values stabilizing at higher scores; Advanced Economies (Panel B) show an exponential decay pattern stabilizing around 0.03; G20 countries (Panel C) exhibit a similar pattern with stabilization near − 0.01; and OECD countries (Panel D) display a flatter exponential relationship, reflecting a minor but consistent role for sequencing. Panels E to H (Cumulative Adoption with Bindingness) highlight varying impacts of policy enforcement: in EMDEs (Panel E), SHAP values rise logarithmically, peaking before stabilizing; Advanced Economies (Panel F) show a negative exponential trend stabilizing around − 0.16; G20 countries (Panel G) demonstrate an initial increase that converges near − 0.01; and OECD countries (Panel H) exhibit a flatter, steady relationship with bindingness converging at approximately − 0.01. These patterns underscore the varying impact of policy measures across economic and institutional contexts.

In EMDEs (Panel A), the fitted curve shows a logarithmic increase, where higher sequencing scores are associated with a positive but diminishing impact on CO2 emissions. This suggests that while sequencing contributes to emissions reductions in other country groups, its effectiveness is hampered in EMDEs, likely due to structural and institutional constraints. Similarly, for CumBind in EMDEs (Panel E), the fitted curve indicates a logarithmic relationship, with higher levels of policy bindingness initially driving steep emissions reductions before plateauing. This highlights the importance of policy adoption and enforcement, even as institutional limitations temper their long-term impact. In Advanced Economies (Panel B), the relationship exhibits exponential decay, where sequencing significantly reduces emissions, stabilizing near a marginal effect of 0.03 at higher scores. Binding policies in these economies (Panel F) display a negative exponential relationship, where emissions are significantly reduced, stabilizing at approximately − 0.16, reflecting the effectiveness of stringent policy implementation. For G20 countries (Panel C), the fitted curve suggests that the impact of sequencing stabilizes at around − 0.01, indicating that beyond a certain point, additional sequencing has minimal incremental effect. A similar pattern is observed for binding policies in G20 countries (Panel G), where SHAP values decline to – 0.01, signaling a sustained but moderate effect of policy enforcement. The trend in OECD countries (Panel D) is relatively flat, stabilizing near − 0.01, suggesting that sequencing plays a consistent yet modest role in emissions reductions. For CumBind in these countries (Panel H), the relationship remains consistently negative, stabilizing near − 0.01. This reflects the modest but reliable contribution of binding policies to emissions reductions within the highly institutionalized settings of OECD economies.

How policy sequencing and cumulative bindingness shape renewable energy transitions: evidence across institutional contexts

In this section, we examine the impact of PSS and CumBind on renewable energy production (REP) across the four country groups, with SHAP summary plots presented in Fig. 5.

Fig. 5
figure 5

SHAP summary plots illustrating the impact of Policy Sequencing Score (Panels AD) and Cumulative Adoption with Bindingness (Panels EH) on Renewable Energy Production across country groups. Panel (A) (EMDEs) highlights GDP, vulnerability, and PSS as key contributors. Panel (B) (Advanced Economies) identifies GDP, openness, and PSS as dominant factors. Panel (C) (G20 countries) emphasizes GDP, openness, and PSS score to the private sector, while Panel D (OECD countries) shows GDP, vulnerability, and PSS as leading contributors. In Panels (EH), cumulative bindingness plays an increasingly significant role: Panel (E) (EMDEs) highlights GDP and domestic credit as dominant, with cumulative bindingness contributing; Panel F (Advanced Economies) shows GDP, vulnerability, and cumulative bindingness as important; Panel (G) (G20 countries) reflects GDP, openness, and cumulative bindingness as key; and Panel (H) (OECD countries) reveals steady contributions from GDP, vulnerability, and cumulative bindingness. These plots underscore the varying importance of PSS and enforcement across institutional and economic contexts.

For EMDEs (Panel A), GDP, PSS, domestic credit, and vulnerability emerge as key drivers of REP. Structural economic factors like GDP and credit access dominate, while PSS contributes by stabilizing the policy environment. Vulnerability reflects the critical role of climate and economic risks in shaping renewable energy transitions. In Panel E, GDP, domestic credit, and vulnerability remain major drivers, while CumBind plays a meaningful but secondary role. For Advanced Economies (Panel B), GDP, openness, and PSS are the primary contributors to REP. Economic capacity and international integration drive renewable energy growth, with structured policies fostering stability for long-term investments. Vulnerability plays a smaller role, reflecting fewer external constraints, though resilience planning remains important. In Panel F, GDP, vulnerability, and CumBind are significant factors, with strong policy enforcement amplifying renewable energy development within robust institutional frameworks. For G20 countries (Panel C), GDP, openness, and domestic credit dominate REP outcomes, with PSS providing consistent policy frameworks that align stakeholders in these complex economies. Similarly, Panel G highlights GDP, openness, and vulnerability as key drivers, with CumBind enhancing renewable energy outcomes by addressing diverse institutional challenges. For OECD countries (Panel D), GDP, vulnerability, and PSS are the main contributors, emphasizing the link between economic scale and renewable energy growth while highlighting the importance of policy stability within institutionalized contexts. In Panel H, GDP, vulnerability, and CumBind drive REP, demonstrating the importance of rigorous policy enforcement even in advanced renewable energy systems (Fig. 5).

Fig. 6
figure 6

Relationship between Policy Sequencing Score (PSS) and Cumulative Adoption with Bindingness and their contributions to renewable energy production across country groups. SHAP values measure the impact of these policy dimensions, highlighting logarithmic trends with diminishing returns. Panels (A and E) (EMDEs) show steep initial increases in the contributions of PSS and cumulative bindingness, which stabilize at higher levels due to institutional constraints. Panels (B and F) (Advanced Economies) indicate more sustained impacts, with diminishing returns reflecting strong institutional frameworks. Panels (C and G) (G20 countries) exhibit significant initial impacts that level off, reflecting diverse governance structures. Panels (D and H) (OECD countries) reveal steady but modest contributions consistent with mature policy and energy systems.

We also analyze the SHAP values, which quantify the contributions of each feature discussed above to emissions outcomes, revealing distinct trends for each group. Results are shown in Fig. 6. For EMDEs, PSS (Panel A) shows a logarithmic increase, with sequencing initially contributing significantly to renewable energy production but with diminishing returns at higher levels, likely due to structural and institutional constraints. Similarly, CumBind (Panel E) demonstrates steep initial gains in renewable energy production, followed by stabilization, reflecting the critical role of early policy enforcement even in the face of institutional challenges. In Advanced Economies, PSS (Panel B) exhibits exponential decay, with sequencing driving substantial improvements in renewable energy production before stabilizing at higher scores, indicating reduced marginal returns. CumBind (Panel F) shows a strong and sustained negative relationship with emissions, with significant early reductions stabilizing at higher bindingness levels, underscoring the effectiveness of policy enforcement in institutionalized contexts. For G20 countries, PSS (Panel C) reveals strong initial impacts that taper off, reflecting a consistent framework across diverse governance structures but with limited incremental benefits at higher sequencing levels. Similarly, CumBind (Panel G) shows significant early gains in renewable energy production that stabilize over time, suggesting a moderate but sustained effect of binding policies, even across heterogeneous economies. In OECD countries, PSS (Panel D) presents a flatter trend, with sequencing providing modest but stable contributions to renewable energy production within mature institutional environments. CumBind (Panel H) maintains a consistently negative impact, with steady and reliable support for emissions reductions, reflecting the strong institutional capacity of OECD countries to enforce and sustain binding policies.

A synthesis of the overall policy impacts across different economic and regional contexts

The analysis and SHAP plot results (Fig. 7) provide a valuable overview of the decarbonization efforts of EMDEs, ADV, G20, and OECD countries and allow for a comprehensive comparison. These findings illustrate how cumulative bindingness-weighted policies and PSS influence emissions (Panels A and C) and REP (Panels B and D), reflecting the structural, institutional, and economic characteristics of each group.

Fig. 7
figure 7

Fitted SHAP values: impact of the policy sequencing score and bindingness-weighted adoption on CO\(_2\) emissions (Panels A and C) and renewable energy production (Panels B and D) across country groups (EMDE, ADV, G20, OECD). The results highlight differing trends based on structural, institutional, and economic contexts, with notable contrasts between EMDEs and advanced economies.

SHAP analysis of policy sequencing and bindingness across country groups CumBind policies exhibit varied impacts across country groups. In EMDEs, such policies are associated with short-term increases in emissions due to the prioritization of financial stability over decarbonization and the absence of robust taxonomies and enforcement mechanisms (Panel C). For instance, policies like green bonds and credit allocation measures may channel financial support to carbon-intensive sectors without strict green standards13,43. Over time, the SHAP value curves stabilize, indicating reduced sensitivity of emissions to additional cumulative bindingness (Panel C). For renewable energy production, CumBind in EMDEs (Panel D) shows a steep initial association with increased production, but this effect levels off as cumulative policies increase, reflecting the structural and institutional constraints in these economies.

CumBind is associated with lower emissions (Panel C), supported by mature financial systems and institutional structures in advanced economies, including ADV and OECD countries. The declining SHAP values suggest that higher levels of bindingness are increasingly correlated with emissions reductions as policies accumulate. For renewable energy production, advanced economies (Panel D) show flatter SHAP curves, indicating more stable relationships between CumBind and renewable energy outcomes, with diminishing incremental contributions at higher levels of bindingness.

The Policy Sequencing Score (PSS) reveals disparities in how emissions and renewable energy production respond across country groups. In EMDEs, higher PSS levels are linked to limited or even positive correlations with emissions in early stages (Panel A). This reflects challenges like weak enforcement and structural reliance on carbon-intensive industries, which can dilute the impact of sequential measures. Over time, the SHAP values flatten, suggesting diminishing marginal impacts of further sequencing as structural barriers persist. For renewable energy production (Panel B), PSS shows initial positive associations that taper off at higher levels, consistent with the limitations posed by institutional and economic challenges.

In advanced economies, PSS is consistently associated with lower emissions across its range (Panel A). The SHAP values highlight a more predictable relationship, as sequencing advances alongside decarbonization pathways. For renewable energy production (Panel B), PSS shows stable and positive associations with increased renewable output, reflecting the structured environments in which these policies are applied.

For G20 countries, CumBind (Panel C) displays early associations with reduced emissions that stabilize over time, likely reflecting the diversity of institutional capacities within the group. For renewable energy production (Panel D), the relationship with CumBind is positive but less pronounced, as the institutional heterogeneity within the G20 moderates the relationship. PSS in G20 countries (Panel A) shows consistent associations with reduced emissions, though the relationship flattens at higher sequencing levels. For renewable energy (Panel B), PSS shows moderate gains, particularly in advanced economies within the group, while emerging economies display more varied outcomes.

In OECD countries, CumBind (Panel C) is associated with consistently lower emissions, with SHAP values stabilizing as bindingness increases. For renewable energy production (Panel D), CumBind exhibits a relatively flat relationship, reflecting limited variation as renewable energy integration becomes saturated. PSS (Panel A) shows steady negative correlations with emissions, with SHAP curves flattening at higher sequencing levels. For renewable energy production (Panel B), PSS is positively associated with stable and predictable increases in renewable energy, aligning with the institutional capacity and advanced policy frameworks of OECD countries.

Regional variations in the influence of policy sequencing and bindingness on CO2 emissions and renewable energy production Finally, we analyze regional differences in the impact of PSS and bindingness-weighted adoption on CO2 emissions and REP, as shown in Fig. 8. Panels A and B highlight the influence of the PSS, which varies across regions in both magnitude and functional form.

Fig. 8
figure 8

Fitted SHAP values illustrating the impact of the policy sequencing score and the bindingness-weighted adoption on CO\(_2\) emissions and renewable energy production across different country groups. Panel (A) displays the SHAP values for the effect of the PSS score on \(\hbox CO_2\) emissions for regions: Latin America and the Caribbean (LAC), Europe and Central Asia (ECA), Sub-Saharan Africa (SSA), South Asia (SA), and East Asia and the Pacific (EAP). Panel (B) shows the SHAP values for the PSS score’s impact on renewable energy production for the same regions. Panels C and D depict the SHAP values for the effect of cumulative bindingness-weighted policy adoption on CO\(_2\) emissions (Panel C) and renewable energy production (Panel D) across these regions. The fitted SHAP values are calculated using logarithmic transformations, capturing region-specific variations in the influence of policy adoption and sequencing.

In Panel A, East Asia and the Pacific (EAP) shows the steepest decline in SHAP values for CO2 emissions as PSS increases, indicating a strong relationship between sequencing and emissions reductions. This may reflect the region’s lower initial baseline for policy implementation, where small improvements in PSS correspond to notable changes in emissions. By contrast, Latin America and the Caribbean (LAC) display a steep increase. Panel B reveals a nonlinear positive relationship between PSS and REP in Sub-Saharan Africa (SSA), suggesting that renewable energy production is highly responsive to sequencing improvements. In East Asia and the Pacific (EAP), the trend is flatter, reflecting a less pronounced relationship, potentially due to the region’s advanced renewable energy sector, where additional policies yield diminishing marginal effects.

Panels C and D examine the effects of CumBind. Panel C shows a positive relationship between CumBind and CO2 emissions in EAP, South Asia (SA), and LAC, as reflected by increased SHAP values. This suggests that binding policies correlate with higher emissions in these regions. In contrast, Europe and Central Asia (ECA) display a downward trend, indicating a less pronounced relationship, which may reflect diminishing returns in regions with established policy landscapes. Panel D illustrates the strongest positive relationship between CumBind and REP in the regions of LAC, SSA, and ECA, as indicated by a sharp increase in SHAP values. This suggests that these regions effectively leverage binding policies to expand renewable energy, likely driven by robust technical expertise and financial capacity. In SSA, however, the increase is more gradual, which may reflect a smaller market size and slower technology diffusion.

Addressing climate risks through policies The analysis highlights significant regional and country-group differences in how climate-related financial policies relate to CO2 emissions and REP. Emerging markets and developing economies (EMDEs) face pronounced physical risks, such as extreme weather events and resource scarcity, alongside structural reliance on carbon-intensive sectors44. Transition risks, including stranded assets and economic disruptions from stricter global climate policies, also present challenges45. In these contexts, Sub-Saharan Africa (SSA) demonstrates a nonlinear positive relationship between PSS and REP, indicating that renewable energy production is highly responsive to improvements in sequencing. However, the moderate increase reflects a still small market and slower technology diffusion. Conversely, SSA shows a less pronounced relationship between CumBind and CO2 emissions, underscoring the ongoing challenges in leveraging binding policies to reduce emissions.

South Asia (SA) exhibits a positive relationship between CumBind and CO2 emissions, as SHAP values indicate that binding policies correlate with higher regional emissions. This finding suggests potential implementation gaps or the difficulty of addressing entrenched reliance on carbon-intensive sectors.

Latin America and the Caribbean (LAC) reveal contrasting dynamics. While PSS correlates with increased CO2 emissions, suggesting challenges in effective sequencing, CumBind shows a strong positive relationship with REP. This highlights the region’s ability to use binding policies to expand renewable energy, supported by technical expertise and financial capacity. The steep trends underscore the region’s responsiveness to these policies, though CO2 reductions remain elusive.

East Asia and the Pacific (EAP) shows a steep decline in SHAP values for CO2 emissions with increases in PSS, reflecting a strong relationship between sequencing and emissions reductions. This may result from the region’s lower initial baseline for policy implementation, where incremental improvements yield significant gains. However, for REP, EAP displays a flatter relationship with both PSS and CumBind, likely due to its advanced renewable energy sector, where additional policies have diminishing marginal effects.

In Europe and Central Asia (ECA), CumBind shows a downward trend in its relationship with CO2 emissions, likely reflecting diminishing returns in regions with well-established policy landscapes. However, CumBind has a strong positive relationship with REP, emphasizing the potential for binding policies to further renewable energy expansion, even in regions with more mature markets.

In advanced economies and OECD countries, lower physical vulnerabilities and established regulatory frameworks contribute to consistent decarbonization trajectories. However, flatter trends in SHAP results for REP and CO2 emissions indicate diminishing sensitivity to additional policies, reflecting policy saturation and existing infrastructure’s limitations for further improvements.

The G20 countries present a dual reality: advanced economies follow patterns observed in OECD countries, with stable emissions reductions and REP growth, while emerging economies display trends similar to EMDEs. This divergence highlights the varying capacities and challenges within the G20, as some countries benefit from established frameworks while others face institutional and financial constraints.

link