The Possibility and Improvement Directions of Achieving the Paris Agreement Goals from the Perspective of Climate Policy

Abstract

Currently, climate change and global warming have significantly impacted human life. In the context of sustainable development, achieving the goals of the Paris Agreement is both urgent and complex. This paper presents a comprehensive review of climate policies worldwide. Based on the global comprehensive climate policy database that we constructed and using global panel data from 1990 to 2019, we predicted the emission reduction due to climate policies using trend and fixed-effects models to identify areas for improvement. The research findings indicate that there is a considerable gap between current climate policies and the targets set by the Paris Agreement, both in terms of quantity and effectiveness. Economic growth and primary energy consumption contribute to increased greenhouse gas emissions, while increasing the proportion of renewable energy in electricity generation and implementing climate policies have the effect of reducing greenhouse gas emissions. Relying solely on increasing the quantity or effectiveness of policies would require an increase of 15–30 times the levels seen before 2019 to achieve the 2 °C warming target of the Paris Agreement. However, simultaneously increasing the number of policy implementations and enhancing their effectiveness would only require about a fourfold increase from the levels seen before 2019. Additionally, the results of the study on national heterogeneity demonstrate significant differences in policy effectiveness among countries, indicating substantial potential for emission reduction. Furthermore, the analysis of policy legal enforceability shows that hard law policies outperform soft law policies, suggesting that increasing the implementation of hard law policies can more effectively reduce emissions.

1. Introduction

As per the most recent Intergovernmental Panel on Climate Change (IPCC) Sixth Assessment Report (AR6) [1] released in 2022, anthropogenic net greenhouse gas emissions have persistently increased from 2010 to 2019. Moreover, cumulative net carbon dioxide emissions since 1850 have exhibited a continual rise, with the average annual greenhouse gas emissions during the period of 2010–2019 surpassing those of any preceding decade. The IPCC AR6 Working Group I report underscores that evidence derived from comprehensive climate models and observational data unequivocally points to human-generated greenhouse gas emissions as the primary catalyst behind global temperature escalation. The rampant utilization of fossil fuels such as coal and oil, coupled with industrial processes like steel and cement production, along with large-scale deforestation, significantly contribute to greenhouse gas emissions. Despite a slowdown in the growth rate of global greenhouse gas emissions over the past decade, the overall emission trajectory continued its upward trend up to 2019.

Simultaneously, nations worldwide are collaboratively striving to identify solutions, with efforts to mitigate climate change gaining momentum. Central to this endeavor is the Paris Agreement, a landmark achievement adopted in 2015. This pivotal accord mandates each participating nation to formulate and periodically update a Nationally Determined Contribution (NDC) plan, setting forth their commitments to combat climate change every five years. The periodic enhancement of NDCs underscores a progressive phase in global multilateral governance of climate change. The overarching goal of the Paris Agreement is to restrict the rise in global temperatures to within 2 °C above pre-industrial levels, with an aspirational target of limiting the increase to 1.5 °C.

Nevertheless, despite the current emission reduction targets outlined in NDCs and climate change mitigation policies, it is evident that, by 2030, cumulative emission reductions are unlikely to sufficiently constrain warming to below 2 °C, let alone meeting the more ambitious target of limiting it to less than 1.5 °C [2,3,4]. The attainment of these goals appears increasingly elusive.

During the twenty-eighth session of the Conference of the Parties (COP28) to the United Nations Framework Convention on Climate Change (UNFCCC), a comprehensive assessment of global efforts to combat climate change was conducted for the first time. The findings of this assessment revealed that progress across all facets of climate action, ranging from greenhouse gas emission reductions to bolstering resilience against climate change impacts, as well as providing support to vulnerable nations, is disappointingly sluggish. Insufficient financial and technical assistance exacerbates the challenges. Moreover, apprehensions persist regarding the feasibility of achieving the specific objectives outlined by various countries.

Additionally, the drivers of emissions exhibit significant variability across countries, further complicating the landscape of climate governance. Figure 1 illustrates the carbon dioxide emissions of different countries and the results of decomposing the emissions into four driving factors. Utilizing the Kaya decomposition method, carbon emissions are parsed into four primary factors, i.e., carbon intensity, energy intensity, per capita GDP, and population, with the growth rate computed relative to 1990 as the baseline. Economic factors, particularly GDP per capita, emerge as the predominant drivers of carbon emissions in diverse nations. In Brazil, South Africa, and India, the growth rate of carbon dioxide outpaces that of GDP per capita, indicating the worrisome synchronization with economic growth reliant on fossil fuel energy—a trend that reverberates globally. Demographic factors assume prominence as significant drivers of carbon emissions in the United States and South Africa. Notably, the population of the United States expanded by 30.52% between 1990 and 2019, while South Africa witnessed a staggering 59.12% increase in its population over the same period. Carbon intensity and energy intensity serve as pivotal driving forces behind the reduction in global emissions. Many countries have made strides in curbing carbon intensity [5], which is poised to drive future emission reductions in select nations. Carbon intensity offers insights into the structural adjustments towards renewable energy, while energy intensity serves as a gauge of efficiency. Technological advancements and economic growth serve as the primary catalysts for the marked decline in energy intensity. Presently, the carbon intensity of the European Union (EU) and Russia continues on a downward trajectory. Russia’s decline in carbon intensity can be largely attributed to the peak in primary energy consumption experienced in 1990. However, Russia’s climate policy demonstrates a phased approach, characterized by significant fluctuations in its carbon intensity growth rate.

Combining the aforementioned background, achieving the goals of the Paris Agreement is both urgent, necessary, and complex. Simultaneously, conducting comprehensive assessment and prediction of the key drivers for climate change mitigation, particularly focusing on efforts in climate policies across nations, understanding the possibility of achieving goals of the Paris Agreement, and identifying areas for improvement in each country, are crucial for further enhancing climate governance. Therefore, our research further expands in this area, drawing from past climate practices to identify critical factors and shortcomings in climate policies, and proposing possibilities for improvement, aiming to provide policy recommendations for climate governance in various countries.

The remaining sections of this paper are as follows: Section 2 introduces relevant literature research and summarizes research gaps. Section 3 provides an explanation of research methods and data, including an introduction to the models used and clarification of core variables. Section 4 analyzes the forecast results of emission reduction from global climate mitigation policies. We make trend assumptions for core variables and utilize models to predict emissions under different scenarios. These predictions are then compared with the goals of the Paris Agreement to identify the disparities in policy quantity and effectiveness. Section 5 summarizes and discusses the aforementioned content and provides some policy improvement suggestions.

2. Literature Review

Currently, the existing research on climate policy assessment and prediction can be categorized into two main types. The first type is ex-post studies, which assess the emission reduction impact of individual policies through retrospective evaluations. These studies typically focus on evaluating the implementation of specific policies [6,7,8,9,10]. For example, Cui et al. [10] utilized the matched Difference-in-Differences (DID) method on a panel of enterprises in China’s manufacturing and utility sectors from 2009 to 2015 to assess the effectiveness of China’s regional carbon Emissions Trading System (ETS) pilot. Their research findings indicate that China’s ETS pilots are generally effective in reducing emissions and emission intensity, with significant heterogeneity observed across different pilot programs. Zhu et al. [9] applied synthetic control and event study methods to investigate the mitigation effects, pathways, and ancillary impacts of four distinct pilot projects and three regions in China. Their findings suggest that carbon market pilots can effectively mitigate air pollution, and they also observed spillover effects of emission reductions.

Another type of research primarily focuses on ex-ante assessment, often utilizing integrated assessment models (IAMs) to simulate and evaluate various targets and emission scenarios. Typical scenarios commonly include Nationally Determined Contributions (NDCs) and the temperature targets set by the Paris Agreement. Building upon these typical scenarios, the further refinement of research can be conducted, such as rating the credibility of commitments like net-zero targets, to derive emission scenarios under different levels of credibility [11]. Roelfsema et al. [3] focused on specific policies of major economies and incorporated them into the parameters of an integrated assessment model, thereby expanding research on corresponding scenarios. The research findings indicate that, although countries are making progress towards their NDC targets, there remains a significant gap between the 2 °C and 1.5 °C targets. Consequently, countries should expedite the implementation of policies promoting renewable energy technologies. Fekete et al. [12] conducted a comprehensive review of the primary performance indicators for sectors such as buildings, transportation, and industry in China, the European Union, India, Japan, and the United States. They used the Integrated Model to Assess the Global Environment (IMAGE) to estimate greenhouse gas emissions for each sector under specific conditions. The study indicates significant disparities among countries in terms of their future policy expectations and current historical performance. Moreover, replicating successful sector-level policies and their performance across countries could accelerate progress in greenhouse gas emission reduction efforts. Roelfsema et al. [13] subdivided policy assumptions and incorporated policy objectives and policy instrument targets into IMAGE, thereby expanding the scenarios for climate policy assessment. Some studies focus on specific countries or regions to assess the gap in climate mitigation policies among these nations [14,15,16,17,18,19,20,21]. Den Elzen et al. [14] focused on the NDCs of G20 countries, suggesting that the greenhouse gas emissions of Brazil, Indonesia, Mexico, and South Korea will peak by 2025, while those of China, India, and South Africa will peak by 2030 or later. Furthermore, den Elzen et al. [21] have been continuously tracking this and comparing the NDCs of G20 economies with their projected greenhouse gas emissions based on current climate policy scenarios. The study suggests that six member states, including China and India, can achieve their unconditional NDCs, while eight member states, including Argentina and the United States, need to further strengthen their climate policy actions.

Meanwhile, other studies have opted not to use integrated assessment models but rather employ methods that are more interpretable. For instance, Liobikienė and Butkus [22] focused on a detailed examination of scenarios regarding renewable energy targets and energy consumption. Their research revealed that even under assumptions of rapid economic growth, the European Union could achieve its strategic goals for 2020 by reducing energy consumption and increasing the proportion of renewable energy.

Recently, scholars have been paying attention to policy density [23,24,25,26,27], which to some extent can reflect a country’s policy ambition [26]. Using the quantity of climate policies as a core variable has expanded the research scope of policy effects and their relationship with greenhouse gases. In this regard, Eskander and Fankhauser [23] are pioneers in this research field. They conducted a study on a global panel from 1999 to 2016 based on climate change law data, suggesting that newly enacted climate laws can reduce emission intensity by 0.78%. Additionally, Chen et al. [24] further investigated the impact of energy policies on Sustainable Development Goals (SDGs), noting that energy policies can effectively contribute to achieving SDGs. Moreover, they found significant heterogeneity in policies across different SDGs.

Through the literature review above, we can derive the following insights: Firstly, regarding the emission reduction characteristics of climate policies, most existing studies based on NDCs’ scenarios suggest that there is still a gap between current policies and the achievement of the temperature targets of the Paris Agreement. By considering subdivisions of different scenarios, such as scenarios focusing on sectors that have successfully implemented policies, recommendations can be made to narrow this gap. Secondly, when analyzing global emission reduction pathways and achieving long-term goals, integrated assessment models can provide a more comprehensive analysis, considering the interactions of various factors, thus possessing evident advantages.

However, this analytical approach also has some limitations. Firstly, the transparency of the model is often criticized, and establishing better connections with social sciences remains a challenging task. Secondly, when the focus of the study is on the feasibility and realism of climate policies, the model still needs improvement [13]. For example, in integrated assessment models, carbon pricing does not represent the policy tool of carbon pricing but rather an idealized marginal abatement cost. On the other hand, a single policy instrument or target cannot fully represent realism, so in some cases, the impact of all implemented policies needs to be considered, which is still an unresolved issue. Lastly, compared to trend extrapolation assessment methods, integrated assessment models typically have higher complexity and weaker interpretability. In summary, in the field of policy assessment and prediction, both integrated assessment models and trend extrapolation methods have their advantages and disadvantages. Therefore, when choosing a method, specific usage scenarios and research needs should be considered. Existing studies have set relatively crude scenarios for climate policies, mostly based on scenarios such as conditional and unconditional NDCs, existing policies, and policies from 2010. This has resulted in a lack of comprehensive assessment of all climate policies. Due to insufficient data on climate policies, the current research can only address the emission gap in terms of the goals of the Paris Agreement but cannot assess the effectiveness and quantity of policies to achieve these goals. Therefore, in this paper, we will fill this gap.

3. Data and Methods

3.1. Emission Gap Model

As mentioned above, the main objective of this study is to forecast emission reductions to assess the probability of achieving the Paris Agreement goals and to identify deficiencies in the current climate policies. Drawing on the greenhouse gas (GHG) emission forecasting method from the existing research [22], the key difference in our study lies in its focus on global climate policies. Therefore, we incorporate policy factors into this methodology. Trend estimation methods are utilized to assess the trends of core variables, and greenhouse gas emissions are predicted after estimating the coefficients of core explanatory variables using a fixed-effects model. The above-mentioned model has two main advantages. Firstly, it exhibits strong explanatory power and employs a relatively concise approach. Through rational model settings, different types of core explanatory variables (percentages, quantities, etc.) undergo various functional transformations, endowing them with economic and statistical interpretative significance. Secondly, the model can adjust focus variables and intuitively address how these variables need to change to achieve the goals of the Paris Agreement. Specifically, a linear form of the following equation is used for trend estimation of core variables:

$${\mathrm{y}}_{\mathrm{i},\mathrm{t}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1\mathrm{t}+{\mathsf{\epsilon }}_{\mathrm{t}}$$

(1)

where ${\mathrm{y}}_{\mathrm{i},\mathrm{t}}$ represents the core variable of interest for country $\mathrm{i}$ in year $\mathrm{t}$, and we focus on ${\mathrm{l}\mathrm{n}\mathrm{P}\mathrm{G}\mathrm{D}\mathrm{P}}_{\mathrm{i},\mathrm{t}}$ (per capita GDP), ${\mathrm{l}\mathrm{n}\mathrm{P}\mathrm{E}\mathrm{C}}_{\mathrm{i},\mathrm{t}}$ (primary energy consumption), ${\mathrm{P}\mathrm{O}\mathrm{L}}_{\mathrm{i},\mathrm{t}}$ (cumulative climate policies), and ${\mathrm{R}\mathrm{E}\mathrm{S}}_{\mathrm{i},\mathrm{t}}$ (proportion of renewable energy generation). For the first two core variables, due to the logarithmic transformation, ${\mathsf{\beta }}_1$ can be interpreted as the percentage change in per capita GDP and primary energy consumption for every one-year increase. For the latter two, ${\mathsf{\beta }}_1$ represents the change in the cumulative climate policies and the proportion of renewable energy generation as the year increases. ${\mathsf{\beta }}_0$ and ${\mathsf{\epsilon }}_{\mathrm{t}}$ represent the intercept and random error term, respectively. By using Equation (1), trends in core variables can be predicted, and subsequently, the following Equation (2) can be used to estimate and forecast greenhouse gas emissions:

$${\mathrm{l}\mathrm{n}\mathrm{G}\mathrm{H}\mathrm{G}}_{\mathrm{i},\mathrm{t}}={\mathsf{\alpha }+\mathsf{\eta }}_1{\mathrm{l}\mathrm{n}\mathrm{P}\mathrm{G}\mathrm{D}\mathrm{P}}_{\mathrm{i},\mathrm{t}-1}+{\mathsf{\eta }}_2{\mathrm{l}\mathrm{n}\mathrm{P}\mathrm{E}\mathrm{C}}_{\mathrm{i},\mathrm{t}-1}+{\mathsf{\eta }}_3{\mathrm{P}\mathrm{O}\mathrm{L}}_{\mathrm{i},\mathrm{t}-1}+{\mathsf{\eta }}_4{\mathrm{R}\mathrm{E}\mathrm{S}}_{\mathrm{i},\mathrm{t}-1}+\mathsf{\gamma }{\mathrm{H}\mathrm{Y}\mathrm{R}\mathrm{D}}_{\mathrm{i},\mathrm{t}-1}+{\mathsf{\theta }}_{\mathrm{i}}{+\mathsf{\delta }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i}\mathrm{t}}$$

(2)

where ${\mathrm{l}\mathrm{n}\mathrm{G}\mathrm{H}\mathrm{G}}_{\mathrm{i},\mathrm{t}}$ represents the greenhouse gas emissions of country $\mathrm{i}$ in year $\mathrm{t}$, in addition to the aforementioned core variables, adding ${\mathrm{H}\mathrm{Y}\mathrm{R}\mathrm{D}}_{\mathrm{i},\mathrm{t}-1}$(proportion of hydroelectric power generation) as a control variable to supplement the proportion of renewable energy generation. Since both the independent and dependent variables are in the logarithmic form, ${\mathsf{\eta }}_1$ and ${\mathsf{\eta }}_2$ can be interpreted as the percentage change in greenhouse gas emissions resulting from a 1% change in per capita GDP and primary energy consumption, respectively. For ${\mathsf{\eta }}_3$ and ${\mathsf{\eta }}_4$, they, respectively, explain the percentage change in greenhouse gas emissions resulting from an increase of one climate policy and a 1% increase in the proportion of renewable energy generation, expressed as $100\ast {\mathsf{\eta }}_3\mathrm{\%}$ and $100\ast {\mathsf{\eta }}_4\mathrm{\%}$. The fixed-effects model can eliminate unobservable heterogeneity, where ${\mathsf{\theta }}_{\mathrm{i}}$ and ${\mathsf{\delta }}_{\mathrm{t}}$ represent country fixed effects and year fixed effects, respectively. $\mathsf{\alpha }$ and $\mathsf{\gamma }$ represent the intercept and the coefficients of the control variables in the model, while ${\mathsf{\epsilon }}_{\mathrm{i}\mathrm{t}}$ is the random error term. All explanatory variables and control variables are lagged by one period to avoid endogeneity resulting from mutual causality. Through the aforementioned setup, the core variables in Equations (1) and (2) are placed within a unified prediction framework.

3.2. Data Source and Descriptive Statistics

To comprehensively review climate policies and thoroughly evaluate the effectiveness of emission reduction policies on a global scale, we have harmonized three renowned climate policy databases as of 31 December 2021: the Climate Change Laws of the World (CCLW) [28], the International Energy Agency (IEA) Policies and Measures Database [29], and the Climate Policy (CP) database [30]. Building upon this foundation, we have incorporated important features such as policy instruments, sectors, objectives, legal enforceability, and policy durability, culminating in the creation of a new comprehensive climate policy database called GCCMPD (Global Climate Change Mitigation Policy Dataset) [31].

Policy formulation and implementation are influenced by various structural factors, such as material endowments, political systems, ideas, values, and belief systems [32,33,34,35,36,37,38,39]. These differences lead to what we term “wide variations with similar policy instruments”. For instance, for low- and middle-income countries, the presence of laws can signal credibility to attract international funding [40]. Therefore, the adoption of soft laws and hard laws depends on the local implementation of environment and legal design. GCCMPD reflects these differences through binding force, as illustrated in

Table 1. Hard and soft law classification.

Hard and Soft Law	Classification	Secondary Classification
Hard Law	Constitution (legislative)
	Statutes/Legislation	International Law (-)
		Law/Act (almost legislative)
		Decree/Order/Ordinance (almost executive)
		Common Law/Case Law (executive)
	Regulations/Rules (almost executive)
Soft Law and Quasi-legislative	Preparatory and Informative Instruments (pre-law function)	Preparatory Instruments (pre-law function)
		Informative Instruments (partially pre-law function)
	Interpretative and Decisional Instruments (post-law function)	Interpretative Communications and Notices (post-law function)
		Decisional Notices and Communications (post-law function)
		Decisional Guidelines, Codes, and Frameworks (post-law function)
	Steering Instruments (partially para-law function)
Other Strategy Plan or Target

Source: compiled by the authors.

, encompassing various forms of hard law and soft law functionalities. The classification of hard law hierarchy and soft law functionalities draws upon the existing research classifications and is further refined and expanded [41,42].

For the ease of comparison with other studies, the dependent variable used is the logarithm of greenhouse gas emissions, sourced from the Emissions Database for Global Atmospheric Research (EDGAR). The core explanatory variables selected are as follows: (1) Per capita GDP (logged), as the issue of decoupling economic growth from greenhouse gas emissions remains a focal point of academic and industrial interests [43]. According to the Environmental Kuznets Curve (EKC), there is an inverted U-shaped relationship between economic growth and pollutant emissions, and greenhouse gases can be understood as a type of pollutant with global negative externalities, thus the relationship between greenhouse gas emissions and economic growth may be more complex [44,45,46]; (2) Primary energy consumption (logged), as there is a long-term relationship between greenhouse gas emissions and energy consumption, with fossil fuel-related activities being the major contributor to greenhouse gas emissions [47,48]; (3) Proportion of renewable energy electricity generation (excluding hydroelectric), as greenhouse gas emissions mainly come from electricity generation and heating, and traditional energy sources have much higher lifecycle greenhouse gas emissions compared to renewable energy sources [49], thus the proportion of renewable energy electricity generation serves as a key factor influencing greenhouse gas emissions; (4) Cumulative policy quantity, i.e., policy density, which can represent the level of national policy activity, reflecting climate governance ambitions, and policy density is also correlated with policy stringency [26]. Recent studies have shown a negative relationship between policy density and greenhouse gas emissions [23,24], thus this variable is chosen as a core variable to explain the effectiveness of climate mitigation policies and the emission gap between policies and the Paris Agreement scenarios. Additionally, based on the policy classification mentioned earlier, cumulative hard law quantity and cumulative soft law quantity are further studied. The proportion of hydroelectric power generation is selected as a control variable to supplement the proportion of renewable energy generation. Among the aforementioned variables, GDP, proportion of renewable energy electricity generation (excluding hydroelectric), and electricity production from hydroelectric sources are sourced from the World Development Indicators (WDI); population data are from the United Nations; primary energy consumption data are from the World Energy Balances; and cumulative policy quantity, cumulative hard law quantity, and cumulative soft law quantity are from the GCCMPD database. Detailed variable descriptions and measurements are depicted in

Table 2. Variable selection and description.

Variable	Definition and Description	Measurement
lnGHG	Greenhouse gases (also known as GHGs) are gases in the Earth’s atmosphere that trap heat.	Gigagramme of carbon dioxide equivalent (GgCO2e, logarithm)
lnPGDP	GDP per capita based on purchasing power parity (PPP)	Constant 2017 international dollars per capita, logarithm
lnPEC	Primary energy consumption refers to the direct use or supply at the source of energy that has not been subjected to any conversion or transformation process.	TCE (ton of standard coal equivalent, logarithm)
RES	Electricity production from renewable sources, excluding hydroelectric, includes geothermal, solar, tides, wind, biomass, and biofuels	Percentages
POL	The accumulation of policy numbers, or policy density, reflects the level of policy activity and the internal differences in policy instruments encompassed within policy areas.	Number
HARD	The accumulation of hard law climate policies encompasses legally binding rules, regulations, statutes, or provisions that are enforceable by law.	Number
SOFT	The accumulation of soft law climate policies refers to the collection of non-binding agreements, guidelines, principles, or declarations that lack the legal force of traditional “hard law”.	Number
HYRD	Electricity production from hydroelectric sources	Percentages

Source: World Development Indicators, United Nations, World Energy Balances, and GCCMPD database.

.

4. Results

4.1. Descriptive Statistical Results

Table 3. Descriptive statistical values of core variables.

Variable	Mean	Std. Dev.	Min	Max
lnGHG	0.163	0.314	0.000	2.748
lnPGDP	9.277	1.171	6.066	11.711
lnPEC	16.532	2.577	0.000	22.315
RES	2.856	6.180	0.000	65.444
HYRD	30.568	33.223	0.000	100.000
POL	40.867	86.238	0.000	634.000
HARD	16.013	34.991	0.000	235.000
SOFT	23.379	51.336	0.000	489.000

Source: compiled by the authors.

presents the descriptive statistics of the core variables in this study. We utilized panel data from 143 countries spanning from 1996 to 2019. The core database of GCCMPD covers 199 countries/entities, while the IEA World Energy Balances encompass 146 countries. Ultimately, the intersection resulted in 143 countries being considered. To broaden the sample size and mitigate estimation errors due to small sample sizes, individual countries were not further excluded. Considering the issue of missing core variables, the study period was chosen from 1990 to 2019. This period broadly covers significant milestones such as the Kyoto Protocol, the Bali Road Map, the Copenhagen Accord, and the Paris Agreement [50]. According to the descriptive statistical results, there is significant heterogeneity among countries in terms of the total number of climate policies, the number of hard law policies, and the number of soft law policies. Over the study period, on average, each country accumulated 40.87 climate policies. During this period, countries such as the United States, Canada, Australia, China, and Germany issued a relatively large number of policies, with cumulative policy releases exceeding 200. Most countries tend to adopt soft law climate policies, which is consistent with findings from some survey and statistical research [50,51]. Additionally, there are notable differences among countries in terms of the proportion of renewable energy (excluding hydropower) in electricity generation and the proportion of hydroelectric power generation.

4.2. Trend Model Results

We return to the core focus of this paper: the probability of achieving the goals of the Paris Agreement. Based on the regression results of Equation (1), we divided the trends into three categories, i.e., long-term trend (1990–2019), medium-term trend (1997–2019), and short-term trend (2009–2019), resulting in the trend assumptions presented in

Table 4. Trend regression results and scenarios of trend changes.

	Long-Term Trend		Medium-Term Trend		Short-Term Trend
	2019–2025	2019–2030	2019–2025	2019–2030	2019–2025	2019–2030
lnPGDP	15.634%	28.663%	14.999%	27.498%	11.459%	21.007%
lnPEC	21.398%	39.229%	16.634%	30.496%	9.210%	16.885%
RES %	1.127	2.066	1.815	3.328	4.233	7.760
POL (Number)	23.699	43.447	29.810	54.652	32.784	60.103

Source: estimated and compiled by the authors.

. The long-term trend of per capita GDP shows an average annual growth of 2.61%, the medium-term trend shows an average annual growth of 2.77%, and the short-term growth is 1.91%; the long-term trend of primary energy consumption shows an average annual growth of 3.57%, the medium-term trend shows an average annual growth of 2.5%, and the short-term growth is 1.53%; the long-term trend of cumulative climate policy quantity shows an average annual growth of 3.95, the medium-term trend shows an average annual growth of 4.97, and the short-term growth is 5.46; the long-term trend of the proportion of renewable energy generation shows an average annual increase of 0.188%, the medium-term trend shows an average annual increase of 0.303%, and the short-term growth is 0.705%. Except for the trend time periods, all three assumptions are derived from the trend regression results. According to the results, long-term economic growth is relatively fast, but primary energy consumption grows even faster, indicating a potential positive relationship between economic growth and energy consumption, and early energy efficiency is not high, resulting in economic growth at the expense of more energy consumption, while medium-term and short-term energy efficiency improves significantly. The proportion of renewable energy generation and the growth in the number of climate mitigation policies show similar trends, both experiencing significant increases in the short term, while long-term trends change more slowly.

Extrapolating from the short-term trend, from 2019 to 2025, on average, each country will accumulate an additional 32.784 policies, increase the proportion of renewable energy generation by 4.233%, achieve a cumulative economic growth of 11.459%, and a cumulative primary energy consumption increase of 9.210%. This analysis indicates that, after the Copenhagen Summit, there has been a significant improvement in the global climate governance process, and there is a possibility of decoupling greenhouse gas emissions from economic growth.

4.3. Greenhouse Gas Emission Scenario Prediction Results and Comparison

4.3.1. Policy Emission Reduction Effects and National Heterogeneity

Based on the extrapolation of the above trends and combined with fixed-effects regression of core variables of greenhouse gas emissions, greenhouse gas emissions under different scenarios can be predicted.

Table 5. Regression results of policy emission reduction effects.

	(1)	(2)	(3)
Variables	lnGHG	lnGHG	lnGHG
L.lnPGDP	0.061 ***	0.060 ***	0.062 ***
	(0.014)	(0.013)	(0.014)
	0.034–0.088	0.034–0.086	0.035–0.089
L.lnPEC	0.010 ***	0.010 ***	0.010 ***
	(0.003)	(0.003)	(0.003)
	0.004–0.016	0.005–0.016	0.004–0.016
L.RES	−0.001 ***	−0.000 ***	−0.001 ***
	(0.000)	(0.000)	(0.000)
	−0.001–−0.001	−0.001–−0.000	−0.002–−0.001
L.HYRD	−0.000	−0.000	−0.000
	(0.000)	(0.000)	(0.000)
	−0.000–0.000	−0.000–0.000	−0.000–0.000
L.POL	−0.008 ***
	(0.003)
	−0.014–−0.003
L.HARD		−0.040 ***
		(0.004)
		−0.048–−0.032
L.SOFT			−0.006
			(0.006)
			−0.018–0.007
Constant	−0.625 ***	−0.621 ***	−0.630 ***
	(0.139)	(0.134)	(0.140)
	−0.897–−0.353	−0.885–−0.358	−0.904–−0.356
Observations	2525	2525	2525
R-squared	0.988	0.988	0.988

Note: *** correspond to significance levels of 1%, with robust standard errors in parentheses. Source: estimated by the authors.

presents the regression results of policy emission reduction effects obtained from Equation (2), where columns (1)–(3), respectively, represent the results using cumulative total policies, cumulative hard law policies, and cumulative soft law policies as the core predictor variables. Since the primary purpose is to predict greenhouse gas emissions, confidence intervals are provided, and an R-squared of around 0.988 also indicates that the model variables can explain greenhouse gas emissions. To avoid excessively small coefficients, the cumulative policy quantities are scaled down by a factor of 100. The coefficients for cumulative total policies and cumulative hard law policies are negative and significant at the 1% level, indicating a clear emission reduction effect of climate mitigation policies. Specifically, every 100 policies and hard law policies can reduce greenhouse gas emissions by approximately 0.85% and 4.00%, respectively, while soft law policies do not have a significant emission reduction effect. From the above results, it can be observed that the current climate mitigation policies rely more on hard law to generate emission reduction effects. The results of other core variables are also as expected: per capita GDP and primary energy consumption coefficients are positive and significant at the 1% level. Taking the cumulative total policies as an example, an increase of 1% in per capita GDP and primary energy consumption would lead to an increase of approximately 0.061% and 0.010% in greenhouse gas emissions, respectively. For every 1% increase in the proportion of renewable energy generation, greenhouse gas emissions would decrease by approximately 0.11%, while increasing the proportion of hydroelectric power generation reduces greenhouse gas emissions, but it has no significant effect.

Building on the regression results mentioned above, we further investigated country heterogeneity. Unlike previous studies that employed predetermined country classification criteria, such as income brackets, we selected features closely related to greenhouse gas emissions and economic growth [25], namely, economic growth (per capita GDP) and per capita greenhouse gas emissions, to cluster countries using the K-means clustering method. Figure 2 displays the clustering results for each year from 2017 to 2019, with a Calinski–Harabasz (CH) score of 205.231 for 2019, indicating a relatively good clustering outcome. Based on Figure 2, the clustering of the same countries remains consistent from 2017 to 2019. We summarize the clustering results into four categories:

• Underdeveloped countries: These nations exhibit relatively low economic development and emissions but constitute a significant proportion (60%) of the total. They are mainly located in certain regions of Africa, South Asia, Southeast Asia, and parts of South America.
• Countries with large emission reduction potential: Representative countries include the United Arab Emirates, Australia, Saudi Arabia, Bahrain, Oman, and Brunei. These countries are typically situated in the upper-left quadrant of Figure 2, indicating that high economic growth leads to excessive emissions. Some of these countries exhibit a pronounced resource curse phenomenon [52]. Due to their high economic growth rates, they possess significant emission reduction potential.
• Developed countries: Representative countries include Germany, Norway, and the Netherlands, mainly distributed in Western Europe. Notably, the United States is also classified as a “developed country” according to the clustering results, while Canada falls into the category of “countries with large emission reduction potential”. However, the two countries are very close in Figure 2. Similarly, countries like Japan, South Korea, Russia, and Kazakhstan are positioned near category boundaries. For instance, some studies suggest that Russia exhibits a clear resource curse problem [53]; hence, these countries cannot be considered as typical examples of their respective categories.
• Developing countries: Representative countries include China, South Africa, Malaysia, etc. This category comprises the second largest proportion (23.87%) and is more widely distributed. The “developing countries” category encompasses nations such as Argentina, Mexico, Turkey, Thailand, etc., with insufficient policies. Therefore, this category also possesses some emission reduction potential. However, most countries face dual challenges of economic development and climate governance, with climate change significantly impacting them [54]. Effectively addressing both objectives has become a challenging aspect of climate policy design for these countries.

Table 6. Results of country heterogeneity in policy emission reduction effects.

	(1)	(2)	(3)
L.lnPGDP	0.063 ***	0.068 ***	0.061 ***
	(0.016)	(0.016)	(0.016)
L.lnPEC	0.012 ***	0.011 ***	0.011 ***
	(0.003)	(0.003)	(0.003)
L.RES	−0.001 ***	0.000	−0.001 ***
	(0.000)	(0.000)	(0.000)
L.HYRD	0.000	0.000	0.000
	(0.000)	(0.000)	(0.000)
L.POL.0	0.181 ***	0.155 ***	0.332 ***
	(0.020)	(0.026)	(0.022)
L.POL.1	0.016 ***	0.051 ***	0.021 ***
	(0.002)	(0.014)	(0.003)
L.POL.2	−0.016 ***	−0.056 ***	−0.025 ***
	(0.002)	(0.005)	(0.003)
L.POL.3	0.005	−0.029 ***	0.024 *
	(0.006)	(0.007)	(0.014)
Constant	−0.692 ***	−0.726 ***	−0.660 ***
	(0.161)	(0.157)	(0.159)
Observations	2232	2232	2232
R2	0.989	0.989	0.990

Note: *** and * correspond to significance levels of 1% and 10%, respectively, with robust standard errors in parentheses. Source: estimated by the authors.

presents the results of country heterogeneity tests for policy emission reduction effects based on the aforementioned clustering analysis. Columns (1) to (3) represent the core explanatory variables of policies, including cumulative total climate policies, cumulative hard legal policies, and cumulative soft legal policies. Due to the relatively small number of countries in some groups (e.g., “countries with large emission reduction potential” consists of only 11 countries), heterogeneity was tested using dummy variable methods. The results indicate that climate policies (including both hard and soft law policies) in the “developed countries” group can significantly reduce emissions. Typically, these countries have also implemented policies earlier, and this category can be understood as “leading countries in climate policies”, leading globally in policy innovation, influence, and effectiveness. In the “developing countries” group, hard law policies have a significant emission reduction effect, while soft law policies show no significant emission reduction effect. This indicates that these countries have taken action on mitigating climate change and have achieved some results, but still rely on legal constraints. As for “underdeveloped countries” and “countries with large emission reduction potential”, climate policies show no significant emission reduction effects.

4.3.2. Forecasting Results and Comparison of Greenhouse Gas Emission Scenarios

The above analysis reveals significant heterogeneity in the emission reduction effects of climate policies across different countries. We further examine the emission gap between climate policies and the targets of the Paris Agreement. Under the trend assumptions in Table 4, combined with the regression results in Table 5 and the economic statistical interpretability of the coefficients, greenhouse gas emission predictions can be made.

Table 7. Forecasts of greenhouse gas emissions in different scenarios of all policies.

Long-Term Trend Scenario
	2025			2030
	Min	Median	Max	Min	Median	Max
PGDP	0.535%	0.954%	1.373%	0.980%	1.749%	2.517%
PEC	0.083%	0.214%	0.345%	0.153%	0.393%	0.632%
REC	−0.158%	−0.122%	−0.079%	−0.289%	−0.223%	−0.145%
POL	−0.337%	−0.200%	−0.064%	−0.617%	−0.367%	−0.117%
Total GHG change	0.124%	0.846%	1.574%	0.227%	1.551%	2.886%
GHG forecast	52.623	53.003	53.386	52.678	53.374	54.075
Medium-Term Trend Scenario
	2025			2030
	Min	Median	Max	Min	Median	Max
PGDP	0.513%	0.915%	1.317%	0.940%	1.678%	2.414%
PEC	0.065%	0.166%	0.268%	0.119%	0.305%	0.491%
REC	−0.254%	−0.196%	−0.127%	−0.466%	−0.359%	−0.233%
POL	−0.423%	−0.252%	−0.080%	−0.776%	−0.462%	−0.148%
Total GHG change	−0.100%	0.634%	1.377%	−0.183%	1.162%	2.525%
GHG forecast	52.506	52.891	53.282	52.462	53.169	53.885
Short-Term Trend Scenario
	2025			2030
	Min	Median	Max	Min	Median	Max
PGDP	0.392%	0.699%	1.006%	0.718%	1.282%	1.844%
PEC	0.036%	0.092%	0.148%	0.066%	0.169%	0.272%
REC	−0.593%	−0.457%	−0.296%	−1.086%	−0.838%	−0.543%
POL	−0.466%	−0.277%	−0.089%	−0.853%	−0.508%	−0.162%
Total GHG change	−0.630%	0.057%	0.770%	−1.156%	0.105%	1.411%
GHG forecast	52.227	52.588	52.963	51.951	52.613	53.3
Extreme Policy Scenario
	2025			2030
	Min	Median	Max	Min	Median	Max
PGDP	0.392%	0.699%	1.006%	0.718%	1.282%	1.844%
PEC	0.036%	0.092%	0.148%	0.066%	0.169%	0.272%
REC	−0.593%	−0.457%	−0.296%	−1.086%	−0.838%	−0.543%
POL	−7.100%	−4.226%	−1.350%	−28.400%	−16.904%	−5.400%
Total GHG change	−7.265%	−3.892%	−0.492%	−28.702%	−16.291%	−3.827%
GHG forecast	48.740	50.513	52.300	37.473	43.996	50.547

Source: estimated by the authors.

presents the forecasted greenhouse gas emissions under different scenarios. Taking the short-term trend as an example, by 2025, climate policies are expected to reduce greenhouse gas emissions by an average of 0.277%. Firstly, based on the trend assumption in Table 4, it is estimated that climate policies will increase by 32.784 units from 2019 to 2025. Secondly, according to Table 5, for every 100 policies implemented, greenhouse gas emissions are projected to decrease by 0.85%. Therefore, the emission reduction contribution of climate policies in the short-term trend scenario can be obtained. According to the total change in greenhouse gas emissions (GHG) in Table 7, under both the long-term and mid-term trend scenarios, GHG emissions are expected to increase in 2025 and 2030 compared to 2019. The contributions of climate policies and the proportion of renewable energy in electricity generation are insufficient to offset the greenhouse gas emissions resulting from economic growth. In comparison, the lower limit of the confidence interval for greenhouse gas emissions predicted by the short-term trend scenario is lower than the actual emissions in 2019. According to the trend assumption, the proportion of renewable energy in electricity generation increased by 4.233% and 7.760%, respectively, during the periods of 2019–2025 and 2019–2030, which are the main driving factors for emission reduction.

Figure 3 depicts the comparison results between the long-term trend, mid-term trend, and short-term trend scenarios in this study and the commonly used scenarios in existing research. In Figure 3, “Current policy scenario” denotes a scenario where no additional policies are assumed, and the current policies are successfully implemented to achieve the expected emission reduction. “Unconditional NDC scenario” represents a scenario where the latest unconditional NDCs submitted by countries are assumed to be successfully implemented. “Conditional NDC scenario” represents a scenario where countries achieve their emission reductions in terms of NDCs with assistance of funding and technology. The scenarios labeled as “2.0 °C scenario”, “1.8 °C scenario”, and “1.5 °C scenario” represent greenhouse gas emission scenarios under various warming targets. The long-term trend estimation in this study yields results similar to the “Unconditional NDC scenario”, while the short-term trend estimation is similar to the “Conditional NDC scenario”. This suggests that the trend extrapolation method used in this study is reasonable. It also indicates that emission reductions in terms of NDCs consider existing policies and policy trends, although not all policies are included in the model. Nevertheless, under these scenarios, there is a significant gap in achieving the 2.0 °C target of the Paris Agreement. For instance, in the short-term trend scenario, there is approximately an 11 GtCO2e gap between the forecasted greenhouse gas emissions of 2030 and the 2.0 °C target.

Furthermore, utilizing the adjustability of the model in this study, extreme policy scenarios can be derived, determining the additional number of climate policies needed to achieve the 2.0 °C target while maintaining short-term trends in economic growth, primary energy consumption, and the proportion of renewable energy in electricity generation. The results indicate that compared to 2019, by 2025 and 2030, on average, each country would need to increase their policies by 500 and 2000, respectively. Even if the other three core factors remain unchanged, an additional 300 and 1900 policies would still be required in 2025 and 2030 to reach the 2.0 °C target. According to the descriptive statistics in Table 3, the average cumulative policies are 41, with a maximum of 634. Therefore, adding 500 and 2000 policies by 2025 and 2030, respectively, seems difficult to achieve. These calculated results align with the estimates in Table 5 and Table 6. Given that the effectiveness of the current climate policies still needs improvement, although the number of policies is a crucial factor influencing emission reduction, the effectiveness of each policy also impacts the achievability of the warming targets set by the Paris Agreement. From this perspective, we further consider hard laws as more effective policies and recalculate the probability of achieving the Paris Agreement targets. The results are shown in

Table 8. Forecasts of greenhouse gas emissions in different scenarios of hard law policies.

Long-Term Trend Scenario
	2025			2030
	Min	Median	Max	Min	Median	Max
PGDP	0.533%	0.938%	1.343%	0.977%	1.720%	2.462%
PEC	0.096%	0.214%	0.336%	0.177%	0.392%	0.616%
REC	−0.079%	−0.045%	−0.011%	−0.145%	−0.083%	−0.021%
POL	−0.457%	−0.384%	−0.311%	−0.838%	−0.704%	−0.570%
Total GHG change	0.094%	0.723%	1.357%	0.172%	1.325%	2.487%
GHG forecast	52.607	52.938	53.271	52.648	53.255	53.865
Medium-Term Trend Scenario
	2025			2030
	Min	Median	Max	Min	Median	Max
PGDP	0.511%	0.900%	1.288%	0.938%	1.650%	2.362%
PEC	0.075%	0.166%	0.261%	0.137%	0.305%	0.479%
REC	−0.127%	−0.073%	−0.018%	−0.233%	−0.133%	−0.033%
POL	−0.603%	−0.507%	−0.411%	−1.106%	−0.929%	−0.753%
Total GHG change	−0.144%	0.487%	1.121%	−0.264%	0.893%	2.055%
GHG forecast	52.483	52.814	53.147	52.420	53.027	53.638
Short-Term Trend Scenario
	2025			2030
	Min	Median	Max	Min	Median	Max
PGDP	0.391%	0.688%	0.984%	0.716%	1.260%	1.805%
PEC	0.041%	0.092%	0.145%	0.076%	0.169%	0.265%
REC	−0.296%	−0.169%	−0.042%	−0.543%	−0.310%	−0.078%
POL	−0.683%	−0.574%	−0.465%	−1.253%	−1.053%	−0.853%
Total GHG change	−0.547%	0.036%	0.621%	−1.004%	0.066%	1.139%
GHG forecast	52.271	52.577	52.885	52.031	52.593	53.157
Extreme Policy Scenario
	2025			2030
	Min	Median	Max	Min	Median	Max
PGDP	0.391%	0.688%	0.984%	0.716%	1.260%	1.805%
PEC	0.041%	0.092%	0.145%	0.076%	0.169%	0.265%
REC	−0.296%	−0.169%	−0.042%	−0.543%	−0.310%	−0.078%
POL	−4.760%	−4.000%	−3.240%	−21.420%	−18.000%	−14.580%
Total GHG change	−4.624%	−3.390%	−2.153%	−21.171%	−16.881%	−12.588%
GHG forecast	50.128	50.777	51.426	41.431	43.686	45.942

Source: estimated by the authors.

, where under extreme policy scenarios, it is necessary to add 100 and 450 hard law policies in 2025 and 2030, respectively. According to the descriptive statistics in Table 3, the average cumulative release of hard laws is 16, with a maximum of 235. While the achievement may still be modest, it is relatively easier to achieve compared to extreme scenarios involving all policies. Furthermore, if we hold the short-term trend of climate policy issuance constant and only enhance the emission reduction effectiveness of climate policies, the effectiveness of policies needs to be increased by 15 and 30 times by 2025 and 2030, respectively. The economic implication of a 15-fold increase in policy effectiveness is that every 100 policies can reduce greenhouse gas emissions by 12.68%, which is challenging to achieve in practice. However, if both policy efficiency and the quantity of policy issuance are increased simultaneously to around four times the short-term trend, there is hope that the warming targets set by the Paris Agreement will be achieved.

5. Discussion and Conclusions

In this paper, we selected representative policy indicators such as policy sectors, policy tools, policy objectives, legal enforceability and functionality, policy implementation persistence, and jurisdictional authority. We harmonized the IEA, CP, and CCLW authoritative datasets to obtain the Global Climate Change Policy Dataset (GCCMPD). Based on this versatile dataset, we constructed a global panel dataset from 1990 to 2019 and used trend models and fixed-effects models to predict emission reductions for the long-term, mid-term, and short-term trend scenarios up to the years of 2025 and 2030.

Through our research, we have found that there is a significant emission gap between the current climate policies and the goals of the Paris Agreement. Economic growth and primary energy consumption contribute to increased greenhouse gas emissions, while an increase in the proportion of renewable energy in electricity generation and the implementation of climate policies can help reduce greenhouse gas emissions. Furthermore, based on cluster analysis, we classified countries into four categories: developed countries, developing countries, countries with large emission reduction potential, and underdeveloped countries. We found that countries in different groups exhibit strong heterogeneity in emission reduction effectiveness, indicating significant potential for improvement in the effectiveness of climate policies. Classifying policies into hard law and soft law reveals that the majority of countries currently tend to adopt soft law policies, whereas hard law, due to its legal enforceability, typically achieves better emission reduction results. According to the calculation results, the long-term policy trends are similar to the unconditional NDC scenario, and the short-term trends are similar to the conditional NDC scenario, both of which are unable to meet the temperature targets of the Paris Agreement. In the short term, sustainable development-related factors have a significant impact on emission growth, and policy issuance can mitigate the emission increase caused by economic growth. However, medium- and long-term scenarios indicate that current policy trends may not fully mitigate the emission growth induced by economic expansion. It is worth noting that, through further analysis, the conclusion can be drawn that simultaneous increases in the adoption and effectiveness of climate policies are necessary to achieve the temperature targets of the Paris Agreement.

The main contributions of this paper are as follows: First, it enriches the research perspectives in the existing literature by using an easy-to-operate model that combines both the quantity and quality of policies to study the gap between current policies, energy, and economic trends and the goals of the Paris Agreement. Second, it emphasizes policy and national differences by objectively studying the effects of policies in different country groups using clustering methods and by investigating the impact of the legal attributes of policies on emission reduction through the distinction between soft law and hard law.

However, this study has some limitations, which can provide potential suggestions for future research. Firstly, the research scenarios in this paper are relatively simple. The advantage is that the scenarios derived from trend model results are objective, but they overlook sectoral differences. Secondly, while this paper emphasizes the legal attributes of policies, policies also have other attributes, such as policy instruments. Future research could further subdivide policies based on these attributes to derive more specific conclusions. Additionally, future studies could explore longer-term predictive research.

This paper focuses on climate policy and combines economic and sustainable development scenarios to quantify the gap between current climate policies and the goals of the Paris Agreement without compromising sustainable development. The conclusions drawn can provide insights for achieving temperature targets under sustainable development conditions. The conclusion of this study can be summarized as follows: to achieve the goals of the Paris Agreement, climate policies are insufficient both in “breadth” and “depth”. Therefore, countries need to adopt more climate policies to reduce emissions and curb temperature rise, and they may prioritize the use of hard law tools to enhance the implementation effectiveness of climate policies.