Sage Journals: Discover world-class research

Abstract

This paper provides insights into the interplay between international and domestic environmental governance by assessing the relationship between China's ratification of the Paris Climate Agreement and the intensity of its domestic environmental policies. Previous literature on the factors influencing environmental policy stringency in China has analyzed levels of bribery, regulatory failings, and information asymmetries among other domestic indicators, but has overlooked the international dimension of environmental governance. By applying a pseudo-panel data model analysis to a new dataset comprising 1721 Chinese environmental policies from 2000 to 2019, we conclude that China's ratification of the Paris Agreement has had a positive and significant impact on environmental policy intensity in China. Consequently, the analysis advances international environmental governance as a key explanatory factor for environmental policy intensity in China.

Keywords

environmental policy intensity Paris Climate Agreement COP21 pseudo-panel data model environmental policy stringency in China

Introduction

The People's Republic of China's (PRC) engagement in international environmental negotiations is fundamental to achieving global climate targets in many respects (Gao, 2016; Hilton and Kerr, 2017; Zhang, 2017). First, China is one of the major contributors to global warming. In 2022, the PRC's total carbon dioxide (CO₂) emissions reached 11,397 megatons, accounting for 31% of the world's share (International Energy Agency, 2024). In response to both domestic and international pressure to prioritize environmental objectives, the Communist Party of China announced that the country would aim to peak carbon emissions by 2030 and accomplish carbon neutrality by 2060 (Chen et al., 2021). Second, China plays a critical role in global low-carbon development due to the large-scale domestic production of low-emissions technologies, which enable the country to reduce manufacturing costs (Fischer, 2014: 77; Garnaut, 2014). The country's cost-leadership in technologies critical to the world's energy and mobility transitions (Lai et al., 2022; Lu et al., 2021) will play a critical role in disseminating affordable products to developing countries. Third, China's dominance in global supply chains paired with its self-identification as a developing country give it significant leverage to influence other countries’ positions in multilateral climate negotiations (Shen and Xie, 2018).

Of the platforms for multilateral climate negotiations, the annual Conferences of Parties (COPs) are where instruments are ironed out that hold signatories accountable to pledges made under Nationally Determined Contributions (NDCs). Of the 30 COPs staged thus far, the 2015 COP21 in Paris stands out, because the gathering resulted in 196 parties signing a legally binding treaty to “limit the temperature increase to 1.5 degrees Celsius above pre-industrial levels” (UNFCCC, 2015). From China's perspective, the Paris Climate Agreement represents a shift in the country's approach to international and domestic climate policy (Lian and Li, 2024). On the one hand, the Paris Agreement precipitated a switch to climate policy that combines a whole-of-system approach in China's NDCs with industry- and sector-specific policies (known as the “1 + N” strategy in China) (Wu, 2023), which aligns international pledges to a domestic economic growth strategy relying increasingly on consumption and services (Hilton and Kerr, 2017). On the other hand, China's positive stance in international climate policies is important in its quest for domestic and international legitimacy and recognition (Lian and Li, 2024).

The interplay between international and domestic politics, in particular the puzzle as to why states comply with international treaties at all, has a long history in the literature (Franck, 1988; Keohane, 2005). Both qualitative (Kingsbury, 1998; Raustiala and Slaughter, 2002; Simmons, 2010) and quantitative (Chayes and Chayes, 1995; Mitchell, 1994) work in this domain have been guided by the “theory of compliance”, where scholars analyze the conditions under which states abide by international treaties (Chayes and Chayes, 1995). In recognition of climate change threats, a rapidly growing body of literature focuses on the interactions between environmental degradation, trade policy, and environmental policy (Aichele and Felbermayr, 2012; Baccini and Urpelainen, 2014; Gallagher, 2004; Morin et al., 2018). Studies have also focused on how domestic factors such as trade openness, democracy, and the degree of integration into the world economy influence countries’ decisions to ratify international climate agreements (Bernauer et al., 2010; Neumayer, 2002). Less research has been conducted on evaluating the effect that a given international environmental treaty has on the severity of policy implementation domestically, a gap in the literature that this study seeks to fill.

Specifically, this paper investigates the dynamics between international and domestic environmental governance through the lens of environmental policy intensity. Treated synonymously in the literature, investigations on environmental policy intensity/stringency seek to evaluate the effectiveness of instruments selected to achieve outcomes stipulated in a policy (Knill et al., 2012; Zhang et al., 2022). As a consequence of research diversity, studies have come up with a variety of indicators to assess the intensity of policies directed at improving the environment, including qualitative yardsticks such as “scope”, “integration”, “implementation”, and “monitoring” (Knill et al., 2012), and quantitative standards such as “whether the instrument has measurable targets, designated budgets, clear objectives and timelines” (Zhang et al., 2022). The aim is to produce more quantifiable assessments of the links between policy design (instrument selection) and outcomes, so as to overcome endogeneity problems related to proxy variables, coding difficulties, and time constraints associated with manually dissecting policy documents (Zhang et al., 2022). This research contributes to these discussions by suggesting “international policy treaties”, in this case the Paris Agreement, as a contributing factor to environmental policy stringency in the PRC. Consequently, this paper proposes the following research question: How has China's ratification of the Paris Agreement affected the intensity of its domestic environmental policies?

To answer the research question, we use a new environmental policy intensity index developed by Zhang et al. (2022) to examine the connection between China's ratification of the Paris Agreement and the intensity of its domestic environmental policy. Zhang et al.'s data are particularly valuable to this research, because their study combines text analysis with machine-learning to produce one of the most comprehensive datasets on Chinese environmental policies to date. The policy intensity in the dataset is systematically calculated to measure the degree of stringency in the implementation of 1912 Chinese environmental policies from 1978 to 2019. Zhang et al.'s index is also beneficial because it allows us to circumvent endogeneity problems common in other studies that use different indicators to measure the stringency of environmental policies (Brunel and Levinson, 2013). As an example, studies that use pollution abatement expenditures as an indicator produce significantly different results than emissions-based indicators (Galeotti et al., 2020; Niemeijer, 2002). The discrepancy in results makes it difficult to assess the effectiveness of environmental policies (Galeotti et al., 2020). Furthermore, spending on pollution control does not necessarily translate into stricter environmental regulation, as countries with the highest pollution rates may spend more on pollution control. In a similar manner, reductions in emissions could be associated with changes in industrial composition or other macroeconomic trends (Brunel and Levinson, 2013). Adopting Zhang et al.'s estimation of Chinese environmental policy intensity allows us to avoid common pitfalls associated with using pollution abatement expenditures or reductions in emissions as indicators of stringency (Brunel and Levinson, 2013).

The dataset developed by Zhang et al. analyzes a new batch of environmental policies each year. In similar cases where historical information on the same unit of observation is not available, the dataset is classified as “repeated cross-sectional” (Lebo and Weber, 2015). While repeated cross-sectional data have advantages, such as providing a large number of observations, the analysis of temporal dynamics is a major limitation (Dutta et al., 2021). To overcome this limitation, we divide the dataset into three cohorts and calculate the mean values for each cohort according to the type of environmental policy instrument. We then use the mean values as the basic unit of observation. The transformation of repeated cross-sectional data into panel data according to the methodology of Deaton (1985) allows us to examine the linkages between China's ratification of the Paris Agreement and its environmental policy intensity. Our pseudo-panel regression model includes other control variables that previous studies have shown to have an impact on environmental policy stringency, such as corruption (Lisciandra and Migliardo, 2017), income inequality (Martínez-Zarzoso and Phillips, 2020), and economic growth (Gatti et al., 2023).

Our estimation shows that there is a positive and significant relationship between China's ratification of the Paris Agreement and its domestic environmental policy intensity. Hence, the study advances international environmental treaties as a determinant of environmental policy stringency in China. The focus on China is especially important, given the country's significance for achieving global climate change objectives. From a theoretical perspective, the study answers multiple calls for research to apply policy stringency indices to environmental policies (Dong et al., 2024; Kruse et al., 2022; Zhang et al., 2022).

The remainder of this paper is structured as follows. The following section outlines the relevant literature by first discussing the link between international and domestic environmental policy, before reviewing the empirical literature on environmental policy intensity. The next section introduces the dataset and methodology, and the section after that turns to the study's data and empirical strategy. The subsequent section then interprets the findings through the lens of theories on compliance and environmental policy intensity. The final section offers our recommendations for future research.

Theory

Compliance theory in international environmental law

Notwithstanding its salience today, the term “international environmental law” only gained traction after the 1972 United Nations Conference on the Human Environment in Stockholm (Gupta et al., 2022; Schachter, 1991). The conference was remarkable in that the resulting declaration held states responsible for “activities within their jurisdiction” that caused “damage to the environment or other states or areas” beyond national borders (United Nations, 1973: n.p.). Yet due to states’ opposition to binding legislation, as well as problems related to free-riding and policy coordination, multilateral treaties in the 1990s and 2000s adopted broad language to define environmental damage, obligations, and global targets. Principles of international environmental law include states’ obligations to prevent, reduce, rectify, or make amends for harm, to notify and consult with others over risks, and to seek peaceful resolution in disagreements. However, for these principles to be recognized as customary law “evidence of general state practice and opinion juris” (Schachter, 1991: 462) would be required. Given the absence of an enforcement body and the non-binding nature of international environmental law, for decades scholarship has questioned whether international law can be considered legal at all (Franck, 1988; Von Stein, 2013).

Despite international relations scholars having since argued that multilateral cooperation without binding laws is possible (e.g. Axelord, 1984; Henkin, 1979) legal scholars continue to question why states abide by international commitments. In most recent times, scholarship on this question has centered on cross-disciplinary investigations that integrate rational choice and behavioral economics with the legal compliance literature (Keohane, 2005; Peat et al., 2022; Von Stein, 2013). Keohane (2005) divides this set of literature into scholars who argue that it is the material costs and benefits of compliance that motivate states, and those who believe states act on normative judgments connected to their identity and legitimacy. Von Stein (2013) integrates the former realist/instrumentalist mindset with the latter constructivist viewpoint by contending that mechanisms for determining norms and costs are not mutually exclusive and that cost-benefit calculations and normative judgments act upon one another in a dynamic fashion. In this sense, reputation (Henkin, 1979), reciprocity, and credibility play a central role in states’ compliance behavior (Schachter, 1991), since committing to environmental promises today pays dividends in future negotiations (Brewster, 2009). Peat et al. (2022: 174) introduce “behavioral insights” as a “third pillar of compliance” to stand alongside material and normative deliberations in international environmental law. By adding emotions as a function of compliance behavior, these authors join others (e.g. Brader and Marcus, 2013) in efforts to incorporate the psychological processes of individuals (leaders) as determinants of whether a state attends to its international legal obligations (or not).

As to the instruments to force (“hard”) or cajole (“soft”) compliance, the literature distinguishes between positive (rewarding) (van Aaken and Simsek, 2021) and negative (punitive) (Hathaway and Shapiro, 2011) actions or a linkage between the two (Leebron, 2002). Zehavi (2012) challenges this dichotomy by distinguishing between “hard” and “soft” instruments (“carrots and sticks”) and “sermons” that countries use to entice compliance. In any case, action is taken because environmental improvements in a foreign jurisdiction are deemed “important enough to a (group of) state(s) that it is willing to pay the cost of inducing others to comply” (Von Stein, 2013: 479). Penalties¹ and rewards can either be internal or external to a treaty and applied upon entry or as a retribution/benefit for (non-)compliance (van Aaken and Simsek, 2021). In establishing punishments and rewards, it is important for exacting parties to establish the targets’ baseline expectations, value position, and desired reputational image (Keohane, 2005). Rewards can hence be thought of as improvements to the baseline, whereas punishments are reductions to the same baseline (van Aaken and Simsek, 2021). For environmental treaty design, difficulties may arise in first identifying and then isolating a reward or punishment to the deserving/violating entity (Barrett, 2007). Concretely, rewards may take the form of aid, money, or technology transfers, reductions in barriers to trade, or other benefits of cooperation. Penalties, meanwhile, include outcasting, sanctioning, cuts in development assistance, reputational damage, and (in rare cases) military intervention (Hathaway and Shapiro, 2011).

Yet while much theorizing has been undertaken to link strategies for inducing compliance with instrument combinations (e.g. Hathaway and Shapiro, 2011; Leebron, 2002; van Aaken and Simsek, 2021), less attention has been paid to developing frameworks that allow for quantitative assessments of treaty implementation (Zhang et al., 2022). Treaty implementation is defined as the “adoption of legal measures by states to translate their international commitments in their domestic legal order” (Weiss and Jacobson, 1998: 4). Treaty effectiveness is then placed along a continuum from the most immediate (translation into domestic law) to more distant standards of policy implementation (e.g. environmental outcomes) (Young, 1994). The benefits of investigations focusing on more immediate measures of effectiveness include the avoidance of endogeneity problems and the ability to disaggregate environmental issues. Analyzing policies rather than environmental outcomes also solves the problem of deciding how much time to compute for a treaty to (reasonably) take effect. A drawback is that the unit of analysis (documents) merely reflects the conditions for, not the outcomes of, treaty implementation (Brandi et al., 2019). Similarly, isolating the effectiveness of individual instruments embedded in environmental policy design remains a problem. To overcome these difficulties, the Organisation for Economic Co-operation and Development (OECD) (2016) has called for research that defines indicators, such as policy intensity, to quantify instruments’ effectiveness for achieving environmental improvements.

Environmental policy intensity

Environmental policy intensity is a cost on environmentally harmful behavior that arises due to the imposition of a policy instrument (Botta and Koźluk, 2014). To measure costs, environmental policy intensity indices reflect “the stringency and importance of policy” (Dong et al., 2024), with weights placed on policies’ objectives and instruments (Zhang et al., 2022). In evaluating the determinants of stringency, empirical literature analyzes several factors, including corruption (e.g. Lisciandra and Migliardo, 2017), the urbanization rate, perceived environmental damage, state capacity, and political ideology (Cadoret and Padovano, 2024), corporate lobbying (Pommeret et al., 2022), as well as public pressure, economic inequality, and growth (e.g. Gatti et al., 2023). With respect to international treaties, the “interaction dividend”, or how a country benefits from spill-over effects from interaction with other signatories’ environmental legislation, is a significant determinant of environmental policy stringency. Cadoret and Padovano (2024: 2) estimate that the dividend “accounts for between one third to more than one half of the measured degree of stringency”. Determinants differ, among other factors, depending on the type of environmental policy, level of government, and instruments selected to achieve objectives (Corrocher and Mancusi, 2021). This multi-dimensionality in measuring environmental policy stringency is a significant challenge authors face in creating indices (Botta and Koźluk, 2014).

As a consequence of the many, varied, and dynamic determinants that need to be considered in quantifying environmental policy stringency, the heterogeneity in objective-instrument patterns, and whether indices measure a single or a collection of environmental legislation (Botta and Koźluk, 2014), several indices to measure environmental policy intensity have emerged (e.g. Huang et al., 2018; Schaffrin et al., 2015; Zhang et al., 2022). Following the OECD's (Botta and Koźluk, 2014) development of an environmental policy stringency index that evaluates 13 instruments across 40 countries, others contributed their own composite indices with unique individual indicators (Dong et al., 2024; Gupta et al., 2022; Kruse et al., 2022; Zhang et al., 2022). What these indices have in common is a separation of policy tools into market-based and non-market-based instruments or a combination of the two.² Market-based instruments include carbon and renewable energy trading schemes, taxes on fuel and greenhouse gases, and financial stimuli in the form of loans, grants, subsidies (Zhang et al., 2022), or price supports (Johnstone et al., 2010). Non-market-based instruments tend to have a command-and-control function (Dong et al., 2024), which is why they take the form of penalties (Turken et al., 2017), standard-setting, licensing, and restrictions on use (such as prohibitions or bans) (Tang et al., 2020), but also permits that determine access to public resources (Knill et al., 2012).

While some studies on environmental stringency in China make use of the OECD's original (Botta and Koźluk, 2014) and updated (Kruse et al., 2022) indices (e.g. Ahmed and Ahmed, 2018; Wang et al., 2020), Zhang et al. (2022) and Dong et al. (2024) develop indices specific to the Chinese case. Similarities can be found in Zhang et al. and Dong et al. both including carbon reduction, energy conservation, capacity utilization, and technology as objectives of environmental policy in China. Both China-specific indices also use broad terms to describe instruments analyzed for intensity. Additionally, overlap exists in the inclusion of fiscal and tax measures, financial measures (investment incentives, loans, subsidies, grants, and price supports) and tradeable certificates (Dong et al., 2024; Zhang et al., 2022). The most striking differences are that Dong et al. split their list of 14 instruments into two categories (market-based and command-and-control), whereas Zhang et al. add public participation as a further category into which they divide seven measures.³

For examining the connection between China's ratification of the Paris Agreement and the intensity of its environmental policies, Zhang et al.'s index is more useful than Dong et al.'s approach for the following reasons. First, Zhang et al.'s definition of policy intensity, with an emphasis on the delivery of a policy from the perspective of designers, is better suited than Dong et al.'s framing of stringency from a policy-taker's viewpoint. Second, this paper explicitly answers Zhang et al.'s call for future comparative research on changes in Chinese environmental policies over time, whereas Dong et al.'s index is better suited to cross-jurisdictional comparisons within the PRC. Third, though Dong et al.'s policy inventory is built on a greater number of documents (7,282), Zhang et al. limit their still-significant lexicon of 1,912 documents to environmental policies by applying more selective criteria to keyword searches. Finally, while both indices cover years that fall within a suitable timeframe around the Paris Agreement, Zhang et al. draw policies from a larger set of databases that also collect sub-national legislation, thereby covering the most obvious benefit of using Dong et al.'s index. While Zhang et al.'s scalar measurement of policy intensity has furthered our understanding of the development of environmental policy intensity in China over time, relying solely on their Environmental Policy Intensity index (EPI index) to evaluate the environmental impacts of policies is inadequate, as policy enforcement remains the primary determinant of environmental outcomes. However, studying policy enforcement would require a far more expansive methodology that goes beyond the scope of this paper.

The existing literature on the determinants of environmental policy stringency in China focuses, among other variables, on levels of bribery, regulatory failings (Peat et al., 2022; Pommeret et al., 2022), information asymmetry between policy-makers and stakeholders (Chen et al., 2022), and regional innovative potential (Chen and Tanchangya, 2022). Though valuable for uncovering the determinants of environmental policy stringency within China, scholarship has so far neglected the international dimension of the Chinese state pledging domestic improvements by signing multilateral climate treaties. The “interaction dividend”, identified as a significant contributor to climate change mitigation elsewhere (Cadoret and Padovano, 2024), has, to the best of our knowledge, not been assessed as a determinant for environmental policy stringency in China. Based on Zhang et al.'s (2022) measurement of environmental policy intensity, this study assesses whether an “interaction dividend” in the form of changes in stringency can be associated with the PRC signing the Paris Agreement in 2016.

Data and empirical strategy

Data description

To understand the relationship between China's ratification of the Paris Agreement and the intensity of its domestic environmental policies, we analyze 1721 Chinese environmental policies from 2000 to 2019. We believe that extending the timeframe of the analysis to before 2000 is not in line with our research question, which focuses on the recent development of environmental policy intensity in China. For robustness measures, we shorten the timeframe to 2005–2019 to compare the findings presented in the Results section with those from an alternative period. Data on the EPI index from Zhang et al.'s (2022) database are not available after 2019; consequently, the post-ratification timeframe is from 2016 to 2019. We believe that this timeframe is sufficient to capture the correlation between China's ratification of the Paris Agreement and environmental policy intensity in China. This relatively short timeframe is consistent with the findings of Baccini and Urpelainen (2014), who demonstrate that the impacts of international treaties can manifest as early as the ratification period.

Data sources

Data on the intensity of environmental policy and on environmental policy instruments during this timeframe are retrieved from Zhang et al. (2022). The dataset classifies environmental policies into three categories, according to the type of environmental policy instrument: (1) Command-and-Control Environmental Policies (CCEPs), (2) Market-based Environmental Policies (MBEPs) and (3) Public Participation Environmental Policies (PPEPs). The main differences between these instruments are that CCEPs rely on direct regulatory measures and mandates, while MBEPs use financial and incentive-based instruments (Goulder and Parry, 2008). Conversely, PPEP instruments adopt a decentralized approach that seeks the active participation of various stakeholders, such as from public consultation sessions and community programs (Santos et al., 2006). Table 1 shows the frequency distribution of each environmental policy instrument in the dataset.

Table 1.

Frequency distribution of environmental policy instruments and cohort sizes, 2000–2019.

Year	Environmental policy instrument			Total
	CCEPn = 1221	MBEPn = 278	PPEPn = 222	N = 1721
2000	17 (85%)	2 (10%)	1 (5%)	20
2001	25 (96%)	0 (0%)	1 (4%)	26
2002	19 (86%)	0 (0%)	3 (14%)	22
2003	14 (78%)	3 (17%)	1 (6%)	18
2004	25(89%)	2 (7%)	1 (4%)	28
2005	34(83%)	3 (7%)	4 (10%)	41
2006	55 (90%)	5 (8%)	1 (2%)	61
2007	74 (73%)	18 (18%)	10 (10%)	102
2008	52 (81%)	6 (9%)	6 (9%)	64
2009	54 (72%)	15 (20%)	6 (8%)	75
2010	119 (76%)	25 (16%)	13 (8%)	157
2011	106 (72%)	16 (11%)	25 (17%)	147
2012	75 (60%)	20 (16%)	29 (23%)	124
2013	78 (67%)	21 (18%)	17 (15%)	116
2014	99 (77%)	9 (7%)	21 (16%)	129
2015	69 (51%)	32 (24%)	33 (25%)	134
2016	159 (85%)	14 (7%)	14 (7%)	187
2017	51 (49%)	32 (30%)	22 (21%)	105
2018	45 (57%)	28 (35%)	6 (8%)	79
2019	51 (59%)	27 (31%)	8 (9%)	86

Data source: Zhang et al. (2022).

The frequency distribution within the dataset shows that CCEP instruments are still the dominant type of environmental governance mechanism in China, but the shares of MBEP and PPEP instruments are rapidly growing. This is due to the importance of engaging various actors in carbon emissions reduction and because CCEP instruments, despite their high enforceability, are designed to expire after a certain period of time (Jiang et al., 2023). MBEP instruments are more cost-effective and provide greater flexibility in meeting environmental targets (Goulder and Parry, 2008). PPEP instruments are the least enforceable by design, since they intend to strengthen democracy and build legitimacy among stakeholders (Newig, 2007; Santos et al., 2006).

Figure 1 illustrates the relationship between Zhang et al.'s (2022) EPI index and different environmental policy instruments. The visualization of mean stringency values from 2000 to 2020 shows an upward trajectory for environmental policy intensity across the three environmental policy instruments (Zhang et al., 2022). In visually comparing environmental policy stringency pre- and post- China's ratification of the Paris Agreement, the data show that during the timeframe 2016 to 2019, the mean values of stringency increased significantly for CCEPs and MBEPs. Environmental stringency for PPEPs, meanwhile, does not show significant variation from 2016 to 2019. In fact, from the highest level upon the treaty's inception, slight decreases in intensity for measures designed to incorporate a broader set of stakeholders can be observed for the years after 2015.

Figure 1.

EPI index by environmental policy type, 2000–2019.

Variables

The dependent variable in our regression model is the EPI index developed by Zhang et al. (2022). As outlined above, the choice of this variable is in line with theoretical considerations (see Theory section of this paper) that suggest a link between treaty implementation and environmental policy in other countries (e.g. Cadoret and Padovano, 2024). The EPI index ranges from 13 to 506 in the timeframe of the dataset, where higher values mean a greater level of policy intensity. The main explanatory variable relevant to our study is a binary variable, which takes the value 1 after China's ratification of the Paris Agreement and 0 otherwise. In the dataset, this corresponds to 0 from 2000 to 2015 and 1 from 2016 to 2019. We add three extra control variables (corruption, income inequality, and economic growth) to improve the explanatory power of the regression model and to reduce the error variance (Gormley and Matsa, 2014). These control variables are important to capture the effects that domestic-level factors in China could have on the intensity of its environmental policy. Adding corruption, income inequality, and economic growth as control variables also allows us to compare the Paris Agreement as a determinant of environmental policy stringency in the PRC to other, more established contributing factors.

For the selection of control variables, we draw on previous literature that analyzes corruption, income inequality (Gatti et al., 2023; Martínez-Zarzoso and Phillips, 2020), and gross domestic product (GDP) per capita (Gatti et al., 2023; Lisciandra and Migliardo, 2017) as predictors of environmental policy stringency in China. First, to indicate corruption levels in China, we use the Transparency International (2024) Corruption Perceptions Index (CPI), which ranks the perception of corruption on a scale from 0 to 100, where 0 indicates the highest and 100 the lowest level of perceived corruption (Budsaratragoon and Jitmaneeroj, 2020). Second, we draw on the Gini coefficient to assess the effect of income inequality on the EPI index. The variable is defined as the “Nationwide Gini Coefficient of Per Capita Disposable Income” and obtained from the “People's Living Conditions” online database by the National Bureau of Statistics of China (2024a). Finally, to examine the linkage between economic growth and environmental policy intensity, we use per capita GDP, measured in Chinese yuan (National Bureau of Statistics of China, 2024b).

Empirical strategy and model specification

Pseudo-panel data model

Zhang et al.'s (2022) archive is analogous to population survey data, where different cross-sections are observed at consecutive points in time (Verbeek, 2008). A major limitation of this type of data (i.e. repeated cross-sectional data), is the difficulty in inferring temporal dynamics in comparison to real panel data, where information about the same individuals is available across different points in time (Dutta et al., 2021). To overcome this limitation, we follow Deaton's (1985) methodology for constructing pseudo-panel data. The unit of observation in pseudo-panels is the intra-cohort mean in comparison to individuals or entities in real panel data (Deaton, 1985). The cohorts must share common characteristics that do not change over time, such as birth year (Guillerm, 2017) or households who follow certain criteria (Bernard et al., 2011). We group data on 1721 Chinese policies during the time frame 2000–2019 into three cohorts according to the classification of their environmental policy type by Zhang et al. (2022), namely, CCEPs (Cohort 1), MBEPs (Cohort 2), and PPEPs (Cohort 3).

A key limitation of the pseudo-panel methodology occurs when the mean values of cohorts are not representative of population-level means (Khan, 2018). Verbeek and Nijman (1992) suggest that the cohort population should be at least 100 to minimize the sampling error. Table 1 shows that the cohorts in this study exceed this quantitative barrier, with cohort sizes of 1221, 278, and 222 for Cohort 1, Cohort 2, and Cohort 3, respectively. The transformation of a total of 1721 individual observations from 2000 to 2019 into cohorts collapses the dataset into 58 synthetic observations. We observe three cohorts over 20 years with 1-year intervals; calculated as 20 + 18 + 20 = 58 observations for Cohort 1, Cohort 2, and Cohort 3, respectively.⁴

Model specification

To construct pseudo-panels, we aggregate all observations to the cohort levels. Following Deaton's (1985) approach, we obtain the following model:

{\bar{y}}_{c t} = {\bar{x}}_{c t} + {\bar{θ}}_{c t} + {\bar{ε}}_{c t}, c = 1, \dots, C; t = 1, \dots, T .

(1)

In this equation, ${\bar{y}}_{c t}$ represents the averages of the EPI index for Cohort 1, Cohort 2, and Cohort 3 from 2000 to 2019, and ${\bar{x}}_{c t}$ are the averages of the explanatory variables explained above for cohort c at time t. Since the regression coefficients vector, ${\bar{θ}}_{c t}$ , may be correlated with the error term, ${\bar{ε}}_{c t}$ , treating ${\bar{θ}}_{c t}$ as random might lead to inconsistent estimators (Verbeek and Nijman, 1992). In our model, we assume that ${\bar{θ}}_{c t}$ and ${\bar{ε}}_{c t}$ are fixed, since there are time-invariant unobserved effects, such as political variables, that could affect the predictors used in our model (corruption, income inequality, and economic growth). To decide whether to use Fixed Effects (FE) or Random Effects (RE), we apply the Hausman test. The Hausman test assumes that the individual effects, ${\bar{θ}}_{c t}$ , and the error term, ${\bar{ε}}_{c t}$ , are uncorrelated. Therefore, the null hypothesis of the Hausman test is that there is no difference between FE and RE. A statistically significant p-value rejects the null hypothesis of the Hausman test and confirms that the FE estimator is more appropriate than the RE estimator (Wooldridge, 2010: 288–289). Our Hausman test results show a p-value of 1.000; accordingly, we expect that the results from FE and RE to be similar. We estimate the two models to compare results in the next section.

Results

Population-level analysis

Before conducting the cohort-level regression analysis, we first analyze Zhang et al.'s (2022) dataset at the population level, which includes 1721 environmental policies in China from 2000 to 2019. A two-sample t-test is conducted to compare the mean values of environmental policy intensity in China before and after China's ratification of the Paris Agreement. The two-sample t-test assumes that the means of the two populations are the same (Kim, 2015). The results in Table 2 show that the mean value of environmental policy intensity significantly increased after China's ratification of the Paris Agreement. The p-value shown in Table 2 allows us to reject the null hypothesis of equal means with statistical significance; however, we cannot discern that these differences are correlated with China's ratification of the agreement or other socio-economic determinants. We address this concern in the following section, by controlling for major socio-economic factors in the cohort-level regression analysis.

Table 2.

Two-sample t-test of environmental policy before and after the Paris Agreement entered into force, 2000–2019.

EPI index	Number of observations	Mean value	Standard error
Environmental policies before the ratification of the Paris Climate Agreement	1264	53.346	37.521
Environmental policies after the ratification of the Paris Climate Agreement	457	64.889	56.073
P-value = 0.000

Data source: Zhang et al. (2022).

Pseudo-panel regression analysis (cohort-level)

Using panel data estimation techniques, we apply equation (1) to obtain the pseudo-panel regression results shown in Table 3. The dependent variable is the EPI index in China and the explanatory variables are a dummy variable indicating China's ratification of the Paris Agreement, the net Gini coefficient, log GDP per capita, and the CPI. The panel dataset is unbalanced due to missing data for the net Gini coefficient in China for the years 2000, 2001, and 2002. Table 3 shows the analysis during the timeframe 2000–2019. For robustness checks, we estimate different timeframes in the next section. We estimate FE regression in model (1) and address potential endogeneity between economic growth and environmental policy intensity by including the lagged variable “log GDP per capita” in model (2). Model (3) presents the results of the RE regression model.⁵

Table 3.

Pseudo-panel data regression results, 2000–2019.

Variables	FE	FE	RE
Variables	(1)	(2)	(3)
Paris Agreement (dummy variable)	7.849**(3.558)	7.902**(3.601)	7.849**(3.558)
Net Gini coefficient of per capita disposable income	−207.086(164.302)	−192.581(170.544)	−207.086(164.302)
Log GDP per capita (Chinese yuan)	5.426(3.725)		5.426(3.725)
Corruption Perceptions Index (CPI)	−1.093(0.832)	−1.120(0.859)	−1.093(0.832)
Log GDP per capita (lagged)		5.504(3.896)
Constant	131.491(95.433)	125.496(99.609)	131.491(95.545)
Observations	51	50	51
R-squared	0.345	0.343	0.211
Number of cohorts	3	3	3
Fixed effects	YES	YES	NO

Note: Standard errors in parentheses.

** p < 0.05.

The results of the three estimated models show that China's ratification of the Paris Agreement had a positive and significant correlation with environmental policy intensity at the 5% significance level. The regression coefficient of the dummy variable indicates that, all other things being equal, the average difference in the EPI index after China's ratification of the Paris Agreement relative to the EPI index before the ratification is approximately 8 points. The outcome of the regression suggests that the Paris Agreement holds more explanatory value for policy stringency compared to the control variables corruption, income inequality, and economic growth.

The results of the net Gini coefficient show a negative correlation with environmental policy intensity with no statistical significance. The negative coefficient sign implies that higher levels of income inequality could explain the lower levels of environmental policy intensity, but our analysis cannot confirm this result with statistical significance.

The results also demonstrate that log GDP per capita has a positive relationship with environmental policy intensity in China (see Figure 2), although the coefficient is not statistically significant. We assume that a larger sample size is needed to confirm this result with statistical significance. To exclude potential endogeneity between log GDP per capita and the EPI index, we replaced the log GDP per capita with its lagged values in model (2): this measure did not have a significant impact on the results.

Figure 2.

Correlation between EPI index and GDP per capita, 2000–2019.

The CPI coefficient unexpectedly yields a negative sign. The CPI ranks perception of corruption within a country from 0 to 100, where 0 indicates the highest and 100 the lowest level of perceived corruption. Therefore, the intuitive expectation is that higher values of the CPI should be positively correlated with the EPI index. To exclude that the CPI has a multicollinear relationship with the net Gini coefficient, we dropped the net Gini coefficient from the model to see whether the CPI still retained its negative sign despite this measure. Even after dropping the Gini coefficient, the variable CPI remained negative. The unexpected coefficient sign could be explained by the small variation of the values within the sample: from 31 as a minimum to 41 as a maximum value. The number of observations does not seem large enough to produce an efficient estimation.

Model (1) and model (3) validate the results of the Hausman test and show no difference between the FE and RE estimators. The similarity between FE and RE estimates could be explained by the low inter-cohort variability that resulted from pooling the data into cohorts (Verbeek and Nijman, 1992).

Robustness checks

In empirical studies, it is important to test the rigor of the regression estimates by including additional variables or testing the regression model over different time periods. Due to the focus of our study on the intensity of environmental policy in China and its linkage to the Paris Agreement, we test the robustness of the regression model under a different timeframe. Table 4 reports the regression analysis for the period 2005 to 2019. The results from Table 4 are statistically no different from outcomes in Table 3 and therefore confirm our findings.

Table 4.

Pseudo-panel data regression, 2005–2019.

Variables	FE (1)	RE (2)
Paris Agreement (dummy variable)	7.927**(3.787)	7.927**(3.787)
Net Gini coefficient of per capita disposable income	−329.975(273.891)	−329.975(273.891)
Log GDP per capita (Chinese yuan)	1.405(8.054)	1.405(8.054)
Corruption Perceptions Index (CPI)	−0.822(1.010)	−0.822(1.010)
Constant	222.178(186.105)	222.178(186.168)
Observations	45	45
R-squared	0.321	0.194
Number of cohorts	3	3
Fixed effects	YES	NO

Note: Standard errors in parentheses.

** p < 0.05.

Discussion

Enforcement of environmental regulation is necessary for achieving global climate change mitigation objectives and for assessing the effectiveness of international environmental governance (Ascher, 2001; Baccini and Urpelainen, 2014). In empirical studies, indicators of environmental policy stringency are used to assess the effectiveness and outcomes of environmental regulation. In line with others’ contributions (e.g. Zehavi, 2012), our results show that China has deployed a mix of instruments (CCEPs and MBEPs), as well as sermons (PPEPs), to enforce domestic environmental legislation since ratifying the Paris Agreement. The results therefore challenge the hard–soft dichotomy that is traditional in the literature on environmental policy compliance (Keohane, 2005). Though still expressing a preference for CCEPs, our data suggest movements toward other, less intrusive tools that incorporate a greater population base. The results show a positive and significant relationship between China's ratification of the Paris Agreement and its domestic environmental policy intensity, which reflects the novel, system-wide (“1 + N”) approach adopted in China's NDCs. Moreover, our results indicate a positive but statistically nonsignificant relationship between log GDP per capita and environmental policy intensity, as illustrated in Figure 2. The positive coefficient sign is consistent with the previous literature on the relationship between economic growth and environmental stringency, for instance on the Environmental Kuznets Curve (EKC) (Kuznets, 2019). The EKC hypothesizes that environmental degradation initially increases alongside economic growth in the early stages of development. However, as countries reach a certain level of income, the relationship between economic growth and environmental degradation becomes negative, as increased investments in technology, innovation, and stricter environmental regulations lead to improved environmental outcomes (Leal and Marques, 2022). Our results are also consistent with Dang and Serajuddin's (2020) findings that the levels of micrograms per cubic meter of PM2.5 (particulate matter) in China started to decouple from GDP per capita in 2016 (Dang and Serajuddin, 2020). A growing body of literature, however, questions the adequacy of the EKC in explaining the complexity of contemporary environmental challenges (Leal and Marques, 2022). Economic growth and technological progress, for instance, do not fully address critical issues such as resource allocation and environmental injustice (Biely and Chakori, 2025).

Since one of the most significant challenges for the identification of robust indicators for environmental policy stringency is endogeneity (Zhang et al., 2022), we use a new measure for environmental policy intensity that estimates the degree of intensity in delivering and enforcing environmental regulations (Zhang, 2017). Applying the pseudo-panel methodology, we use the EPI index to evaluate “international environmental treaties” as a further element of environmental policy intensity. The regression also includes corruption, income inequality, and GDP growth as control variables, which the previous literature has identified as key drivers of environmental policy intensity in China. By integrating the international dimension in addition to other key factors recognized as critical in Chinese climate policy-making, the analysis generates important insights into the interaction between multilateral and domestic environmental policy. While we expected the coefficient of the CPI to be positive, the regression analysis produced a negative sign. This unexpected result supports Budsaratragoon and Jitmaneeroj's (2020) view that continuous change in the calculation methodology of the CPI disturbs the consistency of the year-to-year assessment. Therefore, the CPI is more suitable for cross-country analyses. Net Gini coefficient results, meanwhile, show an expected negative sign. Although statistically nonsignificant, the negative coefficient indicates an inverse relationship between income inequality and environmental policy intensity. This result is consistent with Gatti et al. (2023), who in their analysis of the determinants of environmental policy stringency in OECD countries also find that the net Gini coefficient has a negative and statistically nonsignificant coefficient at the 5% significance level.

Conclusion

Going back to our research question—how does China's ratification of the Paris Agreement affect the intensity of its domestic environmental policy?—this study's most impactful conclusion is that the Paris Agreement has had a positive and significant effect on the Chinese state's design of domestic climate legislation. Table 2's comparison of mean values of environmental policy intensity pre- and post- China's ratification of the Paris Agreement shows a significantly higher intensity level after the treaty's signing. The pseudo-panel regression estimates presented in Table 3 also reveal that the Paris Agreement is a strong explanatory variable for environmental policy stringency compared to more established control variables, such as corruption, income inequality, and economic growth. Temporal analysis of instruments deployed in Chinese climate legislation also indicates that the Chinese state has intensified its use of tools associated with CCEPs, while moderating the employment of measures that incentivize a broader set of stakeholders to participate in policy design and implementation.

The results contribute to discussions in the literature on compliance theory, by showing that the Paris Agreement, as a multilateral climate treaty, has been an important factor in increasing the stringency of Chinese environmental policies. In a time when the effectiveness of COPs as a vehicle of environmental improvement has come under scrutiny, our study of policy stringency in China underlines the continued significance of multilateral climate treaties in tackling climate change. The study also shows that, contrary to condemnation from politicians and media personalities, China has adopted (primarily command-and-control) measures to demonstrate compliance to its multilateral commitments. Based on the Chinese case, the findings also offer lessons for treaty design. Even though the Paris Agreement contains no enforcement mechanisms to hold countries accountable, we show that China, one of the world's most consequential actors in the struggle against climate change, has still translated treaty objectives into domestic policy. The reasons why China has done so, whether for genuine concern for the environment, to placate civilian demands, for economic benefits, reputational gains, or entirely other reasons, require further study. Future studies should also concentrate on how China's political context, characterized by fragmented authoritarianism under a centralized hierarchy, builds institutions that contribute to (or hinder) the translation of multilateral goals into domestic policy.

Conceptually, by investigating domestic policy intensity in one of the world's most important contributors to climate change mitigation, this study's focus on immediate measures also adds to the scholarship on treaty implementation and effectiveness. In drawing on policy intensity as a dependent variable, the inquiry answers multiple calls for practical applications of one of the largest databases on Chinese environmental policies. The investigation's focus on the Paris Agreement also opens up novel avenues of inquiry to uncover contributing factors to policy intensity beyond a country's borders. In doing so, the study contributes to discussions on the “interaction dividend”, as well as discourses on why multilateral treaties remain an important mechanism to induce compliance.

The research suffers from multiple limitations that raise possibilities for future avenues of inquiry. First, policy intensity as a measurement device is limited in its ability to certify implementation of a captured instrument. To back up the quantitative findings of tools stipulated in policy design, qualitative studies involving field research can shed light on the effectiveness and modes of implementation on the ground. Qualitative content analysis could also shed light on nuances within environmental policies in order to identify subcategories of tools and mitigate the literature's broad division of legislation into command-and-control and market- and non-market-based indices. Second, temporal analysis of policy intensity is obstructed by the multi-dimensionality in variables that contribute to the design of an index. Additional studies on determinants of policy intensity are consequently required to build more robust indices that in turn can be used for practical purposes, including comparisons of stringency over time and space. Third, our study frames the Paris Agreement as a single case where a multilateral treaty causes a change in a signatory's domestic environmental policy stringency. Given our results, we call on others to replicate or innovate methods to examine other multilateral treaties’ effects on policy intensity in China and other countries. Fourth, for the Chinese case, recognition of differences in policy intensity across government levels and industries generate possibilities for local and regional indices, which would facilitate the identification of model jurisdictions and environmental paragons among local governments. Finally, pseudo-panel techniques have some limitations. For instance, relying on aggregate data causes a loss of information and aggregation bias. The use of pseudo-panel models to assess environmental policy stringency has yet to become an established method for analyzing policy intensity. Given the method's effectiveness in producing this study's results, we call on others to verify, replicate, and optimize the methodology in other areas related to environmental policy stringency and treaty implementation.

Footnotes

Acknowledgments

We acknowledge the institutional support provided by the Chair of China Business and Economics at the University of Würzburg. Additionally, we are grateful to Prof. Doris Fischer for her professional guidance, and to the Center on Contemporary China at Princeton University for its assistance in enabling us to conduct this research.

Contributorship

Passant Aboubakr conceived the study and was responsible for the research design and data analysis. Hannes Gohli contributed to the literature review, theoretical framework, and discussion. Both authors jointly wrote the manuscript and were involved in drafting and revising the original version. Both authors reviewed and approved the final manuscript prior to submission.

ORCID iDs

Passant Aboubakr

Hannes Gohli

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Notes

References

Ahmed

(2018) A predictive analysis of CO₂ emissions, environmental policy stringency, and economic growth in China. Environmental Science and Pollution Research 25(16): 16091–16100.

Aichele

Felbermayr

(2012) Kyoto and the carbon footprint of nations. Journal of Environmental Economics and Management 63(3): 336–354.

Ascher

(2001) Coping with complexity and organizational interests in natural resource management. Ecosystems 4: 742–757.

Axelord

(1984) The Evolution of Cooperation. New York: Basic Books.

Baccini

Urpelainen

(2014) Before ratification: Understanding the timing of international treaty effects on domestic policies. International Studies Quarterly 58(1): 29–43.

Barrett

(2007) Why Cooperate? The Incentive to Supply Global Public Goods. New York: Oxford University Press.

Bernard

J-T

Bolduc

Yameogo

N-D

(2011) A pseudo-panel data model of household electricity demand. Resource and Energy Economics 33(1): 315–325.

Bernauer

Kalbhenn

Koubi

, et al. (2010) A comparison of international and domestic sources of global governance dynamics. British Journal of Political Science 40(3): 509–538.

Biely

Chakori

(2025) Sustainability transition theories: Perpetuating or breaking with the status quo. Sustainable Development 33(1): 52–68.

10.

Botta

Koźluk

(2014) Measuring environmental policy stringency in OECD countries: A composite index approach. OECD Economics Department Working Paper No. 1177. Available at: https://www.oecd-ilibrary.org/economics/measuring-environmental-policy-stringency-in-oecd-countries_5jxrjnc45gvg-en (accessed 25 September 2024).

11.

Brader

Marcus

(2013) Emotion and political psychology. In: Huddy

(eds) The Oxford Handbook of Political Psychology. Oxford: Oxford University Press, pp. 165–204.

12.

Brandi

Blümer

Morin

J-F

(2019) When do international treaties matter for domestic environmental legislation? Global Environmental Politics 19(4): 14–44.

13.

Brewster

(2009) Unpacking the state’s reputation. Harvard International Law Journal 50: 231–269.

14.

Brunel

Levinson

(2013) Measuring environmental regulatory stringency. OECD Trade and Environment Working Papers, August 22. Available at: https://doi.org/10.1787/5k41t69f6f6d-en (accessed 10 September 2024).

15.

Budsaratragoon

Jitmaneeroj

(2020) A critique on the corruption perceptions Index: An interdisciplinary approach. Socio-Economic Planning Sciences 70: 100768.

16.

Cadoret

Padovano

(2024) Explaining the stringency of environmental policies: Domestic determinants or international policy coordination? European Journal of Political Economy 85: 102596.

17.

Chayes

(1995) The New Sovereignty: Compliance with International Regulatory Agreements. Cambridge: Harvard University Press.

18.

Chen

Cui

, et al. (2021) Long-term temperature and sea-level rise stabilization before and beyond 2100: Estimating the additional climate mitigation contribution from China’s recent 2060 carbon neutrality pledge. Environmental Research Letters 16(7): 074032.

19.

Chen

Tanchangya

(2022) Analyzing the role of environmental technologies and environmental policy stringency on green growth in China. Environmental Science and Pollution Research 29(37): 55630–55638.

20.

Chen

Sohail

Majeed

(2022) Revealing the effectiveness of environmental policy stringency and environmental law on environmental performance: Does asymmetry matter? Environmental Science and Pollution Research 29(60): 91190–91200.

21.

Corrocher

Mancusi

(2021) International collaborations in green energy technologies: What is the role of distance in environmental policy stringency? Energy Policy 156: 112470.

22.

Dang

H-AH

Serajuddin

(2020) Tracking the sustainable development goals: Emerging measurement challenges and further reflections. World Development 127: 104570.

23.

Deaton

(1985) Panel data from time series of cross-sections. Journal of Econometrics 30(1-2): 109–126.

24.

Dong

Wang

Zhang

, et al. (2024) China’s low-carbon policy intensity dataset from national-to prefecture-level over 2007–2022. Scientific Data 11(1): 13.

25.

Dutta

Quax

Crielaard

, et al. (2021) Inferring temporal dynamics from cross-sectional data using langevin dynamics. Royal Society Open Science 8(11): 211374.

26.

Fischer

(2014) Green industrial policies in China—the example of solar energy. In: Pegels

(ed.) Green Industrial Policy in Emerging Countries. London: Routledge, pp. 69–103.

27.

Franck

(1988) Legitimacy in the international system. The American Journal of International Law 82(4): 705–759.

28.

Galeotti

Salini

Verdolini

(2020) Measuring environmental policy stringency: Approaches, validity, and impact on environmental innovation and energy efficiency. Energy Policy 136: 111052.

29.

Gallagher

(2004) Free Trade and the Environment: Mexico, NAFTA, and Beyond. Stanford, CA: Stanford University Press.

30.

Gao

(2016) The Paris agreement and global climate governance: China’s role and contribution. China Quarterly of International Strategic Studies 2(03): 365–381.

31.

Garnaut

(2014) China’s role in global climate change mitigation. China and World Economy 22(5): 2–18.

32.

Gatti

Del Lo

Serranito

(2023) Unpacking the green box: Determinants of environmental policy stringency in European countries. EconomiX Working Paper 2023–25. Available at: https://hal.science/hal-04208688 (accessed 10 September 2024).

33.

Gormley

Matsa

(2014) Common errors: How to (and not to) control for unobserved heterogeneity. The Review of Financial Studies 27(2): 617–661.

34.

Goulder

Parry

(2008) Instrument choice in environmental policy. Review of Environmental Economics and Policy 2(2): 152–174.

35.

Guillerm

(2017) Pseudo-panel methods and an example of application to household wealth data. Economie et Statistique 491(1): 109–130.

36.

Gupta

Vegelin

Pouw

(2022) Lessons learnt from international environmental agreements for the Stockholm +50 conference: Celebrating 20 years of INEA. International Environmental Agreements: Politics, Law and Economics 22(2): 229–244.

37.

Hathaway

Shapiro

(2011) Outcasting: Enforcement in domestic and international law. Yale Law Journal 121: 252–349.

38.

Henkin

(1979) How Nations Behave: Law and Foreign Policy. New York: Columbia University Press.

39.

Hilton

Kerr

(2017) The Paris agreement: China’s ‘new normal’role in international climate negotiations. Climate Policy 17(1): 48–58.

40.

Huang

Yang

(2018) Policy change analysis based on “policy target–policy instrument” patterns: A case study of China’s nuclear energy policy. Scientometrics 117(2): 1081–1114.

41.

International Energy Agency (2024) Energy statistics. Available at: https://www.iea.org/data-and-statistics (accessed 20 September 2024).

42.

Jiang

Zhang

Hui

(2023) The impact of market and non-market-based environmental policy instruments on firms’ sustainable technological innovation: Evidence from Chinese firms. Sustainability 15(5): 4425.

43.

Johnstone

Haščič

Popp

(2010) Renewable energy policies and technological innovation: Evidence based on patent counts. Environmental and Resource Economics 45: 133–155.

44.

Keohane

(2005) After Hegemony: Cooperation and Discord in the World political economy. Princeton, NJ: Princeton University Press.

45.

Khan

(2018) Assessing cohort aggregation to minimise bias in pseudo-panels. Centre for Research in Economic Development and International Trade Research Paper, No. 18/01.

46.

Kim

(2015) T test as a parametric statistic. Korean Journal of Anesthesiology 68(6): 540–546.

47.

Kingsbury

(1998) The concept of compliance as a function of competing conceptions of international law. Michigan Journal of International Law 19: 345–372.

48.

Knill

Schulze

Tosun

(2012) Regulatory policy outputs and impacts: Exploring a complex relationship. Regulation and Governance 6(4): 427–444.

49.

Kruse

Dechezleprêtre

Saffar

, et al. (2022) Measuring environmental policy stringency in OECD countries: An update of the OECD composite EPS indicator. OECD Economics Department Working Paper, March 14. Available at: https://doi.org/10.1787/90ab82e8-en (accessed 24 September 2024).

50.

Kuznets

(2019) Economic growth and income inequality. In: Seligson

(ed.) The Gap Between Rich and Poor. London: Routledge, pp. 25–37.

51.

Lai

Chen

, et al. (2022) Investigating greenhouse gas emissions and environmental impacts from the production of lithium-ion batteries in China. Journal of Cleaner Production 372: 133756.

52.

Leal

Marques

(2022) The evolution of the environmental Kuznets curve hypothesis assessment: A literature review under a critical analysis perspective. Heliyon 8(11): e11521.

53.

Lebo

Weber

(2015) An effective approach to the repeated cross-sectional design. American Journal of Political Science 59(1): 242–258.

54.

Leebron

(2002) Linkages. American Journal of International Law 96(1): 5–27.

55.

Lian

(2024) Legitimacy-seeking: China’s statements and actions on combating climate change. Third World Quarterly 45(1): 171–188.

56.

Lisciandra

Migliardo

(2017) An empirical study of the impact of corruption on environmental performance: Evidence from panel data. Environmental and Resource Economics 68: 297–318.

57.

Chen

Nielsen

, et al. (2021) Combined solar power and storage as cost-competitive and grid-compatible supply for China’s future carbon-neutral electricity system. Proceedings of the National Academy of Sciences 118(42): e2103471118.

58.

Martínez-Zarzoso

Phillips

(2020) Freedom of the press, inequality and environmental policy. Environment and Development Economics 25(6): 537–560.

59.

Mitchell

(1994) Intentional Oil Pollution at Sea: Environmental Policy and Treaty Compliance. Cambridge, MA: MIT Press.

60.

Morin

J-F

Dür

Lechner

(2018) Mapping the trade and environment nexus: Insights from a new data set. Global Environmental Politics 18(1): 122–139.

61.

National Bureau of Statistics of China (2024a) Nationwide Gini Coefficient of Per Capita Disposable Income. People’s Living Conditions online database. Available at: https://data.stats.gov.cn (accessed 15 September 2024).

62.

National Bureau of Statistics of China (2024b) Per Capita GDP (yuan). National Accounts. Available at: https://data.stats.gov.cn (accessed 15 September 2024).

63.

Neumayer

(2002) Does trade openness promote multilateral environmental cooperation? World Economy 25(6): 815–832.

64.

Newig

(2007) Does public participation in environmental decisions lead to improved environmental quality? Towards an analytical framework. Communication, Cooperation, Participation (International Journal of Sustainability Communication) 1(1): 51–71.

65.

Niemeijer

(2002) Developing indicators for environmental policy: Data-driven and theory-driven approaches examined by example. Environmental Science and Policy 5(2): 91–103.

66.

Peat

Fikfak

van der Zee

(2022) Behavioural compliance theory. Journal of International Dispute Settlement 13(2): 167–178.

67.

Pommeret

Zhang

(2022) Stringency of environmental policy in China: When pollution drives bribery. Economic Modelling 117: 106048.

68.

Raustiala

Slaughter

A-M

(2002) International law, international relations and compliance. In: Carlsnaes

Risse

Simmons

(eds) Handbook of International Relations. London: Sage, pp. 538–558.

69.

Santos

Antunes

Baptista

, et al. (2006) Stakeholder participation in the design of environmental policy mixes. Ecological Economics 60(1): 100–110.

70.

Schachter

(1991) The emergence of international environmental law. Journal of International Affairs 44(2): 457–493.

71.

Schaffrin

Sewerin

Seubert

(2015) Toward a comparative measure of climate policy output. Policy Studies Journal 43(2): 257–282.

72.

Shen

Xie

(2018) Can China lead in multilateral environmental negotiations? Internal politics, self-depiction, and China’s contribution in climate change regime and Mekong governance. Eurasian Geography and Economics 59(5-6): 708–732.

73.

Simmons

(2010) Treaty compliance and violation. Annual Review of Political Science 13(1): 273–296.

74.

Tang

Qiu

Zhou

(2020) Does command-and-control regulation promote green innovation performance? Evidence from China’s industrial enterprises. Science of the Total Environment 712: 136362.

75.

Transparency International (2024) Corruption Perceptions Index. Available at: https://www.transparency.org/en/cpi (accessed 20 September 2024).

76.

Turken

Carrillo

Verter

(2017) Facility location and capacity acquisition under carbon tax and emissions limits: To centralize or to decentralize? International Journal of Production Economics 187: 126–141.

77.

UNFCCC (United Nations Framework Convention on Climate Change) (2015) Adoption of the Paris Agreement Available at: https://unfccc.int/resource/docs/2015/cop21/eng/l09r01.pdf (accessed 23 September 2024).

78.

United Nations (1973) United Nations Conference on the Human Environment Stockholm, 5-16 June 1972. Available at: https://documents.un.org/doc/undoc/gen/nl7/300/05/pdf/nl730005.pdf (accessed 23 September 2024).

79.

van Aaken

Simsek

(2021) Rewarding in international law. American Journal of International Law 115(2): 195–241.

80.

Verbeek

(2008) Pseudo-panels and repeated cross-sections. In: Mátyás L and Sevestre P (eds) The Econometrics of Panel Data: Fundamentals of Recent Development in Theory and Practice. Heidelberg: Springer, pp.369–383.

81.

Verbeek

Nijman

(1992) Can cohort data be treated as genuine panel data? Empirical Economics 17(1): 9–23.

82.

Von Stein

(2013) The engines of compliance. In: Dunoff

Pollack

(eds) Interdisciplinary Perspectives on International Law and International Relations. Cambridge: Cambridge University Press, pp. 477–501.

83.

Wang

Yan

Wang

, et al. (2020) The impact of environmental policy stringency on air quality. Atmospheric Environment 231: 117522.

84.

Weiss

Jacobson

(1998) Engaging Countries: Strengthening Compliance with International Environmental Accords. Cambridge, MA: MIT Press.

85.

Wooldridge

(2010) Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT press.

86.

(2023) A systematic review of climate policies in China: Evolution, effectiveness, and challenges. Environmental Impact Assessment Review 99: 1–12.

87.

Young

(1994) International Governance: Protecting the Environment in a Stateless Society. Ithaca, NY: Cornell University Press.

88.

Zehavi

(2012) New governance and policy instruments: Are governments going “soft”? In: Levi-Faur

(ed.) The Oxford Handbook of Governance. Oxford: Oxford University Press, pp. 242–254.

89.

Zhang

Gao

, et al. (2022) China’s environmental policy intensity for 1978–2019. Scientific Data 9(1): 75.

90.

Zhang

(2017) Are China's climate commitments in a post-Paris agreement sufficiently ambitious? Wiley Interdisciplinary Reviews: Climate Change 8(2): e443.

The Paris Climate Agreement and China's environmental policies: A pseudo-panel data model approach to analyzing the predictors of environmental policy intensity in China

Abstract

Keywords

Introduction

Theory

Compliance theory in international environmental law

Environmental policy intensity

Data and empirical strategy

Data description

Data sources

Variables

Empirical strategy and model specification

Pseudo-panel data model

Model specification

Results

Population-level analysis

Pseudo-panel regression analysis (cohort-level)

Robustness checks

Discussion

Conclusion

Footnotes

Acknowledgments

Contributorship

ORCID iDs

Funding

Declaration of conflicting interests

Notes

References