3.1 Interactions between elements in the process of decomposing governance targets

The process of decomposing governance targets in China involves two modes. The first is to directly allocate climate targets to actors, which is realized through China's ‘target responsibility system’ (TRS). In 2007, the Outline of the 11th Five-Year Plan (FYP) for National Economic and Social Development established a TRS centered on energy intensity. Subsequently, the TRS also included the total energy consumption and carbon intensity. Under the TRS, the central government sets the overall national climate target, and allocates it to provinces. Each province then distributes its target to municipalities until all levels of government are assigned a target (Cheng et al., Reference Cheng, Gu, Tan, Yan and Sheng2022).

The second is to translate governance targets into climate actions, and guide and incentivize the participation of a wider range of actors. In 2011, the Outline of 12th FYP formally included targets and priorities for addressing climate change. On this basis, the State Council took the lead in releasing the Plan for Controlling Greenhouse Gas Emissions during the 12th FYP Period. Subsequently, several ministries collaborated in formulating sector-specific or technology-specific FYPs according to the climate change priorities set out in the Outline of 12th FYP. Currently, China's carbon abatement actions mainly involve four aspects: improving the energy efficiency of energy-intensive industries, increasing the share of low-carbon industries to realize the transformation of the economic structure, increasing the share of non-fossil energy in the primary energy mix, and increasing carbon sequestration through afforestation.

3.2 Interactions between elements in the process of guiding the participation of various actors

China guides and incentivizes the participation of various actors in climate actions mainly through the formulation and implementation of a set of climate policies. Climate policies can be divided into three types, i.e. regulation instruments, economic instruments, and information instruments (Rogge & Reichardt, Reference Rogge and Reichardt2016). Regulation instruments can directly curb emissions from companies in key industries. For example, China has continuously improved energy efficiency standards for coal-fired power plants.

Compared with regulation instruments, economic instruments are more commonly used in China (Tan et al., Reference Tan, Wang, Gu, Kong and Zeng2022a, Reference Tan, Guo, Fan, Guo, Wang, Zeng, Su and Sun2022b). For energy-intensive and high-emitting enterprises that take responsibility for reducing emissions, there will be additional costs associated with reducing product output, purchasing specialized pollution control equipment, or adjusting the production process. To incentivize these actors, China tends to introduce financial support and subsidy policies. In addition, in order to upgrade industrial structure and develop emerging industries, China has also implemented economic policies, such as tax relief, for low-carbon technology and service providers. In addition, the scope of information instruments is relatively broad, encompassing aspects such as education, social advocacy, and voluntary mechanisms, making it challenging to provide a unified characterization of their function mechanisms. Consequently, the theoretical framework to be constructed in the following section will primarily focus on regulation instruments and economic instruments.

4.1 Module 1: decomposing the carbon neutrality target and identifying climate actions

In China's climate governance practice, key climate actions to meet climate targets are usually determined by drawing lessons from previous FYPs and collecting information from a series of consultations with experts and subnational governments (IEA, 2021). Given that the actions are all target-oriented, the framework needs to explore possible future scenarios. Consequently, this paper attempts to use the low-carbon pathways stimulated by the IAMs adopted by several governments to assess climate governance around the world (Farmer et al., Reference Farmer, Hepburn, Mealy and Teytelboym2015).

With advances in modeling techniques, current IAMs can characterize different climate change response systems and project their low-carbon transition pathways. The output variables include technology and funding portfolio, etc. Therefore, this framework translates the output variables into a set of climate actions P [p₁, …, p_h] for China to realize the carbon neutrality target. For a given action p, the government can track and evaluate the progress of its implementation by setting several sub-targets, denoted by O^p [O^p₁, …, O^p_y]. For example, China has set installed capacity targets of different renewable energy sources in FYPs for renewable energy since 2007.

4.2 Module 2: selecting actor's objectives and assessing the effects

After obtaining the action set P, the second module will focus on analyzing the participation process of various actors, including how to characterize actors' motivation to engage in climate actions, and the impact of different institutions and policies on actors' motivation. In China, the government, businesses, the public, research institutions, think tanks, and NGOs, all play a role in climate actions within climate governance system (Wang et al., Reference Wang, Liu and Wu2018; Zhang, Reference Zhang2024). For each actor i involved in climate action p, it is assumed that its motivation is determined by the effects of action p on its objectives. For example, objectives with greater influence on the government's decisions mainly include investment and employment. For companies, profit may be the main concern.

The set of actors involved in climate action p is denoted by A [1, …, a], and the set of objective of actor i is denoted by w_i. The concatenation of w_i must satisfy the condition that it can fully cover the sub-target set O^p, as shown in Equation 2–1:

(2-1)$$O^p\subseteq \bigcup\limits_{i = 1}^a {w_i} $$

On this basis, referring to Jakob et al. (Reference Jakob, Flachsland, Christoph Steckel and Urpelainen2020), the motivation of actor i for climate action p depends on the impact of p on objective k and the relative weight of objective k for actor i, as shown in Equation 2–2:

(2-2)$$G_i( p) = \sum\limits_{k = 1}^{b_i} {\alpha _{i, k}\cdot F_{i, k}( p) } $$

G_i(p) is the total target function that measures the motivation of actor i toward climate action p. b_i denotes the number of objectives of actor i. The weight α_i,k indicates the importance of objective k to actor i. If objective k is not important to actor i, α_i,k will take a small value; if objective k is among the core interests of actor i, α_i,k will take a relatively large value among the objectives in w_i. At the same time, the total objective function G is also affected by the specific effects of p on objective k, denoted by F_k(p).

Based on the characterization of actors' motivation in Equation 2–2, the mechanism of action of governance mechanism is expressed as increasing actors' motivation toward climate action p through the implementation of institution C and policy Q. In terms of institutional arrangements, the theoretical framework mainly considers the TRS (C_t), as shown in Equation 2–3:

(2-3)$$G_i( p\vert C_t) = \sum\limits_{k = 1}^{b_i} {\alpha _{i, k}( C_t) \cdot F_{i, k}( p) } $$

G_i(p|C_t) represents the motivation of actor i affected by institution C_t. C_t can increase the weight α_i,k by assigning the objective k to actor i (e.g. energy intensity and carbon intensity targets) and tracking target performance through the Statistics Indicators, Monitoring, and Examination (SME) system. In the TMA framework, the mechanism of action is expressed by setting the weight α_i,k as a function of C_t. The institutional arrangements can not only directly adjust the weight of the climate objective, but also indirectly enhance the relative weight of the climate objective by changing the weight of other objectives, so as to increase actors' motivation toward climate actions.

As for climate policies, different policy instruments can impose various effects on the total target function G_i(p). The mechanism of action of regulation policy instrument (denoted by Q_o) can be described as requiring actor i to take the climate action, through which specific climate targets can be met, as shown in Equation 2–4.

(2-4)$$G_i( p\vert Q_o) = \sum\limits_{k = 1}^{b_i} {\alpha _{i, k}\cdot F_{i, k}( Q_o) } $$

As the same time, the economic policy instrument (denoted by Q_e) occupies a dominant position in China's current climate policy mix. In the TMA framework, its mechanism of action is expressed as increasing the motivation of actor i toward climate action p by affecting F_i,k, i.e. increasing the positive effects of objective k, as shown in Equation 2–5:

(2-5)$$G_i( p\vert Q_e) = \sum\limits_{k = 1}^{b_i} {\alpha _{i, k}\cdot F_{i, k}( p\vert Q_e) } $$

4.3 Module 3: Designing the governance mechanism

The implementation of institution C or policy Q may impose different impacts on different actors, leading to distributional effects between them. In particular, ambitious climate actions or stringent climate policies may entail losses for certain actors, thus hindering the implementation of policies. Therefore, the finalization and introduction of policies often needs to go through an interactive process where the distributional effects between different actors will be coordinated. In response to this process, it is hypothesized that the most feasible governance mechanism is the solution that can best satisfy actors' objectives with strong policymaking influence. The coordination process is then characterized through the following two equations:

(2-6)$$J( p) = \sum\limits_{i = 1}^a {\delta _i( C) \cdot } G_i( p\vert Q, \;C) $$

(2-7)$$Z_p^\ast ( C^\ast , \;Q^\ast ) = \mathop {\max }\limits_{C, Q} J( p) $$

J(p) denotes the political support for institution C and policy Q aimed at promoting action p, reflecting the political feasibility of the governance mechanism. δ_i reflects actors' decision-making influence on action p. The finally obtained governance mechanism Z*_p (C*, Q*) is the option receiving the greatest political support and thus the most feasible one. According to Equation 2–6, there are two ways to obtain Z*_p (C*, Q*). The first is to change δ_i. The second is to reduce negative impacts of p or the original policy Q on actors that have the most pronounced influence on governance decision-making.

4.3 Application of the theoretical framework

By integrating the above three modules, we have obtained the theoretical framework, as shown in Figure 1. In general, after translating China's carbon neutrality target into climate actions P [p¹, …, p^z], the TAM framework takes each specific action p as the unit of analysis, intending to find out the governance mechanism Z*_p that can effectively promote action p.

Figure 1.

Graphical representation of the TAM framework.

The obtaining of governance mechanism Z*_p can be summarized in Equation 2–8. The target function shows that, compared to traditional governance research that emphasizes obtaining the solution with the least cost or the greatest benefit, the TAM framework developed in this paper pays more attention to the support of actors, in order to find out a more feasible solution. The acquisition of this solution takes into account the objective effects F_i (p), the weight of objectives α_i, and the influence on the decision-making process δ_i. At the same time, several constraints have to be taken into account. In particular, the set of objectives W of all actors involved in action p must fully cover the set of sub-targets O^p. Institution C mainly includes the TRS C_t, and policy includes regulation instruments Q_o and economic instruments Q_e. In fact, these constraints represent to some extent the characteristics of each element in China's climate governance system.

(2-8)$$\eqalign{& \max J( C, \;Q) = \sum\limits_{i = 1}^a {\delta _i( C) \cdot } G_i( \sum\limits_{k = 1}^{b_i} {\alpha _{i, k}( C) \cdot F_{i, k}( p\vert Q) } ) \cr & s.t.\;C\in C_t \cr & Q\in ( Q_o, \;Q_e) \cr & O^p\subseteq W } $$

when applying the theoretical framework to case studies, the three modules correspond to three steps. Step 1 is to translate the transition pathways of IAMs into climate actions. As differences between IAMs, especially in their solution methods, parameter settings, and scenario design, often lead to inconsistent conclusions on key issues such as technology choices and socio-economic impacts (Kong et al., Reference Kong, Tan, Gu and Yan2023), the theoretical framework considers the use of multi-model comparison analysis to obtain more robust results.

Step 2 involves selecting actors' main objectives and assessing the effects of the objectives, which mainly relies on quantitative modeling. The commonly used methods include the system dynamics model, multiple-actor evolutionary game model, and other quantitative models that can represent the interactions between various actors.

Compared with regulation instruments and economic instruments characterized by quantitative modeling, the weight of objectives α, and the influence on decision-making process δ are usually characterized as relatively ‘large’ and ‘small’, which are difficult to assign precise values, and thus are not suitable for quantitative modeling. As a result, Step 3, referring to Lockwood et al. (Reference Lockwood, Kuzemko, Mitchell and Hoggett2016) and Andrews-Speed (Reference Andrews-Speed2016), is to introduce historical institutionalism to analyze the decision-making process. It focuses on how institutions shape the distribution of power and power structures, which is compatible with the hypotheses of maximum political support in this theoretical framework. To be specific, Step 3 begins with an analysis of institutions involved in climate action p to clarify the power structure between the actors, which can help understand the weight of objectives α and the influence on the decision-making process δ. Then, the policy feedback theory under historical institutionalism is applied to analyze how to ultimately obtain a feasible governance mechanism by creating positive policy feedback or mitigating negative policy feedback. In conclusion, the TAM framework is able to accommodate different perspectives on climate governance, as summarized in Table 1.

Table 1.

Climate governance from different perspectives represented in the ATM framework