Navigating Uncertainty: How Climate Policy Uncertainty Drives Firms’ Green Transformation

Abstract

As climate change intensifies, climate-related policy instruments have become essential tools for governments to mitigate climate risks. The formulation of such policies typically involves balancing economic growth and ecological goals while ensuring fairness and feasibility, often resulting in frequent policy adjustments. Drawing on panel data from A-share listed firms in China between 2008 and 2023, this study conducts a comprehensive investigation into how climate policy uncertainty affects corporate green transformation, with a particular focus on the underlying mechanisms. The empirical evidence shows that climate policy uncertainty serves as a significant catalyst in promoting firms’ green upgrading. Mechanism analysis indicates that this uncertainty has enhanced enterprises’ environmental awareness, encouraged their green investment, and promoted strategic planning oriented towards environmental sustainability. Further heterogeneity analysis reveals that the positive effect is especially pronounced in state-owned enterprises, labor-intensive industries, and firms exposed to high levels of climate risk. This study enriches the understanding of corporate behavioral responses under climate-related policy uncertainty and offers practical implications for optimizing policy design to support sustainable transformation in the corporate sector.

1. Introduction

Climate-related risks have escalated into a pressing global concern [1]. In 2024, carbon dioxide emissions reached an unprecedented 40.6 billion tons, with atmospheric CO₂ levels now exceeding pre-industrial concentrations by over 50%. According to the World Meteorological Organization (WMO), the average global surface temperature has risen by 1.45 °C relative to the pre-industrial baseline, approaching the critical 2 °C threshold outlined in the Paris Agreement. This persistent warming trend has not only escalated the occurrence of severe weather phenomena, but also exerted far-reaching effects across ecosystems, public health systems, agricultural productivity, and water availability. Recent climate anomalies—such as prolonged heatwaves in Europe, widespread wildfires in Australia, and the record-breaking rainfall in Henan, China in 2021—have resulted in significant human and economic losses. In response, it is essential to curb the pace of global warming, proactively tackle the multifaceted challenges posed by climate instability, and foster development pathways that are both inclusive and ecologically sustainable.

As a major country committed to global responsibility, China has actively engaged in international efforts toward climate governance. In 2020, it announced its targets for carbon peaking and carbon neutrality, and the 20th National Congress of the Communist Party of China further emphasized progress toward these “dual carbon” goals and China’s deeper involvement in international climate collaboration. Since then, China has introduced a range of climate-related policy measures. Governments across administrative levels have also intensified efforts to advance climate-related initiatives. Nonetheless, climate change remains a highly intricate issue, and policy design must navigate the trade-offs between economic growth, environmental integrity, equity, and operability. Moreover, substantial regional disparities exist in how local authorities interpret and enforce such policies—particularly concerning enforcement intensity, subsidy expectations, and implementation boundaries. These inconsistencies have amplified climate policy uncertainty and introduced operational risks for firms [2]. Consequently, enhancing firms’ responsiveness to policy fluctuations has become crucial for sustaining long-term green development and preserving competitive advantage.

Corporate green transformation refers to a development path shaped by the principles of ecological enhancement and sustainable economic progress. This approach aligns closely with China’s national strategy for “green development” [3], emphasizing the integration of economic advancement and environmental stewardship [4]. By adopting green transformation strategies, firms can elevate their environmental reputation, attract greater public visibility, and obtain benefits from environmental incentives or subsidies—while also improving their overall operational performance [5]. Given the rising emphasis on sustainable growth and the uncertainties associated with evolving climate policies, transitioning toward green practices has become a strategic necessity for enterprises [6]. Yet, due to firms’ limited ability to anticipate the scope and timing of policy interventions, climate-related policy ambiguity continues to exert substantial influence on corporate strategic planning—particularly in areas such as innovation, capital deployment, and market positioning [7,8].

Golub et al. (2020) [9] introduced a corporate behavioral framework to illustrate how policy uncertainty drives firms to adopt a “dual delay strategy”—namely, postponing both emission reduction actions and the acquisition of emission permits. Similarly, Zhao et al. (2025) [10] observed that under uncertain policy environments, firms are more likely to perceive elevated operational risks, which in turn discourages investment. These findings raise key questions: To what extent does climate policy ambiguity influence firms’ transition toward green practices? What mechanisms mediate this relationship? Furthermore, how do such effects vary depending on firm ownership, sector characteristics, or exposure to climate-related risks? A comprehensive examination of these issues can help policymakers identify the core constraints firms face in advancing sustainability efforts, refine climate policy design, and deliver clearer strategic signals to guide businesses in navigating green transformation and achieving long-term sustainable objectives.

This paper selects A-share listed firms in China from 2008 to 2023 as the research sample to examine the impact of climate policy uncertainty on firms’ green transformation. The specific research route is detailed in Figure 1 The main marginal contributions of this study are as follows:

(1)

Shifting the focus of climate risk research from physical risks to policy uncertainty risks. While existing studies have largely focused on how climate-related disasters—such as extreme heat, floods, and other events—influence firms’ green transition, the intensifying global climate crisis has brought growing attention to climate policy as a crucial driver. With China emphasizing its “dual carbon” goals, the government has increasingly prioritized ecological development. Accordingly, the uncertainty surrounding climate policy has become a rising concern in shaping corporate strategies. In response, this paper broadens the climate risk research agenda by incorporating policy-related risks, thus offering a complementary perspective to the existing body of literature.

(2)

Earlier empirical studies often relied on the U.S.-based Climate Policy Uncertainty (CPU) Index proposed by Gavriilidis (2021) [11] to quantify policy-related risk in China. However, as this index is constructed using U.S. climate policy data, its relevance for China-specific research remains limited. To address this gap, we utilize the CPU Index developed by Ma et al. (2023) [12], which is specifically tailored to reflect China’s policy environment. This approach allows for a more precise evaluation of policy-driven uncertainties in China, enhances the empirical robustness of climate risk studies, and improves the applicability of research outcomes.

(3)

In terms of internal mechanisms, our analysis has found that the uncertainty of climate policies can influence the green transition by enhancing enterprises’ environmental awareness, stimulating green investment and promoting sustainable development strategies. These results not only contribute to theoretical discussions but also provide meaningful insights for policymakers and enterprises seeking to effectively address policy uncertainties.

2. Literature Review

With climate change exerting increasingly profound influences on the global economy, scholarly interest in this topic has surged. The current literature typically investigates climate-related issues from both macroeconomic and microeconomic dimensions. At the macroeconomic scale, Dell et al. (2012) [13] analyzed how historical fluctuations in temperature affect the economy, revealing that rising temperatures significantly hinder economic growth in low-income nations. Later, Arbex and Batu (2020) [14] incorporated temperature shocks into a resource-augmented real business cycle framework, reaffirming that abnormal warming trends pose adverse effects on national output. Dietz et al. (2016) [15] further highlighted that climate-related risks reduce the value of financial assets, potentially triggering macroeconomic instability. Kahn et al. (2021) [16] showed that national economic outcomes are closely linked to variations in temperature and precipitation patterns. Gong et al. (2025) [17] emphasized how extreme heat events can reshape social dynamics and influence societal adaptation strategies in the context of climate change.

At the microeconomic level, climate policy uncertainty also exerts a significant influence on firm behavior. Nguyen and Phan (2020) [18], using the example of Australia’s Kyoto Protocol ratification, observed a marked reduction in financial leverage among high-emission enterprises, particularly those with tight financial constraints. Pankratz et al. (2023) [19] analyzed temperature patterns across 93 countries and demonstrated that extreme weather events can substantially affect firms’ financial outcomes. Similarly, Javadi and Masum (2021) [20] found that climate risk contributes to higher borrowing costs and intensifies financial stress among firms. In contrast, Wang et al. (2023) [21] showed that firms increase risk-taking under extreme heat, and that greater transparency in climate risk disclosure is positively associated with firm valuation.

In response to the escalating global climate crisis, nations worldwide have adopted various regulatory policies aimed at managing human-induced environmental impacts. Yet, the unpredictability stemming from frequent shifts in climate-related regulations has emerged as a major challenge for economic stability. As a result, understanding the economic consequences of climate policy uncertainty has become an increasingly urgent area for academic investigation. One essential component of this inquiry lies in accurately quantifying policy uncertainty. In climate economics, the ability to precisely assess such uncertainty forms a foundational basis for empirical inquiry. With advancements in computational tools, including big data analytics, text mining, and machine learning, these methods have been widely applied in quantitative studies of policy unpredictability. Baker et al. (2016) [22] pioneered the use of text-based metrics to create the Economic Policy Uncertainty Index, which offered a novel methodological structure for subsequent research. Building on this, Gavriilidis (2021) [11] further extended the framework for climate-specific applications. Subsequently, Gavriilidis (2021) [11] constructed a U.S. Climate Policy Uncertainty Index based on the number of articles related to climate change, uncertainty, and legislative regulation published in eight major newspapers, and verified the accuracy and reliability of this measurement by comparing the peaks of the index with actual climate policy changes. This approach laid the foundation for the quantitative analysis of climate policy uncertainty.

Furthermore, Lee and Cho (2023) [23] argued that China, as a major carbon emitter, warrants the development of a localized climate policy uncertainty index. To this end, they constructed the Twitter-based China Climate Policy Uncertainty Index (TC-CU), derived from tweets by global users. The index’s validity was confirmed by examining its correlation with the U.S. Climate Policy Uncertainty Index [23]. Nevertheless, since Twitter content primarily captures international discourse and lacks alignment with China’s domestic policy context, concerns about its representativeness remain. To overcome this limitation, Ma et al. (2023) [12] proposed an alternative approach by extracting features from six mainstream Chinese newspapers. They utilized the MacBERT deep learning algorithm to identify policy-relevant signals associated with climate policy uncertainty. This method not only improves model adaptability and robustness but also helps reduce dictionary-based biases commonly found in earlier textual analysis techniques. Consequently, their work facilitated the creation of a multi-level climate policy uncertainty index tailored to China’s national, provincial, and municipal contexts, offering a novel empirical foundation for climate-related policy analysis.

Recent research has increasingly focused on how uncertainty in climate-related policies influences firm behavior. Jung et al. (2018) [24] demonstrated that rising uncertainty in climate governance elevates corporate financing costs [24]. This suggests that climate policy ambiguity not only affects firms’ capital structure decisions but may also intensify operational vulnerabilities. Building on this, Ren et al. (2022) [25] reported that such uncertainty significantly reduces firms’ total factor productivity and limits available cash flow, with the adverse effects being especially evident among low-efficiency firms. These findings emphasize the direct consequences of regulatory unpredictability on firms’ operational efficiency, particularly through channels such as resource allocation and business strategy. Persakis (2024) [26] found that climate policy ambiguity undermines both financial outcomes and carbon emission control, although it may positively influence firms’ ESG (Environmental, Social, and Governance) indicators to some extent. Siddique et al. (2024) [27], applying advanced econometric methods including wavelet coherence and quantile-based models, explored how regulatory uncertainty related to climate issues influences fossil fuel and alternative energy markets. Their results indicate that this uncertainty significantly impacts returns in fossil energy derivatives markets.

In conclusion, much of the current literature on climate policy uncertainty relies on the index developed by Gavriilidis (2021) [11], which, while influential, lacks contextual relevance for China’s institutional and policy environment. As a result, its applicability to China’s economic and social conditions is limited. Moreover, prior studies have predominantly concentrated on the physical consequences of climate change, with insufficient attention to policy-induced risks. To bridge this gap, this study adopts a policy-oriented perspective and employs the Climate Policy Uncertainty Index constructed by Ma et al. (2023) [12] as a proxy variable to quantify climate-related policy uncertainty in the Chinese context. This approach offers a meaningful supplement to existing empirical frameworks. More critically, earlier research has yet to fully clarify the underlying mechanisms through which policy uncertainty affects firms’ transitions toward green development. Accordingly, this study investigates how internal channels transmit the influence of policy uncertainty on corporate green strategies, thereby addressing both the pressing demand for theoretical insight and the practical need for firms to cope with regulatory volatility—ultimately supporting the broader promotion of sustainability in the business sector.

3. Theoretical Analysis and Research Hypotheses

Amid the growing emphasis on global climate governance, the regulatory landscape in which firms operate has become increasingly intricate. Within this context, uncertainty surrounding climate-related policies has become a key variable shaping corporate strategic decisions. While prior research has primarily examined its impact on firm-level investment, innovation, and financing behavior, the theoretical understanding of how such uncertainty drives firms’ transition toward green practices remains underdeveloped. To fill this academic void, this study proposes an integrated theoretical framework—drawing upon uncertainty theory, the resource-based view, and institutional theory—to examine how climate policy uncertainty influences firms’ green transformation through multiple transmission channels.

3.1. Corporate Environmental Attention

When firms face ambiguity about the direction and specifics of climate policy implementation, they may worry that forthcoming policy changes could disrupt current operations. As a result, businesses are incentivized to prioritize environmental strategies to buffer against regulatory volatility. To address such uncertainty, firms often enhance their environmental initiatives to reduce exposure to climate-related policy shifts. As regulatory ambiguity escalates, shifts may occur in both investor and consumer preferences. From the investment perspective, stakeholders might perceive climate policy uncertainty as a key determinant of firms’ long-term stability and value creation, prompting a preference for firms with robust climate risk management practices [28]. On the consumer side, increasing environmental consciousness may lead to greater demand for sustainable products [29]. Consequently, in the face of regulatory unpredictability, companies can manage uncertainty more effectively and appeal to environmentally conscious investors and consumers by actively improving their sustainability commitments.

Firms with high environmental attention typically possess a stronger forward-looking advantage in responding to climate policy risks. By optimizing emission management and adjusting production models, firms can effectively reduce the operational risks caused by climate policy uncertainty [30]. Meanwhile, leveraging their green advantages, these firms have greater competitiveness in securing government green subsidies and green credit support, thus reducing financing costs and providing strong support for green transformation [31]. In addition, enhanced environmental attention encourages firms to optimize resource allocation, promote green supply chains, adjust energy structures, and refine operational processes, thereby accelerating the green transformation process [32]. These measures not only enhance a firm’s market competitiveness, but also shape a positive green brand image, attracting more investors and consumers with green preferences [33]. Therefore, improving corporate environmental attention is not only an effective strategy to cope with climate policy uncertainty, but also a key pathway to promote green transformation and enhance long-term competitive advantage. Accordingly, the following hypothesis is proposed:

Hypothesis 1.

The uncertainty of China’s climate policy promotes Chinese firms’ green transformation by enhancing their environmental attention.

3.2. Enterprise Green Investment

Climate policy uncertainty is primarily reflected in the unpredictability of policy formulation timing, implementation intensity, and related details, which presents firms with greater challenges in decision-making. Firms must not only cope with increasing operational risks but also bear higher compliance costs, inevitably increasing the complexity of decision-making and future planning. According to dynamic competition theory, competition among firms is not static, but rather a dynamic process that continuously evolves under the influence of external information. Firms must continuously innovate in order to maintain a leading position in intense market competition. Therefore, when confronted with climate uncertainties, enterprises will no longer rely solely on existing technologies to maintain their competitive edge. Instead, they will increase investment in green research and development to form professional R&D teams, thereby seizing the initiative in the green economic transformation and achieving a “first-mover advantage” [34]. At the same time, firms will also promptly adjust their production methods and R&D strategies, introducing advanced environmentally friendly production technologies to optimize production processes and achieve green production. Through these measures, firms can remain competitive in an environment of policy uncertainty.

Furthermore, according to prospect theory, firms’ decision-making is not only based on current information but is also driven by expectations of future scenarios. In the context of increasingly severe climate issues, firms may anticipate that the government will introduce stricter environmental regulations, carbon reduction requirements, or other green policies. To cope with such possible changes, enterprises usually manage expectations by increasing green investment, thereby reducing the risks and costs associated with future policy shifts [9]. This forward-looking strategy not only enables firms to rapidly adjust when policies are implemented, but also allows them to seize the initiative in the process of green economic transformation.

The green investment of enterprises directly determines the level of green innovation and is the core driving force for their green transformation. Continuous R&D investment not only helps firms accumulate technological expertise and provides a solid foundation for green transformation, but also enhances their ability to adapt to market demand and policy changes [35]. In summary, green investment enables firms to respond quickly to policy changes, optimize production processes, and ensure the smooth progress of green transformation. Accordingly, the following hypothesis is proposed:

Hypothesis 2.

The uncertainty of China’s climate policy promotes Chinese firms’ green transformation by increasing their green R&D investment.

3.3. Corporate Green Strategic Orientation

Amid intensifying climate policy uncertainty, firms face the dual pressures of resource constraints and rising costs, requiring more prudent evaluation of business models, investment decisions, and resource allocation. In this context, a green strategy, as a long-term sustainable development path, becomes an important choice for firms seeking to enhance competitiveness and reduce policy risks [32]. First, a green strategy helps strengthen firms’ adaptability to climate policy changes and operational resilience, thereby reducing potential losses from policy adjustments [36]. Second, a green strategy drives firms to optimize production processes, introduce circular economy models, and improve resource utilization efficiency. This not only enhances firms’ survival ability under resource constraints but also promotes the overall shift in firms towards low-carbon and environmentally friendly development, forming a more sustainable competitive landscape [37]. In addition, a green strategy can enhance corporate brand value. As consumers’ environmental awareness increases, firms that take the lead in green initiatives can effectively attract environmentally oriented consumer groups and gain a more advantageous position in market competition [29].

Corporate green strategic orientation plays a key role in promoting green transformation. First, it helps firms clarify green development goals, provides guidance for action, and ensures the systematic advancement of green transformation [38]. Second, a green strategic orientation encourages firms to strengthen internal management, establish sound green management systems, and ensure the effective implementation of green transformation. Moreover, optimizing resource allocation is a crucial step in the realization of green transformation. By optimizing resource allocation, firms can promote greening across the entire production, marketing, and service chain, improve resource use efficiency, and reduce operational costs, thus advancing green transformation [39]. In summary, a green strategic orientation not only enhances firms’ adaptability to policy uncertainty, but also promotes their long-term sustainable development, thereby facilitating green transformation. Accordingly, the following hypothesis is proposed:

Hypothesis 3.

The uncertainty of China’s climate policy promotes Chinese firms’ green transformation by encouraging a green strategic orientation.

4. Research Design

4.1. Data Sources and Processing

The core explanatory variable—China’s Climate Policy Uncertainty Index—adopts the version developed by Ma et al. (2023) [12], which was published in Scientific Data, a sub-journal of Nature [14]. This index is used to capture policy-related uncertainty. The dependent variable, firms’ green transformation performance, is derived from textual disclosures of listed companies, with the raw data sourced from the Shanghai and Shenzhen Stock Exchanges. Additional firm-level and financial information is retrieved from the China Stock Market & Accounting Research (CSMAR) database and the China Research Data Service (CNRDS). Industry classifications follow the coding standards of the China Securities Regulatory Commission (CSRC), while geographic attributes such as province and city are identified based on firm registration details contained in the CSMAR database.

This study selects A-share listed firms on the Shanghai and Shenzhen Stock Exchanges from 2008 to 2023 as the empirical sample. To improve data integrity and ensure the robustness of the empirical findings, the following sample screening and data preprocessing procedures are implemented: (1) Firms labeled as ST or PT are excluded. (2) Companies operating in the financial and real estate industries are removed. (3) Observations with substantial data omissions or abnormal values—such as those with a debt-to-asset ratio exceeding 1—are excluded. (4) All continuous variables are winsorized at the 1st and 99th percentiles. After applying the above procedures, the final sample includes 29,830 firm-year observations.

4.2. Baseline Regression Model

To investigate the impact of climate policy uncertainty on the level of firms’ [green transformation/other specific variable], the following baseline regression model is constructed, as shown in Equation (1):

$${GTS}_{i,t}={\alpha }_0+{\alpha }_1{CCPU}_{i,t}+\sum {\alpha }_{k}Contro{l}_{it}+{\lambda }_i+{\eta }_{j,t}+{\epsilon }_{i,t}$$

(1)

where the subscripts i, t, and j denote firm, year, and industry, respectively. ${GTS}_{i,t}$ represents the green transformation of firm i in year t; ${\alpha }_0$ is the intercept term; ${CCPU}_{i,t}$ denotes the climate policy uncertainty faced by firm i in year t at the prefecture-level city; the coefficient ${\alpha }_1$ is the main focus of interest—if ${\alpha }_1$ is significantly greater than zero, it indicates that climate policy uncertainty promotes firms’ green transformation. $Contro{l}_{it}$ represents a set of control variables. In addition, firm fixed effects ${\lambda }_i$ and industry-year joint fixed effects ${\eta }_{j,i}$ are controlled for ${\epsilon }_{i,t}$ is the random error term.

4.3. Variable Selection and Measurement Methods

4.3.1. Dependent Variable

Corporate green transformation describes the process through which firms align with green development principles and transition away from conventional production patterns characterized by high energy use, heavy pollution, and excessive emissions. Instead, they adopt environmentally responsible, energy-efficient, and low-carbon operational models to realize high-quality growth that integrates both environmental sustainability and economic performance [40]. Consistent with prior literature, this study utilizes textual data analysis to measure firms’ engagement in green transformation activities [41,42]. The specific quantification process is as follows. First, drawing on authoritative policy documents such as the Environmental Protection Law, Made in China 2025, and the White Paper on Green Manufacturing Standards, we construct a feature word dictionary that covers the core dimensions of corporate green transformation (see

Table 1. Feature Words for Green Transformation.

Dimension of Green Transformation	Feature Words
Advocacy and Promotion	green development, circular economy, low-carbon pathways, sustainable development, sustainable growth strategies, eco-awareness, green lifestyle, green consumption, ecological protection awareness, environmental protection awareness, ecological civilization system, environmental protection obligations, environmental commitment, ecological protection concept, ecological protection obligations
Strategic Vision	eco-revitalization, sustained growth trajectory, low-emission development plans, green financial tools, picturesque towns and villages, beautiful China campaign, clean water initiative
Technological Innovation	resource efficiency enhancement, advanced recycling capabilities, green tech solutions, pollution prevention strategies, innovation in environmental protection, governance-driven green systems
Pollution Control	reducing energy consumption, resource conservation, ecological restoration, recycling, energy conservation and emission reduction, pollution prevention, ecological restoration, environmental governance system, environmental remediation, environmental justice, co-governance by society, co-governance by all people, source prevention, ecological remediation, pollution control, energy-saving and emissions mitigation actions, pollution control mechanisms, ecosystem restoration, environmental governance structure, eco-remediation practices, environmental equity, inclusive co-governance
Monitoring and Management	environmental risk, biodiversity, ecosystem, ecological function, ecological service functionality, environmental safeguarding, nature-respecting philosophy, adaptation to ecological changes, preservation of natural systems, respect for nature, adapt to nature, protect nature, corporate environmental responsibility, corporate environmental governance, corporate green upgrading, corporate ecological responsibility

for details). Second, annual report texts are collected through Python 3.9-based web scraping, and keyword frequencies are counted based on the dictionary. To enhance measurement validity, several adjustments are applied: (i) sentences containing negations or conditional expressions (e.g., “not,” “has not yet,” “planned”) are excluded to avoid overstating unimplemented actions; (ii) keywords are required to co-occur with action-oriented verbs (e.g., “invest,” “upgrade,” “certify,” “achieve,” “commission”) in the surrounding context, thereby emphasizing actual implementation rather than rhetorical slogans; and (iii) irrelevant uses of the term “green transformation” are filtered out. Finally, the logarithm of the adjusted and normalized keyword frequencies (e.g., standardized by report length) is taken to construct the indicator of corporate green transformation. This method captures firms’ disclosed signals of substantive green practices while minimizing noise from “greenwashing”-type statements. To further validate the reliability of this indicator, we incorporate alternative or external measures such as green patents in robustness checks, ensuring that the textual proxy reflects firms’ genuine engagement in green transformation rather than merely symbolic disclosure. It is worth noting that we also measured the mechanism variable ‘environmental concern’ using text analysis, but its keywords are not those in the table below.

4.3.2. Independent Variable

This study adopts the methodology developed by Ma et al. (2023) [12] to construct China’s Climate Policy Uncertainty Index. Specifically, a total of 1,755,826 news articles were initially retrieved from six major Chinese media outlets—People’s Daily, Guangming Daily, Economic Daily, Global Times, Science and Technology Daily, and China News Service—via the WiseNews database. Advanced natural language processing techniques, including BERT and HanLP, were employed to extract location-specific entities from the texts and map them to corresponding provinces or cities. Based on this, the city-level index of climate policy uncertainty was developed through a multi-step procedure involving text preprocessing, manual review, index modeling, and robustness checks. To ensure index reliability, additional newspapers such as China Energy News, China Science Daily, and China Environment News were incorporated in the validation process. The new index demonstrated a strong correlation with existing benchmarks, with a match rate of 84.47%, supporting its robustness. Compared with prior approaches, this method draws from a broader and more authoritative data foundation, and offers improved accuracy in capturing climate-related policy uncertainty within the Chinese context. See Figure 2 for more details. It is noteworthy that this study employs prefecture-level city indexes for the analysis.

4.3.3. Control Variables

Referring to the existing literature, the following control variables are selected to ensure analytical accuracy and control for potential bias: firm size (Size), ownership concentration (Top5), profitability (ROA), leverage (Lev), years since listing (ListAge), revenue growth rate (Growth), and monetary executive compensation incentive (Pay). In addition, to exclude the confounding effect of climate uncertainty on the results, the Climate Physical Risk Index (CCRI) at the prefecture-level city in China is also included. Detailed definitions of these variables are provided in

Table 2. Definitions of Variables.

Variable Type	Variable Name	Variable Symbol	Variable Definition
Dependent Variable	Firms’ Green Transformation	GTS	Natural logarithm of (1 plus the frequency count of green transformation keywords)
Independent Variable	Climate Policy Uncertainty	CCPU	China Climate Policy Uncertainty Index calculated by Ma et al. (2023) [12]
Control Variables	Firm Size	Size	Natural logarithm of total assets
	Ownership Concentration	Top5	Shareholding ratio of the top five shareholders
	Return on Assets	ROA	Net profit divided by total assets
	Leverage	Lev	Total liabilities divided by total assets
	Years Since Listing	ListAge	Natural logarithm of years since listing
	Revenue Growth Rate	Growth	(Current year operating revenue—Previous year operating revenue)/Previous year operating revenue
	Executive Monetary Compensation Incentive	Pay	Natural logarithm of the total monetary compensation of the top three executives
	Climate Physical Risk Index	CCRI	Prefecture-level city climate risk index based on Guo et al. (2024) [43]

.

4.3.4. Descriptive Statistics

Table 3. Descriptive Statistics.

Variable	N	Mean	Std. Dev.	Min	Median	Max
GTS	29,830	2.2112	1.102	0.00	2.08	6.46
CCPU	29,830	1.7889	0.714	0.03	1.78	6.99
Size	29,830	22.1526	1.309	19.82	21.95	26.31
Top5	29,830	0.5348	0.156	0.19	0.54	0.89
ROA	29,830	0.0415	0.065	−0.22	0.04	0.22
Lev	29,830	0.4125	0.207	0.01	0.40	1.00
ListAge	29,830	2.0200	0.947	0.00	2.20	3.37
Growth	29,830	0.1498	0.349	−0.53	0.10	1.97
Pay	29,830	15.3558	0.736	13.44	15.35	17.26
CCRI	29,830	27.8244	5.695	14.61	27.88	43.88

reports summary statistics for the key variables used in this analysis. The average value of the dependent variable—corporate green transformation—is 2.21, with a standard deviation of 1.082, indicating that most firms exhibit green transformation levels clustered around 2. Nonetheless, variation exists across firms: some demonstrate higher engagement with a maximum value of 6.46, while others show no green transformation efforts (minimum value = 0), revealing disparities in performance. Regarding the primary explanatory variable—climate policy uncertainty—the calculated mean and median are 1.7178 and 0.718, respectively, suggesting moderate distributional stability. The dispersion, as reflected by the standard deviation and the gap between the extreme values, implies potential heterogeneity in perceived policy risks. Summary statistics for the remaining control variables align well with theoretical expectations and prior findings.

5. Empirical Analysis

5.1. Baseline Regression Results

Table 4. Baseline Regression Results.

Variable	(1)	(2)	(3)	(4)
	GTS	GTS	GTS	GTS
CCPU	0.0200 **	0.0156 *	0.0194 *	0.0197 **
	(0.0083)	(0.0084)	(0.0104)	(0.0080)
Size		0.2888 ***	0.3378 ***	0.2901 ***
		(0.0239)	(0.0138)	(0.0248)
Top5		0.3198 ***	0.1378 *	0.3575 ***
		(0.0929)	(0.0803)	(0.0992)
ROA		0.4734 ***	−0.0283	0.4231 ***
		(0.0959)	(0.1242)	(0.0822)
Lev		−0.0736	−0.0976	−0.0564
		(0.0641)	(0.0685)	(0.0643)
ListAge		−0.0116	−0.0074	−0.0092
		(0.0181)	(0.0127)	(0.0183)
Growth		−0.0520 ***	−0.0456 *	−0.0557 ***
		(0.0155)	(0.0242)	(0.0157)
Pay		0.0618 ***	0.1269 ***	0.0644 ***
		(0.0160)	(0.0206)	(0.0155)
CCRI		−0.0005	0.0003	−0.0000
		(0.0013)	(0.0015)	(0.0015)
Constant	2.1754 ***	−5.2797 ***	−7.2742 ***	−5.3973 ***
	(0.0148)	(0.5403)	(0.3077)	(0.5500)
Firm Fixed Effects	No	Yes	No	Yes
Time Fixed Effects	No	Yes	Yes	No
Industry Fixed Effects	No	No	Yes	No
Industry-Year Joint Effects	No	No	No	Yes
N	29,830	29,830	29,830	29,830
R2	0.750	0.747	0.475	0.762

Note: standard error are reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The same notation applies to all subsequent tables.

details the benchmark regression estimates examining how climate policy uncertainty influences corporate green transformation. In model (1), no controls are included for firm-level, temporal, or year-specific effects. Models (2) to (4) build upon model (1), progressively incorporating additional controls. Specifically, model (2) accounts for firm and time fixed effects; model (3) further includes industry-specific fixed effects; and model (4) incorporates joint controls for firm, industry, and time dimensions.

5.2. Robustness Checks

5.2.1. Removing Samples from Pandemic-Affected Years

Since the outbreak of COVID-19 in early 2020, the global economy and various industries have suffered unprecedented shocks. During this period, the business environment became highly complex and volatile, and data from the pandemic may not accurately reflect normal economic patterns. Consequently, directly using such data could bias the results and undermine the generalizability of the findings. Similarly, the global financial crisis of 2007–2009 and the energy crisis of 2022 may also distort research conclusions. To address this concern, following the approach of Zhang (2024) [44], we exclude the periods of the COVID-19 pandemic, the financial crisis, and the energy crisis to avoid potential adverse effects. Columns (1)–(3) of

Table 5. Robustness Analysis.

Variable	Exclude Special Years			Exclude Provincial Capitals	Alternative Model
	(1)	(2)	(3)	(4)	(5)
CCPU	0.0278 ***	0.0173 **	0.0243 **	0.0258 **	0.057 **
	(0.0103)	(0.0086)	(0.0094)	(0.0108)	(0.0277)
Constant	−4.8779 ***	−5.6027 ***	−5.0693 ***	−5.2950 ***	−14.5839 ***
	(0.5388)	(0.665)	(0.5392)	(0.4407)	(1.3426)
Control Variables	Yes	Yes	Yes	Yes	Yes
Firm Fixed Effects	Yes	Yes	Yes	Yes	Yes
Time Fixed Effects	No	No	No	No	No
Industry Fixed Effects	No	No	No	No	No
Industry-Year Joint Effects	Yes	Yes	Yes	Yes	Yes
N	20,506	28,014	26,412	19,439	19,708
R2	0.751	0.768	0.758	0.772
Variable	Alternative Clustering	Alternative Fixed Effects	Alternative Dependent Variable
	(6)	(7)	(8)	(9)	(10)
CCPU	0.0197 *	0.0156 *	0.0194 **	−0.0024 ***	0.2710 ***
	(0.0103)	(0.0084)	(0.008)	(0.0003)	(0.0086)
Constant	−5.3973 ***	−5.2797 ***	−5.3974 ***	0.0564 ***	−0.2143
	(0.3836)	(0.5403)	(0.5491)	(0.0022)	(0.2250)
Control Variables	Yes	Yes	Yes	Yes	Yes
Firm Fixed Effects	Yes	Yes	Yes	Yes	Yes
Time Fixed Effects	No	Yes	No	No	No
Industry Fixed Effects	No	Yes	No	No	No
Industry-Year Joint Effects	Yes	No	Yes	Yes	Yes
N	29,830	29,830	29,784	27,732	29,814
R2	0.762	0.747	0.762	0.929	0.722
Variable	Exclude Climate Policy Effects		Lagged by One Period	Instrumental Variable Method
	(11)	(12)	(13)	(14)	(15)
CCPU	0.0195 **	0.0221 **	0.0220 **	0.621 ***	0.5960 ***
	(0.0081)	(0.0101)	(0.0101)	(0.0149)	(0.2023)
Constant	−5.5447 ***	−4.6177 ***	−4.6339 ***
	(0.3327)	(0.5971)	(0.5702)
Kleibergen–Paap rk LM				20.91
Cragg–Donald Wald F statistic				90.54
Control Variables	Yes	Yes	Yes	Yes	Yes
Firm Fixed Effects	Yes	Yes	Yes	Yes	Yes
Time Fixed Effects	No	No	No	No	No
Industry Fixed Effects	No	No	No	No	No
Industry-Year Joint Effects	Yes	Yes	Yes	Yes	Yes
N	29,830	25,571	25,571	26,440	26,440
R2	0.763	0.771	0.0763

Note: standard error are reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The same notation applies to all subsequent tables.

present the regression results after excluding these events, showing that the coefficient of climate policy uncertainty remains significantly positive. This further confirms that higher climate policy uncertainty promotes enterprises’ green transformation.

5.2.2. Excluding Special Economic Zones (Excluding Provincial Capitals or Special Economic Zones)

Provincial capital cities, as administrative and economic hubs, often benefit from distinctive policy preferences and resource endowments. Consequently, firms situated in these cities are subject to different regulatory environments and development opportunities than those in other urban areas. This unique context may bias empirical findings due to unobservable city-level heterogeneity. To mitigate such concerns, this study follows Chen et al. (2025) [6] and re-estimates the benchmark model after excluding observations from provincial capitals. The corresponding estimation results are reported in column (4) of Table 4. Notably, the coefficient for climate policy uncertainty remains positive and statistically significant at the 5% level, even after excluding these special regions. This finding underscores the consistency and robustness of the positive relationship across heterogeneous urban contexts, further validating the influence of climate policy uncertainty on firms’ green transformation initiatives.

5.2.3. Alternative Econometric Model

Following Lin (2024) [45], a binary indicator variable (GTSDummy) is introduced. Specifically, this dummy equals 1 if a firm’s green transformation score exceeds the median, and 0 otherwise. Based on this definition, a probit model is utilized to re-evaluate the results. Compared to linear regression, the probit specification, which is commonly applied to binary dependent variables, enables a more nuanced estimation of the dummy’s marginal impact on green transformation behavior. As reported in column (5) of Table 5, the estimated coefficient of climate policy uncertainty remains significantly positive under the alternative model. Moreover, the effect appears stronger among firms with higher levels of green transformation, suggesting that these firms are more responsive to shifts in policy-related uncertainty.

5.2.4. Changing the Clustering Level

In empirical analyses, issues like heteroscedasticity and autocorrelation may bias standard errors, compromising result reliability. While baseline regressions cluster at the year and industry level, firms in similar sectors but distinct geographic areas may experience different conditions due to regional variation. To account for this, we re-estimate the model with clustering at the city level for robustness. As presented in column (6) of Table 5, the coefficient on climate policy uncertainty remains significantly positive, confirming that the positive effect of climate policy uncertainty on green transformation persists even after adjusting for city-level heterogeneity and local policy diversity.

5.2.5. Changing Fixed Effects

Considering the influence of dynamic city-level characteristics on corporate green transition, we further introduce time × city interaction fixed effects, drawing on Zhou et al. (2024) [46], to mitigate potential biases. This specification accounts for both spatial and temporal unobservable heterogeneity, thereby improving estimation credibility. As indicated in column (7) of Table 5, the estimated coefficient for climate policy uncertainty remains significantly positive under this specification, further validating that heightened climate policy uncertainty continues to foster green transformation among firms.

5.2.6. Alternative Dependent Variable

To further verify our findings, we adopted the method of Yue et al. (2023) [47], replacing the dependent variables with the logarithm of (1 + the number of green patents) submitted by the enterprise in a given year, the carbon dioxide emissions of the enterprise, and the green background of the enterprise’s senior executives. As shown in columns (8) to (10) of Table 5, the regression results all indicate that climate policy uncertainty is conducive to promoting the green transformation of enterprises.

5.2.7. Excluding Climate Policy Interference

To isolate the effect of climate policy uncertainty, we address potential interference from other concurrent climate policies, such as carbon trading programs and low-carbon city pilot initiatives. We introduce dummy variables based on the list of carbon trading cities (launched in 2013) and the three stages of low-carbon pilots implemented in 2010, 2012, and 2017. The empirical results, as shown in columns (11) and (12) of Table 5, continue to confirm the significantly positive influence of climate policy uncertainty on green transformation, indicating the robustness of our core findings.

5.2.8. The Core Explanatory Variable Lagged by One Period

There may exist a bidirectional causality between climate policy uncertainty and corporate green transformation. To mitigate this concern, this study employs the one-period lag of the core explanatory variable and control variables as robustness test variables and re-estimates the model. As reported in column (13) of Table 5, the coefficient of climate policy uncertainty remains significantly positive, consistent with the baseline regression, thereby confirming the robustness of the main conclusion.

5.2.9. Instrumental Variable Method

To mitigate the concern of reverse causality, this study adopts the Bartik instrumental variable approach to construct a composite instrument. The core logic of the Bartik instrument is to combine external shocks with region-specific share weights in order to identify causal effects and alleviate endogeneity issues. Specifically, we use U.S. climate policy uncertainty as the external shock and the share of each prefecture’s carbon emissions in national emissions in 2010 as the baseline weight [48]. The validity of this instrument rests on two main considerations. (1) Exogeneity of the shock: The U.S. climate policy uncertainty index represents an external shock and has no direct link to Chinese firms’ green transformation, thereby satisfying the exogeneity requirement of instrumental variables. (2) Rationality of the baseline share: The 2010 carbon emission share reflects each region’s relative position in the national emission structure. As it is based on pre-sample data, it is not subject to reverse influence from firms’ decisions during the study period.

Columns (14) and (15) of Table 5 report the two-stage regression results of the instrumental variable estimation. The first-stage results show that the coefficient of China’s climate policy uncertainty is significantly positive, indicating a positive correlation between U.S. climate policy uncertainty and China’s climate policy uncertainty. In the second stage, the coefficient of climate policy uncertainty remains significantly positive. Meanwhile, the Kleibergen–Paap rk LM statistic (20.9) and the Cragg–Donald Wald F statistic (90.54) suggest that the selected instrumental variables do not suffer from identification problems or weak instrument issues. Taken together, these results demonstrate that even after addressing potential endogeneity concerns, climate policy uncertainty continues to promote corporate green transformation.

6. Mechanism Analysis

Existing studies have confirmed that climate policy uncertainty can significantly promote firms’ green transformation, but the specific mechanisms underlying this effect still lack systematic elaboration. To fill this gap, it is necessary to conduct an in-depth analysis of its potential pathways.

First, a higher level of corporate environmental attention not only drives firms to attach greater importance to ecological issues but also significantly enhances their sensitivity to climate risks and policy signals. This enables them to identify potential policy trends earlier and adjust their business strategies in advance. More importantly, environmental attention strengthens firms’ alignment with key stakeholders such as investors, consumers, and governments: investors are more willing to allocate capital, consumers are more inclined to purchase green products, and governments may provide preferential support through subsidies and policy-based financing. Over time, sustained environmental attention often evolves into a strategic commitment, embedding green concepts institutionally into all aspects of production and management, thereby fundamentally driving green transformation.

Second, from an opportunity-oriented perspective, high uncertainty is often perceived by firms as external pressure for innovation and renewal. Firms tend to increase their R&D investment to cope with the possibility of stricter future environmental constraints. Such R&D activities not only facilitate the development of green technologies and products, thereby reducing compliance risks, but also help accumulate green knowledge and technological reserves, improving production efficiency and green supply capacity. More critically, firms that take the lead in expanding green R&D often capture emerging markets, establish technological barriers, and secure first-mover advantages, thereby reaping higher returns in the process of green transformation. In contrast, insufficient R&D may result in marginalization in the green competition.

Third, proactive green strategies provide firms with a sense of direction and institutional safeguards under uncertainty. On the one hand, they help firms avoid strategic drift when setting environmental performance goals, ensuring that limited capital, labor, and technological resources are prioritized for green domains. On the other hand, strategic positioning enhances firms’ institutional legitimacy in policy bargaining and regulatory processes, strengthening their advantages in competing for green subsidies, green credit, and broader social recognition. Without such institutionalized green strategies, firms may not only fall into passivity in resource allocation but also be disadvantaged in terms of policy support and market acceptance.

Taken together, corporate environmental attention provides the cognitive basis and external support, R&D investment establishes the technological and market foundation, and green strategic positioning ensures institutionalized resource allocation and long-term commitment.

These three elements are interconnected and mutually reinforcing, jointly forming the mechanism chain through which climate policy uncertainty promotes firms’ green transformation. This offers a more systematic and comprehensive theoretical explanation of corporate green transformation behavior in the context of uncertainty. For corporate environmental attention, following Li (2025) [42], this study uses the frequency of environment-related keywords in annual reports of listed firms as a proxy indicator. Following the research methods of Zhang (2019) and Fu (2024) [49,50], this study uses two proxy indicators to measure firms’ green investment: the ratio of annual green investment expenditure to year-end total assets, and the ratio of annual green investment expenditure to operating revenue. These indicators can effectively capture the level of firms’ investment in green technological innovation. Green strategic orientation is assessed using proxies such as disclosure of environmental management systems and ISO14000 certification, based on the method of Kweh et al. (2024) [51].

The empirical results are shown in

Table 6. Mechanism Analysis.

Variable	Environmental Attention	Enterprise Green Investment		Corporate Green Strategic Orientation
	(1)	(2)	(3)	(4)
CCPU	0.0017 **	0.1633 **	0.1138 ***	0.0238 *
	(0.0008)	(0.0800)	(0.0333)	(0.0129)
Constant	0.2549	−18.0398 ***	−2.6106 ***	−7.0631 ***
	(0.0248)	(3.8818)	(0.9240)	(0.4673)
Control Variables	Yes	Yes	Yes	Yes
Firm Fixed Effects	Yes	Yes	Yes	Yes
Industry-Year Joint Effects	Yes	Yes	Yes	Yes
N	24,156	21,748	21,748	29,830
R2	0.653	0.789	0.825	0.672

Note: standard error are reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The same notation applies to all subsequent tables.

. Column (1) indicates that climate policy uncertainty significantly increases corporate environmental attention at the 10% level, suggesting that firms facing greater policy uncertainty are more likely to emphasize environmental issues and pursue green transformation. Columns (2) and (3) demonstrate that climate policy uncertainty also positively influences R&D investment, with coefficients statistically significant at the 10% level. This implies that uncertainty stimulates firms to raise R&D input, thereby accelerating green technology development. Column (4) reveals a significantly positive impact of climate policy uncertainty on green strategic orientation, indicating that uncertainty leads firms to clarify development goals, improve resource allocation, and ultimately promote green transformation.

7. Heterogeneity Analysis

The previous findings indicate that climate policy uncertainty can effectively stimulate firms’ green transformation. Nonetheless, firms may exhibit varying responses to such uncertainty depending on their specific characteristics. These variations are primarily shaped by firm-level factors such as ownership type, financial constraints, and exposure to climate risk. Ownership structure influences how firms respond to policy signals and the intensity of their behavioral adjustments. For instance, due to differing resource endowments and policy contexts, state-owned and non-state-owned firms may pursue distinct strategies. The characteristics of different industries determine the extent to which firms can access and allocate resources needed for green transformation. Moreover, firms exposed to higher levels of climate risk often exhibit different resource allocation and risk management behavior. Given these differences, this study further examines the heterogeneous effects of climate policy uncertainty on firms’ green transformation across ownership structures, Industry characteristics, and climate risk levels. It is worth noting that, for some firms, information on ownership type is missing in certain years. The classification of industry attributes relies on industry codes and related data matching, which excludes some ST firms or firms in special industries. In addition, for some firms, missing annual reports or unavailable text processing results prevent the construction of the index. Consequently, when conducting subgroup regressions, it is unavoidable to exclude these missing observations, leading to a sample size smaller than 29,830.

7.1. Ownership Structure

Following Zhao et al. (2024) [52], we divide the sample into state-owned enterprises (SOEs) and non-state-owned enterprises (non-SOEs) for subgroup analysis. Regression results in columns (1) and (2) of

Table 7. Heterogeneity Analysis.

Variable	State-Owned Enterprises	Non-State-Owned Enterprises	Technology-Intensive	Asset-Intensive	Labor-Intensive
	(1)	(2)	(3)	(4)	(5)
CCPU	0.0534 ***	0.0007	0.0156	−0.0300 *	0.0507 ***
	(0.0142)	(0.0092)	(0.0120)	(0.0157)	(0.0139)
Constant	−5.4105 ***	−5.4514 ***	−5.5403 ***	−6.0704 ***	−4.9434 ***
	(0.4762)	(0.4198)	(0.5596)	(0.7394)	(0.5238)
Control Variables	Yes	Yes	Yes	Yes	Yes
Firm Fixed Effects	Yes	Yes	Yes	Yes	Yes
Industry-Year Joint Effects	Yes	Yes	Yes	Yes	Yes
N	11066	18576	14728	5004	9722
R2	0.7898	0.7520	0.745	0.786	0.779
Variable	Low Climate Risk	High Climate Risk
	(6)	(7)
CCPU	0.0087	0.0217 **
	(0.0130)	(0.0104)
Constant	−3.4050 ***	−6.3147 ***
	(0.4045)	(0.5118)
Control Variables	Yes	Yes
Firm Fixed Effects	Yes	Yes
Industry-Year Joint Effects	Yes	Yes
N	14536	14540
R2	0.702	0.789

Note: standard error are reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The same notation applies to all subsequent tables.

reveal that the coefficient of climate policy uncertainty on green transformation is significantly positive at the 1% level for SOEs but not statistically significant for non-SOEs. This indicates that climate policy uncertainty facilitates green transformation in SOEs, while its influence on non-SOEs is negligible.

A potential explanation lies in institutional differences. First, SOEs and non-SOEs vary in access to policy resources and support. SOEs generally maintain closer ties with government institutions, enabling them to obtain financial assistance and policy backing more easily. Under high climate policy uncertainty, SOEs are better positioned to rely on public resources to advance green transformation [44]. Second, there are differences in decision-making mechanisms. Decision-making in SOEs is more likely to be guided by the government, allowing them to respond more quickly to government initiatives and adjust their green transformation strategies accordingly. In contrast, non-SOEs rely more on market mechanisms and, when facing uncertainty, may be more inclined to avoid risks and adopt a more conservative approach to transformation. Finally, the two types of enterprises differ in their risk-bearing capacities. SOEs, backed by government support, can access more funding and policy guarantees, and thus possess stronger risk resilience. Non-SOEs, with relatively limited funding sources, tend to be more cautious in responding to policy changes and may adopt more conservative strategies in green transformation.

We further examine the distribution of enterprises by ownership type and factor intensity, as reported in

Table 8. Distribution of Enterprises by Ownership Type and Factor Intensity.

	Technology-Intensive	Asset-Intensive	Labor-Intensive	Total
State-Owned Enterprises	3866	2646	4554	11,066
Non-State-Owned Enterprises	10,879	2370	5222	18,471
total	14,745	5016	9776	29,537

. The overall sample comprises 29,537 firms, among which non-state-owned enterprises substantially outnumber state-owned enterprises, with 18,471 and 11,066 firms, respectively. In terms of industry composition, technology-intensive enterprises account for the largest share (14,745), followed by labor-intensive enterprises (9776), while asset-intensive enterprises represent the smallest group (5016). A closer comparison reveals that non-state-owned enterprises dominate in technology-intensive industries (10,879), whereas state-owned enterprises hold a relatively larger share in labor-intensive industries. This discrepancy underscores the pronounced heterogeneity in factor endowments and industrial positioning between enterprises with different ownership structures.

7.2. Industry Heterogeneity

Following Zhang et al. (2024) [44], firms are categorized based on industry characteristics into technology-intensive, asset-intensive, and labor-intensive groups. The results, presented in columns (3) to (5) of Table 7, show that the impact of climate policy uncertainty on green transformation varies across different types of firms: it has a promoting effect on labor-intensive firms, a suppressing effect on asset-intensive firms, and an insignificant effect on technology-intensive firms. The possible reasons are as follows: Labor-intensive enterprises operate in highly competitive markets with thin profit margins, making them highly sensitive to policy signals. When uncertainty rises, they tend to pursue green transformation through low-cost and quick-return measures such as process optimization or equipment upgrades, in order to reduce compliance risks and secure external resource support. This aligns with the logic of rapid response under resource constraints emphasized by uncertainty theory. In contrast, asset-intensive enterprises are constrained by high sunk costs and long investment payback periods. Under policy ambiguity, the expected returns on green investments decline, leading to significantly weaker transformation incentives, which reflects the restrictive effect of “resource rigidity” highlighted in the resource-based view.

7.3. Corporate Climate Risk

Following Hussain et al. (2023) [53], we construct a climate risk index through a combination of text analysis and machine learning methods. Based on the median value of this index, firms are categorized into high-risk and low-risk groups for regression estimation. As reported in columns (5) and (6) of Table 7, climate policy uncertainty significantly promotes green transformation only for firms with high exposure to climate risk. For firms with low climate risk, no statistically significant effect is observed. This heterogeneity can be explained from the perspectives of risk-driven factors and policy sensitivity. First, firms facing high climate risk, due to long-term exposure to environmental shocks, pay closer attention to changes in the external environment and the uncertainty of policy fluctuations. As a result, they tend to adopt forward-looking green transformation measures to enhance adaptability and reduce potential future losses. Second, such firms typically exhibit higher policy sensitivity, especially in industries that rely heavily on natural resources or are closely linked to climate change, making them more directly affected by policy adjustments [54]. Consequently, they are more proactive in promoting green innovation and transformation under uncertain conditions. In contrast, firms with low climate risk, facing limited external pressure, lack the same degree of risk-driven motivation and institutional incentives, and thus demonstrate relatively weaker momentum for green transformation.

8. Conclusions and Implications

In light of the evolving global climate policy environment, uncertainty surrounding climate-related regulations has emerged as a crucial external factor affecting firms in the real economy. Motivated by this, the present study investigates how climate policy uncertainty influences corporate green transformation and explores the channels through which this effect materializes. Utilizing panel data of A-share firms listed on the Shanghai and Shenzhen Stock Exchanges from 2008 to 2023, we provide empirical evidence on the role of climate policy uncertainty in shaping green strategic behavior. Our main findings are as follows: (1) Climate policy uncertainty significantly stimulates firms’ green transformation. This result remains robust across various specifications, including alternative clustering approaches, exclusion of abnormal years, and fixed effects adjustments. (2) Mechanism analysis identifies three pathways through which climate policy uncertainty fosters green transition: it raises firms’ environmental awareness, promotes green investment, and promotes long-term green strategic orientation. (3) The effect of policy uncertainty exhibits heterogeneity across firm types. The positive influence is more evident among state-owned firms, labor-intensive firms, and those with high climate risk exposure, which are typically more adaptive to policy shifts and better equipped in terms of resources and financial flexibility. These firms respond more actively to climate policy uncertainty by accelerating their green transformation processes as a strategy to manage future risks.

This study contributes to the literature by offering empirical support for the hypothesis that climate policy uncertainty can serve as a driver of green transformation. Based on these insights, several policy implications are put forward in the subsequent section:

(1)

The government should take effective measures to promote corporate green transformation. Local governments are not only implementers of climate policy but also guides for corporate green development. In the face of climate policy uncertainty, governments should further strengthen their role in climate governance. First, they should enhance environmental supervision, set stricter environmental standards and assessment mechanisms, and encourage firms to proactively undertake green transformation initiatives. Second, governments should optimize green subsidy policies, increase fiscal support for green technology R&D and clean energy projects, and ensure that subsidies effectively promote green technological innovation and transformation. In addition, tax incentives and green credit support should be used to reduce the costs of green transformation for firms, thereby encouraging greater R&D investment and enabling them to meet the challenges posed by climate policy uncertainty through technological innovation.

(2)

Improve climate policy communication mechanisms to enhance policy transparency and expectation management. A significant part of climate policy uncertainty stems from information asymmetry and insufficient communication during policy formulation and implementation. It is recommended that government departments establish regular and systematic policy communication mechanisms to promptly disclose policy rationales, implementation progress, and future plans to firms, the public, and financial institutions, thereby reducing uncertainty caused by information asymmetry. Considerations may include setting up dedicated policy interpretation sections, industry dialogue platforms, or holding enterprise forums to collect feedback and suggestions from firms regarding climate policies. These measures can improve policy transparency and predictability, guide firms in forming reasonable expectations, alleviate their “wait-and-see” attitudes, and reduce the disruptive impact of policy fluctuations on corporate green transformation.

(3)

The government should formulate differentiated policies based on local conditions. Heterogeneity analysis shows that the impact of climate policy uncertainty varies among different types of firms; thus, governments should provide customized, targeted policy support. Specifically, more policy resources should be directed toward non-state-owned enterprises to reduce the policy treatment gap between ownership types. For asset-intensive firms, targeted support measures should be provided to broaden their financing channels, for example, through low-interest loans and green bonds, to help these firms better respond to policy requirements and achieve green transformation. For firms facing high climate risk, governments should strengthen policy guidance, encourage enhanced climate risk management, and increase support for investment in green technologies to help them better address the challenges brought by climate change.