Impact of Climate Policy Uncertainty on Regional New Quality Productive Forces in China

Abstract

In the context of China’s strategic push toward high-quality development, the concept of new quality productive forces (NQPF)—which emphasizes technological innovation, green transformation, and digital upgrading—has received a lot of attention. However, the increasing volatility and ambiguity in climate-related policymaking present a serious institutional challenge. This study addresses the underexplored question of how climate policy uncertainty (CPU) affects the regional development of NQPF in China. Unlike traditional productivity, NQPF relies on long-term innovation and sustainable investment, which are highly sensitive to external policy signals. Drawing on panel data from 30 Chinese provinces between 2013 and 2021, this paper uses fixed-effects regressions to empirically assess the influence of CPU on NQPF. The findings reveal that CPU significantly suppresses the development of NQPF, but this effect is mitigated by financial inclusion, carbon market participation, and strong local government sustainability performance. This paper provides new insight into the risks posed by climate uncertainty to economic development and highlights institutional tools that can buffer its negative effects.

1. Introduction

In September 2023, China introduced the concept of new quality productive forces (NQPF). With the continuous promotion of the innovation-driven strategy, a growth model that emphasizes quality and efficiency has begun to become the new theme of China’s economic development [1]. Against this backdrop, the concept of new quality productive forces has emerged and has received a lot of attention [2]. NQPF is a new form of productivity based on scientific and technological progress, knowledge precipitation, and management optimization, which is capable of profoundly changing the economic structure and mode of production [3]. Compared with traditional productivity, NQPF is more innovation-driven, emphasizing the application of new technologies, the introduction of new thinking, and the practice of new modes to improve production efficiency and resource allocation effects, thereby promoting overall social progress and economic development [4]. As an important driving force for high-quality economic growth, NQPF plays a key role in industrial transformation and upgrading [5]. The enhancement of NQPF has enabled society to shift from a crude growth model based on resource consumption to an intensive and innovative growth path that relies on efficiency improvement and technological innovation [6]. This has not only enhanced the country’s economic competitiveness but also significantly improved the quality of life of the population. In particular, against the backdrop of global challenges such as environmental pollution and shortage of energy resources, NQPF provides a reliable path to sustainable development. By promoting the efficient and recycled use of resources, NQPF promotes green development while reducing environmental pressures [7]. Such development will continue to lay a solid foundation for sustainable and high-quality economic development.

Climate change is one of the most serious and urgent global challenges of the twenty-first century and has received great attention from the international community [8]. In recent years, climate problems such as global warming, frequent occurrences of extreme weather, and rising sea levels have intensified, making combating climate change a common task for all countries [9]. In order to reduce carbon emissions and achieve a green transition, governments around the world are gradually realizing the urgency of adopting effective climate policies. The Chinese government has formulated a series of climate policies in response to this trend, the most representative of which include the establishment of “carbon peak” and “carbon neutral” targets and the promotion of a carbon emissions trading system. These policies aim to gradually promote industrial restructuring, energy structure optimization, environmental protection, and sustainable development [10]. However, the locally tailored policy implementation strategies of governments at all levels also make the policy impacts show significant differences among different regions [11]. The formulation and implementation of climate policies are affected by a variety of factors, including the diversity of policy objectives, the game of stakeholders, and changes in the international market environment, leading to some uncertainty in the continuity and consistency of policies [12]. In addition, regions show significant implementation differences in the process of policy implementation due to different resource conditions, economic development levels, and industrial structures, further exacerbating policy uncertainty. The dynamic nature of policy adjustment makes it difficult for enterprises to predict the direction of future policies and challenging to plan investment and business decisions [13]. This policy uncertainty has become one of the most important risk factors affecting economic activities, not only weakening the implementation effect of government policies but also profoundly affecting the business strategies and market behaviors of enterprises [14]. For businesses, climate policy uncertainty increases the difficulty and cost of environmental compliance, affecting their confidence to invest and their willingness to innovate [15]. For the regional economy, such uncertainty has a complex impact on economic dynamism, the business environment, and the innovative capacity of regions, which in turn creates a series of knock-on effects that affect the long-term development and productivity of the region.

While studies have examined the impact of CPU on economic growth, investment decisions, and firm behavior [16,17,18,19], there is no literature on how it affects new quality productive forces. As an important indicator of high-quality economic growth and sustainable development, new quality productive forces are affected by a combination of factors, including policies, innovation resources, and the regional economic environment. However, the effect of policy uncertainty on new, quality productive forces has not been examined. Exploring the specific impacts of climate policy uncertainty on regional new quality productive forces will not only contribute to a more comprehensive understanding of the multilevel effects of policy changes on economic development but can also provide a scientific basis for the formulation of more effective regional policies, which will in turn promote the development of regional economies in a greener and more sustainable direction.

While climate policies aim to guide long-term environmental and economic transitions, their uncertainty can introduce significant ambiguity to economic development. This ambiguity may reduce the willingness of economic agents to invest in innovation and long-term green initiatives, thus potentially hindering the development of NQPF.

Drawing on panel data from 30 Chinese provinces between 2013 and 2021, this paper uses fixed-effects regressions to empirically assess the influence of CPU on NQPF. In doing so, this paper contributes to the growing literature on environmental institutional quality and green development by bridging the gap between climate policy volatility and regional productivity upgrading. The remainder of the paper is structured as follows. Section 2 reviews the theoretical mechanisms linking CPU to NQPF and presents testable hypotheses. Section 3 describes the data sources, variables, and empirical strategy. Section 4 reports and discusses the main regression results. Section 5 concludes and offers policy recommendations.

2. Literature Review and Hypothesis Development

2.1. New Quality Productive Forces

NQPF is a concept relative to traditional productivity [20]. Wang [21] points out that NQPF is productivity driven by the deepening application of new technologies and characterized by the rapid emergence of new industries, new forms of business, and new models, and compared to traditional productivity, NQPF is higher in technology, better in quality, more efficient, and more sustainable. The concept of NQPF stems from changes in the stage of economic development, technological change, and resource and environmental pressures that make it difficult for traditional productivity theories to fully explain modern economic phenomena [22]. Ren et al. [23] pointed out that with the deepening of globalization and the wide application of information technology, especially the development of emerging technologies such as the Internet, artificial intelligence, big data, biotechnology, clean energy, etc., the composition and driving force of productivity have undergone qualitative changes, and the traditional mode of production that relies on resource expansion is no longer applicable. Lu et al. [24] pointed out that the new quality of productivity emphasizes the achievement of high-quality economic development through scientific and technological innovation and the accumulation of knowledge, rather than relying solely on the input of material capital. Cao and Dai [25] define NQPF as the additional increase in green total factor productivity (GTFP) beyond the original growth trajectory. Xie et al. [26] argue that science and technology are core elements of new productivity, which is essentially green productivity.

At present, there is no unified and authoritative measurement system for NQPF. Scholars from different research perspectives have proposed various measurement methods and evaluation frameworks. Based on existing research, the main measurement approaches can be categorized as follows.

Some scholars construct an evaluation system for NQPF from a comprehensive perspective that integrates technological productivity, green productivity, and digital productivity [27,28,29]. Additionally, some scholars have constructed a measurement framework for NQPF from the perspective of the disruptive innovation ecosystem [30]. This framework posits that the formation of NQPF relies on technological revolutions driving profound transformations in production methods and industrial structures. Therefore, the measurement system should comprehensively cover breakthroughs in fundamental research, technological innovation and transformation, industrial upgrading, and the interaction between technology and industry. Some other scholars adopt a Marxist labor productivity theory-based approach, constructing a measurement system for NQPF based on three dimensions: labor force, labor objects, and labor means [22,31].

There is already a body of literature that finds the development of NQPF to be critical to the socio-economic system in many ways. Xu et al. [32] emphasized that the NQPF significantly promotes green development. Liu [33] highlighted that the characteristics of the NQPF—such as digitalization, networking, intelligence, informatization, and intensification—are not only direct reflections of technological advancements and productivity growth but are also key drivers of industrial transformation and sustainable future development. Zhang et al. [34], by analyzing data from 30 provinces across China spanning from 2012 to 2021, affirmed that the NQPF is fundamental for realizing high-quality development in the manufacturing sector, particularly by improving technological standards and efficiency. Yang [35] suggested that the introduction of the NQPF serves as an important catalyst for advancing China’s scientific and technological landscape, laying the foundation for future innovation. Similarly, Shao et al. [29] utilized data from 30 provincial administrative regions between 2011 and 2021 and found that the implementation of the NQPF significantly reduces disparities in industrial structure, contributing to regional economic balancing.

Current research primarily focuses on the impacts of NQPF on economic and social development, yet there is a notable lack of systematic investigation into the constraints on its growth. In particular, climate policy uncertainty poses significant challenges, as fluctuating regulatory environments and inconsistent policy signals create uncertainty in investment and innovation decisions. This uncertainty hinders the effective allocation of resources and disrupts the long-term strategic planning necessary for the development of NQPF. Therefore, despite the attention paid to new and qualitative productivity as an important driver of high-quality economic development, many of its theoretical and practical issues need to be explored in greater depth.

2.2. Research Related to Climate Policy Uncertainty (CPU)

Climate policy uncertainty (CPU) refers to the inability of economic agents to accurately predict if, when, and how governments will change current climate policies [36]. In recent years, the CPU has gradually become a focus of attention for academics and policymakers. In the existing literature, scholars have explored climate policy uncertainty and its impacts on the economy, investment, innovation, the energy mix, and financial markets from a variety of perspectives. Liang et al. [12] found that CPU has a significant negative effect on the long-term volatility of renewable energy, suggesting that uncertain policies can destabilize the growth trajectory of this sector. Similarly, Dai and Zhang [17] analyzed data from 210 commercial banks in China between 2009 and 2020, concluding that CPU significantly reduces both passive and active risks for banks, while paradoxically increasing their insolvency risk. This indicates that while certain types of risks are mitigated, the uncertainty surrounding policies may create vulnerabilities for banks’ financial stability. Fried et al. [37] argued that climate policy risks tend to reduce firms’ capital stocks, while also pushing them towards relatively cleaner operations, ultimately contributing to lower carbon emissions. Fuss et al. [38] emphasized that investors are exposed not only to the uncertainty caused by traditional market price fluctuations but also to the unpredictable consequences of increasingly stringent climate change policies, which present unique and substantial challenges. The sensitivity of capital markets to policy uncertainty can also be reflected in share price volatility. Bouri et al. [39] identify CPU as a key determinant of stock market performance, indicating that policy uncertainty around climate issues can significantly influence stock returns and investor sentiment. Similarly, Chen et al. [40] find that CPU plays an important role in stock price volatility, and they highlight that accounting for CPU improves the accuracy of volatility predictions, underlining the importance of incorporating policy factors into financial forecasting models. In response to the challenges posed by climate policy uncertainty, researchers have also explored various mitigation strategies. For instance, Azimli and Cek [41] argued that investments in environmental, social, and governance (ESG) initiatives can help firms build reputational capital, which in turn mitigates the negative effects of economic policy uncertainty on firm valuation. Wellman [42] also found that firms with established political connections are better positioned to obtain relevant and timely information regarding legislative changes, thus reducing the adverse effects of policy uncertainty. Similarly, Wang et al. [43] reported that increased transparency and disclosure can effectively mitigate the negative impacts of policy uncertainty on stock liquidity, suggesting that greater openness helps investors manage their risk more effectively.

To date, there is a noticeable gap in the existing literature regarding the relationship between CPU and new quality productive forces. While numerous studies have examined the impact of CPU on traditional economic indicators such as investment, innovation, financial markets, and energy consumption, little attention has been paid to its potential influence on the development and transformation of new quality productive forces—a key driver of high-quality, sustainable economic growth in the current era. As this concept becomes increasingly central to national development strategies, particularly in the context of green and innovation-led growth, understanding how CPU may affect its formation, efficiency, and resilience is both timely and necessary. However, no systematic academic research has yet addressed this issue, leaving an important area underexplored.

2.3. Hypothesis Development

From the perspective of the formulation of the category of NQPF, its essence is still productivity. NQPF is one in which innovation plays a leading role, and which is free from the traditional mode of economic growth and path of productivity development, and which is characterized by high technology, high efficiency, and high quality [21]. The impact of climate policy uncertainty on regional new and qualitative productivity can work through a number of mechanisms, including innovation-driven, resource allocation and etc.

First, from an innovation-driven perspective, NQPF relies on continued technological innovation and green R&D investment by firms. However, climate policy uncertainty may discourage firms from investing in this area [21]. This is because investment in innovation is usually subject to long lead times and a high risk of stranded assets [44]. When there are frequent changes in climate policy or a lack of clear direction, firms find it difficult to predict future environmental standards and market prospects and are thus wary of investing in innovation. In addition, many climate policies directly target green innovation activities of firms by providing them with policy incentives or financing facilities. However, when climate policies become uncertain, managers are concerned that green R&D activities will not receive the expected policy support, leading to a reduction in the potential benefits of green innovation. Technological progress is an important source of new qualitative productivity gains, and climate policy uncertainty hinders the development and application of new technologies and innovation-driven development strategies. This innovation crowding-out effect is directly detrimental to regional NQPF growth.

Second, when there is uncertainty in the external environment, the way firms allocate resources changes [45]. Compared with innovative investments, investments in financial assets are often seen as an ideal choice for avoiding uncertainty shocks because of their higher liquidity and lower policy relevance. Financial assets have a reservoir function and can better meet the liquidity needs of firms in times of increased external uncertainty. In addition, financial investment also has certain speculative arbitrage qualities, and driven by profit-seeking motives, managers tend to increase investment in financial assets when climate policy uncertainty intensifies, in order to maximize the short-term benefits of enterprises. Enterprises devoting excessive resources to financial investment may distort the normal production and operation activities of enterprises [46]. This could crowd out resources for new models, new business models, and other areas of new productivity, thereby negatively affecting new productivity.

An important process in the development of new quality productive forces is the optimization of industrial structure, i.e., the elimination of inefficient and highly polluting industries and the development of high-tech and green industries [47]. However, climate policy uncertainty may hinder industrial restructuring, affecting the cultivation of new industries and the transformation of traditional industries. Backward production capacity may be difficult to exit due to policy ambiguity. Inefficient industries should be gradually withdrawn under technological upgrading and market competition, but if the climate policy is unstable, some high-polluting and high-emission enterprises may take the opportunity to maintain their operations, postpone their green transformation, and continue to rely on traditional modes of production, which will squeeze the space for the development of high-value-added industries, and affect the development of new-quality productive forces.

In terms of capital flows, financial capital typically seeks higher returns and reallocates across industries and regions [48]. Compared with corporate investment decisions, financial capital is more liquid and is more significantly influenced by market expectations and policy changes. Capital usually seeks higher returns and flows from traditional industries to high-technology industries [49]. This flow promotes the research and development and application of new technologies, accelerates the implementation of innovation-driven development strategies, and raises the overall level of new quality productivity. When policies are unstable, investors in the market tend to overreact, leading to increased volatility in the capital market. Regions with higher stability of climate policy policies tend to attract capital more easily, while regions with higher volatility of climate policy implementation may face the risk of capital outflow. Changes in the allocation of capital due to climate policy uncertainty can hinder the level of regional new quality productivity.

Based on the above theoretical analyses, this paper proposes the following research hypotheses:

Hypothesis 1.

Climate policy uncertainty has a significant negative effect on regional NQPF.

Inclusive finance refers to efforts to make financial services accessible and affordable to all individuals and businesses, regardless of their personal net worth or company size [50]. China’s financial system is still very poorly developed, which has severely inhibited sustainable economic development [51]. Inclusive finance can make up for the shortcomings of traditional finance and enable underdeveloped regions to enjoy convenient financial services [52]. Inclusive finance may play an important role in mitigating the dampening effect of climate policy uncertainty on NQPF. First, inclusive finance improves access to finance for enterprises by expanding the coverage of financial services, especially support for small and medium-sized enterprises and innovative enterprises, enabling them to obtain the necessary financial support in an uncertain policy environment. Secondly, by providing favorable lending rates and designing green financial products, inclusive finance reduces the cost of financing and capital risk for enterprises and enhances their willingness to invest in innovation and high-quality production. In addition, the diversified funding sources of inclusive finance make enterprises more resilient and can effectively diversify the financial risks associated with policy uncertainty, thus guaranteeing the continued improvement of their NQPF. This paper proposes Hypothesis 2.

Hypothesis 2.

Inclusive finance reduces the negative impact of CPU on NQPF.

The carbon market provides a clear carbon pricing mechanism, which reduces firms’ uncertainty about future policies and costs and allows them to better plan their investment and operational decisions, especially in environmental innovation [53]. The carbon market offers direct financial incentives that encourage firms to adopt energy-efficient and emission-reduction technologies. For example, companies that successfully lower their emissions can sell excess carbon allowances to other firms, generating additional revenue while simultaneously contributing to overall emission reduction goals. This creates a market-driven motivation for businesses to transition toward low-carbon operations. Through the carbon market, firms are able to obtain financial incentives, which directly incentivize firms to invest more in energy-efficient and emission-reduction technologies, thus mitigating the negative impacts of policy uncertainty on innovation. In addition, the flexibility of the carbon market allows firms to achieve their emission reduction targets through different pathways, thereby effectively managing compliance costs and reducing the impact of policy changes. The transparency of the carbon market also raises firms’ expectations for the future, increasing their confidence to make long-term investments. In these ways, the opening of carbon markets provides a more stable and predictable environment for firms to cope with climate policy uncertainty through economic incentives and transparent rules, thereby effectively reducing the dampening effect of such uncertainty on NQPF.

Hypothesis 3.

Opening the carbon market reduces the negative impact of CPU on NQPF.

3. Research Methodology

3.1. Key Variable Definitions

Ma et al. [54] used the MacBERT deep learning model to quantify climate policy-related uncertainty by extracting relevant information from news texts using text analytics to calculate a macro-level climate policy uncertainty index for China. In this paper, we use the provincial-level climate policy uncertainty index for China published by Ma et al. [54].

New quality productive forces: In the existing literature, new quality productivity is widely understood as a novel type of productivity driven by technological innovation that enhances and transforms traditional production processes. Building on this concept and referring to the practice of Zhao et al. [27], Yu and Zhang [28], Peng and Mariadas [55] and Zhang and Pu [56], this study constructs an evaluation index system for digital new quality productivity and uses the entropy weighting method to measure its status. The detailed results are presented in

Table 1. New quality productive forces tectonic indicators.

Level 1	Level 2	Serial Number	Level 3	Description	Attribute
New quality productive forces	Technological productive forces	A1	Innovation R&D	Number of domestic patents granted	+
		A2	Innovative Industry	High-tech industry business revenue	+
		A3	Innovative Products	Industrial innovation expenditure by large enterprises	+
		A4	Technology Development	Full-time equivalent R&D personnel in large enterprises	+
	Green productive forces	B1	Energy consumption Intensity	Energy consumption per GDP	−
		B2	Industrial water use intensity	Industrial water usage per GDP	−
		B3	Waste Utilization	Ratio of industrial solid waste used to generate	+
		B4	Wastewater Discharge	Utilized amount/generated amount of industrial solid waste	−
		B5	Waste Gas Emission	Industrial SO2 emissions per GDP	−
	Digital productive forces	C1	Electronic Information Manufacturing	Integrated circuit production	+
		C2	Telecommunications	Total telecommunications business volume	+
		C3	Internet Penetration	Number of broadband internet access ports	+
		C4	Software Services	Software business revenue	+
		C5	Digital Information	Length of optical cable lines per regional area	+
		C6	E-commerce	E-commerce sales volume	+

.

3.2. Model Construction

The baseline regression model of this paper is set as follows:

NQPF_it = β₀ + β₁CPU_it + γControl_it + dyear + dprovince + e_it

(1)

where i and t represent province and year, respectively. NQPF_it measures the level of new quality productive forces in year t. CPU represents climate policy uncertainty. Control_it represents control variables. In addition, the model controls for province fixed effects dprovince and year fixed effects dyear, and e_it is a random error term. Referring to the practice of Zhang and Jia [57], Zhao et al. [27], Cheng et al. [47] and Xie et al. [58], a range of provincial control variables has also been selected for this paper, as shown as follows.

These variables are (1) Financial development level (Fdl), measured by the ratio of gross loans to financial institutions over GDP; (2) Foreign direct investment (FDI), measured as the ratio of FDI to GDP; (3) Higher education level (Educa), measured by the logarithm of the number of students enrolled in higher education institutions; (4) Science and technology expenditure (Ste), captured by the ratio of local fiscal expenditure on science and technology to GDP; (5) Fixed-asset investment (Fai), measured as the logarithm of total investment in fixed assets; and (6) Share of secondary industry (Seci), proxied by the ratio of output from the secondary sector to GDP.

Additionally, the model incorporates province fixed effects (dprovince) to control for time-invariant unobservable provincial characteristics, and year fixed effects (dyear) to control for common shocks and macroeconomic changes over time. A random error term (ε) is also included to capture unobserved stochastic disturbances.

Climate policy uncertainty may dampen the development of new quality productivity. In order to understand more fully the possible ways to mitigate this dampening effect, this paper further adapts Equation (1) to obtain Equations (2) and (3):

NQPF_it = β₀ + β₁CPU_it + β₂IF_it + β₃CPU_it × IF_it + γControl_it + dyear + dprovince + e_it

(2)

NQPF_it = β₀ + β₁CPU_it + β₂CM_it + β₃CPU_it × CM_it + γControl_it + dyear + dprovince + e_it

(3)

where IF represents inclusive finance, and CM represents whether or not to open up the carbon market.

3.3. Data Selection and Variable Summary

To conduct an empirical investigation on the regional effects of climate policy uncertainty, this study constructs a balanced panel dataset covering 30 provinces in mainland China over the period from 2013 to 2021. Tibet is deliberately excluded due to its geographical isolation and socio-economic structure differences, which would otherwise impair the comparability and reliability of cross-regional analysis.

The reasons for choosing 2013 to 2021 as the study period are as follows. First, the year 2013 coincides with the official rollout of China’s innovation-driven development strategy, a national agenda that places innovation at the heart of economic transformation—an ideal policy backdrop for studying new quality productive forces. Second, due to the limited availability and delayed release of several provincial-level indicators, this study uses data up to the year 2021.

All data used in this research are Chinese provincial data. The index measuring climate policy uncertainty is sourced from the work of Ma et al. [54]. Other indicators are obtained from the National Bureau of Statistics of China, China Statistical Yearbook, China Environmental Statistics Yearbook, and China Energy Statistics Yearbook.

Table 2. Variables summary.

Variable	Symbol	Obs	Mean	Median	Std. Dev.
New quality productive forces	NQPF	270	0.10976	0.07873	0.09731
Climate policy uncertainty	CPU	270	2.26003	2.26383	0.55862
Level of financial development	Fdl	270	1.54417	1.44305	0.81199
Size of foreign investment	FDI	270	1.74081	1.61042	1.40416
Higher education	Educa	270	13.52942	13.63824	0.79548
Scale of science and technology expenditures	Ste	270	0.00479	0.00367	0.00272
Fixed-asset investment	Fai	270	9.64450	9.74737	0.80066
Share of secondary industry	Seci	270	0.41186	0.42308	0.08267

reports a variable summary. Focusing on the core explanatory variable—NQPF—its mean value is 0.10976, and the standard deviation is 0.09731. This indicates that the level of NQPF varies considerably across different provinces in China. The relatively large standard deviation relative to the mean suggests the presence of significant regional disparities in the development of new, quality productive forces. Such variation provides a solid empirical basis for exploring the heterogeneous effects of climate policy uncertainty and financial factors on regional productivity dynamics. The VIFs are all less than 10, according to testing, suggesting that multiple collinearities between the interpretive variables are not a severe problem.

4. Regression Results

4.1. Baseline Regression Results

Table 3. Baseline regression results.

Variable	(1)	(2)
CPU	−0.015 **	−0.018 ***
	(−2.31)	(−2.96)
Fdl		−0.004
		(−1.45)
FDI		−0.004
		(−1.43)
Educa		−0.126 ***
		(−4.25)
Ste		5.299 **
		(2.54)
Fai		0.061 ***
		(6.17)
Seci		0.226 ***
		(2.60)
Constant	0.099 ***	1.098 ***
	(7.62)	(2.90)
Observations	270	270
R-squared	0.497	0.630
Number of regions	30	30
F	25.39	25.57

(Note: t in parentheses; ** p < 0.05, *** p < 0.01).

reports the results of the baseline regression of CPU and NQPF. Among them, column (1) shows the regression result controlling only for year and province fixed effects. The results show that there is a significant negative relationship between CPU and regional NQPF, indicating that CPU significantly hinders NQPF. Column (2) adds control variables on the basis of column (1). The results show that the sign and significance level of CPU are not significantly changed. The above analyses show that CPU has a significant negative impact on regional NQPF. Hypothesis 1 of this paper is verified; that is, CPU significantly hinders regional NQPF.

4.2. Robustness Check

4.2.1. Exclusion of Specific Samples

As part of the robustness strategy, this study modifies the sample to examine whether particular time periods or regional characteristics are unduly influencing the results. First, we address the potential distortions introduced by the COVID-19 pandemic. Since the outbreak severely disrupted business operations across the country, especially in 2020 and 2021, we re-run the baseline regression excluding these two years to control for the exogenous shock of the pandemic.

Second, we consider the unique economic and institutional roles of Beijing and Shanghai. As China’s political and financial centers, respectively, these municipalities often follow policy trajectories and development patterns distinct from those of other provinces. To avoid their outsized influence on the overall results, we exclude these two samples and conduct parallel regression analyses. The aim is to verify whether the relationship between CPU and NQPF holds when these outlier regions are removed from the dataset. The stability of the results under these modifications further supports the robustness of the main conclusions.

Table 4. Exclusion of specific samples.

Variable	(1)	(2)
CPU	−0.011 *	−0.014 **
	(−1.73)	(−2.27)
Fdl	−0.003	−0.003
	(−1.36)	(−1.25)
FDI	−0.005 **	−0.005 *
	(−2.05)	(−1.86)
Educa	−0.109 ***	−0.081 **
	(−3.69)	(−2.38)
Ste	8.958 ***	7.878 ***
	(4.85)	(3.37)
Fai	0.039 ***	0.056 ***
	(4.59)	(5.41)
Seci	0.184 **	0.193 **
	(2.32)	(2.16)
Constant	1.070 ***	0.540
	(2.83)	(1.25)
Observations	210	252
R-squared	0.652	0.598
Number of regions	30	28
F	24.03	20.69

(Note: t in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01).

presents the results of the robustness checks. Column (1) displays the estimation results after excluding data from 2020 and 2021 to mitigate potential distortions caused by the COVID-19 pandemic. Column (2) reports the outcomes after removing observations from Beijing and Shanghai, given their unique political and economic status. In both specifications, the estimated coefficients on CPU remain significantly negative, indicating that the adverse effect of climate policy uncertainty on new quality productive forces is robust.

4.2.2. Endogenous Issues

To address potential endogeneity concerns, this study adopts multiple econometric strategies to ensure the credibility of the causal inference. Specifically, a two-stage least squares (2SLS) approach is employed to mitigate bias in estimating the impact of CPU on NQPF.

In the 2SLS framework, we utilize lagged values of the potentially endogenous regressors as instrumental variables. These lagged instruments are theoretically justified and assumed to be exogenous, as they are predetermined and thus unlikely to be correlated with contemporaneous error terms. In the first stage, the endogenous variable—CPU—is regressed on its lagged instruments, generating predicted values that represent its exogenous variation. In the second stage, these fitted values are substituted into the main regression model to estimate their effect on the dependent variable, thereby correcting for possible endogeneity.

The estimation results, reported in column (1) of

Table 5. Endogenous issues.

Variable	(1) 2SLS	(2) GMM
NQPF		0.996 ***
		(71.53)
CPU	−0.034 **	−0.003 ***
	(−1.98)	(−7.82)
Control variable	YES	YES
Kleibergen–Paap rk LM statistic	17.855
Cragg–Donald Wald F statistic	33.628
Hensen test p value		1.00

(Note: t in parentheses; ** p < 0.05, *** p < 0.01. The Kleibergen–Paap rk LM statistic is under the identification test, and the Cragg–Donald Wald F statistic is a weak identification test).

, show that the coefficient of CPU remains significantly negative and consistent with the baseline regression, indicating that the adverse impact of climate policy uncertainty on NQPF is not driven by endogeneity bias.

As a further robustness check, we employ the system generalized method of moments (System GMM) estimator, which is well suited for dynamic panel data analysis. This method helps to control for unobserved heterogeneity, potential endogeneity of explanatory variables, and autocorrelation in error terms. By incorporating both the level and first-difference equations and using appropriate internal instruments, System GMM enhances estimation efficiency and reliability. Column (2) of Table 5 presents the System GMM results, confirming the robustness of the negative association between CPU and NQPF. The consistency across both 2SLS and System GMM estimations reinforces the credibility of our findings and suggests that the impact of climate policy uncertainty is both statistically and economically significant.

4.3. The Moderating Effects of Inclusive Finance

Inclusive finance aims to provide accessible and affordable financial services to all individuals and businesses, regardless of their wealth or company size [50]. Traditional financial institutions often fail to meet the needs of enterprises in less-developed regions and small enterprises, restricting their access to capital and limiting investment in productivity-enhancing activities. Inclusive finance, by expanding financial service coverage and addressing the shortcomings of traditional finance, can play a crucial role in supporting economic development and reducing regional financial disparities [59]. Inclusive finance may be an essential mechanism to mitigate the negative impact of CPU on NQPF. First, inclusive finance broadens access to credit and financial services, particularly for small and medium-sized enterprises and innovative firms, ensuring that they can obtain the necessary funding even in an uncertain policy environment. Additionally, the diversified funding channels provided by inclusive finance strengthen firms’ resilience and help them hedge against financial risks associated with policy uncertainty. By fostering financial inclusivity and resilience, inclusive finance can serve as a stabilizing force, supporting firms in navigating CPU and promoting sustainable NQPF.

Based on Equation (2), this paper uses the Financial Inclusion Index compiled by Peking University to measure the level of inclusive finance. The regression results of Equation (2) are shown in the first column of

Table 6. The moderating effects analysis of inclusive finance.

Variable	(1)	(2)
CPU	−0.099 ***	−0.112 ***
	(−5.30)	(−6.08)
IF	0.002 ***	0.001
	(6.12)	(2.49)
CPU × IF	0.0003 ***	0.0004 ***
	(4.78)	(5.44)
Control variable	NO	YES
Fixed effect	YES	YES
R2	0.953	0.712

(Note: t in parentheses; *** p < 0.01).

. The coefficient of the cross term is significantly positive in both columns (1) and (2), indicating that inclusive finance slows down the inhibitory effect of CPU. This result proves Hypothesis 2.

Regions with stronger financial inclusion are better equipped to buffer the adverse effects of CPU, as improved access to credit and capital enables firms to maintain investment in innovation and green transformation despite policy ambiguity. These findings underscore the importance of fostering inclusive financial systems to enhance regions’ adaptive capacity under environmental policy uncertainty.

4.4. The Role of Opening Carbon Markets

The carbon market, by establishing a predictable framework, enables businesses to make informed decisions about investment and technological innovation, particularly in areas that drive high-quality, sustainable development. One of the key advantages of the carbon market is its market-driven incentives, which encourage firms to adopt energy-efficient and emission-reduction technologies. Companies that effectively lower their emissions can trade surplus carbon allowances, generating additional revenue while reinforcing their commitment to green innovation. This system not only aligns environmental responsibility with economic benefits but also counteracts the inhibitory effect of CPU on NQPF, as firms operating in a more transparent and financially incentivized environment are more likely to invest in productivity-enhancing innovation. Moreover, the flexibility of the carbon market allows businesses to achieve carbon reduction targets through diverse strategic approaches, helping them manage compliance costs while maintaining competitiveness. Enhanced policy clarity and financial predictability strengthen firms’ confidence in making long-term, high-quality investments.

In 2013, Beijing, Shanghai, Tianjin, Chongqing, Guangdong, Hubei, and Shenzhen launched pilot carbon emissions trading programs. Based on Equation (3), this paper introduces a dummy variable to indicate whether the carbon market is open, with an open carbon market coded as 1 and a non-open market coded as 0. To ensure comparability and data availability, the sample timeframe is extended to start in 2010 (three years prior to the start of the pilot), and the sample period ends in 2020, just before China’s national carbon emissions trading market officially launched in 2021. Accordingly, this section uses data from 2010 to 2020 for analysis.

The regression results of Equation (3) are shown in the second column of

Table 7. The moderating effects analysis of opening carbon markets.

Variable	(1)	(2)
CPU	−0.010	−0.016 ***
	(−1.42)	(−2.65)
CM	−0.034	−0.027
	(−1.28)	(−1.17)
CPU × CM	0.036 ***	0.024 **
	(3.17)	(2.30)
Control variable	NO	YES
Fixed effect	YES	YES
R2	0.504	0.626

(Note: t in parentheses; ** p < 0.05, *** p < 0.01).

. The coefficient of the cross term is significantly positive, indicating that the carbon market slows down the inhibitory effect of CPU. This result proves hypothesis 3: Opening the carbon market reduces the negative impact of CPU on NQPF.

These results reveal that participation in carbon markets significantly moderates the negative effects of climate policy uncertainty. By providing a clear pricing mechanism and a transparent regulatory framework, carbon markets offer firms a more stable expectation of future policy directions, thereby reducing decision-making hesitation. In regions where enterprises actively engage in carbon trading, anticipating long-term returns and policy consistency enhances their willingness to pursue green investments. This finding underscores the importance of market-based environmental instruments in building institutional credibility and strengthening firms’ adaptive capacity in the face of policy uncertainty.

4.5. Heterogeneity Analysis

4.5.1. Regional Heterogeneity Analysis

Given the significant regional disparities in economic development, studying the impact of CPU on NQPF must account for differences in economic, policy, and technological conditions across regions. Since Chinese provinces vary substantially in their development levels, industrial structures, and policy environments, the mechanisms through which CPU affects NQPF may differ accordingly. Ignoring these regional differences may introduce bias into the overall analysis, so it is important to analyze the different regions separately.

According to the division found in the China Statistical Yearbook, this study divides the sample of 30 Chinese provinces into four groups: eastern, central, western, and northeastern regions. The eastern region is the most economically developed part of China, characterized by a high level of industrialization, strong innovation capacity, and a more market-oriented economic structure. It includes major economic hubs such as Beijing, Shanghai, and Guangdong, which are home to a dense concentration of advanced manufacturing, financial services, and high-tech industries. The region also has well-developed infrastructure that supports business growth and technological innovation. Unlike the eastern region, which has a more advanced high-tech sector, the central region relies on resource-intensive production and traditional manufacturing. While infrastructure improvements and policy support have facilitated industrial expansion, the transition toward low-carbon and high-value-added industries remains slow. The western region, which includes provinces such as Xinjiang and Gansu, is less developed due to geographic and infrastructural constraints, and it relies more on resource-based industries and government support. The northeastern region, historically an industrial base, has faced economic restructuring challenges in recent years, leading to slower growth compared to the eastern provinces.

Table 8. Subgroup regression.

Variable	(1)	(2)	(3)	(4)
CPU	−0.033 **	0.011	−0.009 ***	−0.004
	(−2.52)	(1.28)	(−2.74)	(−1.09)
Fin	−0.088 ***	−0.001	0.001	−0.020
	(−3.88)	(−0.54)	(0.15)	(−1.31)
FDI	−0.004	−0.002	−0.015 ***	0.004 **
	(−0.98)	(−0.38)	(−6.01)	(2.38)
Edu	−0.162 *	0.029	−0.044 ***	0.045
	(−1.78)	(0.59)	(−3.58)	(0.75)
Sci	10.152 **	6.872 **	0.895	−5.045
	(2.63)	(2.46)	(0.80)	(−1.17)
Inv	0.245 ***	0.067 ***	0.035 ***	−0.016 ***
	(5.53)	(5.22)	(6.86)	(−3.36)
Ind	0.050	0.112	−0.028	0.058
	(0.17)	(1.04)	(−0.67)	(1.09)
Constant	0.121	−1.076	0.331 **	−0.407
	(0.10)	(−1.53)	(2.32)	(−0.51)
Observations	90	54	99	27
R-squared	0.840	0.799	0.897	0.953
F	22.68	15.38	42.23	12.06

(Note: t in parentheses; * p < 0.1, ** p < 0.05, *** p < 0.01).

presents the regression results for different regions: column (1) corresponds to the eastern region, column (2) to the central region, column (3) to the western region, and column (4) to the northeastern region. The results show that CPU significantly inhibits NQPF in the eastern and western regions, as indicated by the significantly negative coefficients. In contrast, the coefficients for the central and northeastern regions are not statistically significant, suggesting that CPU does not have a discernible impact on NQPF in these areas.

The impact of CPU on NQPF varies significantly across regions due to differences in economic structure, industrial development models, marketization, policy dependence, and government intervention. The eastern and western regions experience a significant inhibitory effect from CPU on NQPF, whereas the central and northeastern regions remain largely unaffected. First, as the most economically developed area in China, the eastern region relies heavily on high-tech industries, advanced manufacturing, and modern services, particularly in sectors such as new energy, intelligent manufacturing, biopharmaceuticals, and information technology. These industries are highly dependent on a stable policy environment, and when the CPU increases (e.g., adjustments in carbon taxes, fluctuations in carbon markets, and changes in environmental subsidies), firms tend to delay investment and R&D, thereby hindering the development of NQPF. Moreover, enterprises in the eastern region operate in a highly market-oriented economy, with a large proportion of private and foreign-owned firms that are more sensitive to policy instability. When faced with policy uncertainty, these firms often adopt conservative strategies, reducing investments in low-carbon technologies, which further constrains NQPF development. In contrast, the central region’s industrial structure remains more traditional, with a strong reliance on manufacturing and agriculture. Businesses in this region are more driven by existing production systems and cost efficiency rather than policy incentives for low-carbon transitions, which weakens the influence of CPU. Similarly, the northeastern region, which has historically relied on heavy industries and state-owned enterprises, has faced economic stagnation and industrial restructuring in recent years. Given that firms in this region already have low incentives for technological innovation and limited investment capacity, policy uncertainty does not have a substantial impact on their decision-making regarding NQPF.

On the other hand, the western region, despite being less developed, experiences a significant inhibitory effect from CPU on NQPF due to its economic structure, which relies heavily on resource-based industries, energy extraction, and government investment. When the CPU increases, uncertainty surrounding new energy projects, emerging manufacturing industries, and infrastructure investment escalates, leading firms to delay investments and stalling the growth of NQPF. In recent years, the western region has accelerated its transition to renewable energy industries such as solar and wind power, but these projects require long-term, stable policy support. Increased CPU usage may deter investors and delay industrial expansion, constraining the development of emerging sectors. Furthermore, government-led infrastructure investments, which are critical for the western region’s economic growth, may become more uncertain under CPU, affecting the allocation of funds and slowing the development of projects related to NQPF. Consequently, heightened CPU lowers investment confidence in the western region, restricting NQPF development. In contrast, firms in the central and northeastern regions are less sensitive to policy uncertainty. The central region’s manufacturing industries depend largely on domestic supply chains rather than policy-driven subsidies, making them less affected by CPU fluctuations. Meanwhile, northeastern firms, due to industrial decline, weak investment incentives, and population outflows, already exhibit low levels of innovation and technological upgrading. Investment and production decisions in the northeast rely more on government financial support rather than market dynamics, reducing the impact of CPU on NQPF.

4.5.2. ESG Performance of Local Governments

Climate policy uncertainty itself is not merely a negative factor; there are underlying positive motives and potential impacts behind it. Governments often face complex and rapidly changing challenges, such as extreme climate events, shifts in international cooperation, and rapid technological advancements. To effectively address these unforeseen situations, governments need to maintain a certain degree of flexibility in their policies in the short term, avoiding rigid policies that could hinder their ability to respond. This flexibility, while essential for dealing with sudden crises and adapting to new circumstances, inevitably results in an increase in policy uncertainty. However, whether climate policy uncertainty leads to positive or negative outcomes largely depends on the government’s commitment to sustainable development. If the government prioritizes sustainability and demonstrates a strong commitment to economic transformation, this policy uncertainty may instead create opportunities for innovation. When firms and stakeholders recognize that the government’s long-term goal is sustainable development and that this goal is consistent and persistent, they are more willing to invest in green technologies and clean energy, even amid short-term uncertainties, to meet future policy demands. The government’s performance in terms of sustainable development is a critical factor driving innovation. A well-performing government typically conveys its commitment to green development through transparent policy, established legal frameworks, and incentives. Such actions send a stable, long-term signal to the market that, despite short-term policy changes, green economic development is the future direction. As a result, in the face of policy uncertainty, firms are encouraged to develop new technologies to meet potentially stricter environmental standards and evolving policy requirements, thereby fostering innovation. This innovation is not limited to technology alone but also extends to management, production processes, and business models. By constantly adapting to climate policy uncertainty, firms enhance their ability to respond to external changes. These factors above all drive the development of NQPF.

Therefore, in provinces where the government demonstrates strong performance in sustainable development, CPU may not necessarily inhibit NQPF.

Government ESG scores provide a good measure of the performance of governments in sustainable development. The 2023 Tsinghua University Sustainable Development Report provides government ESG ratings for Chinese provinces from 2016 to 2020, categorizing them into six levels: AAA, AA, A, BBB, BB, and B. This study refers to these ratings and classifies provinces into high and low ESG groups based on their average ratings over this period. Provinces that consistently maintained an ESG rating of A or higher (i.e., A, AA, or AAA) from 2016 to 2020 are classified as high ESG provinces. These include Beijing, Shanghai, Jiangsu, Guangdong, and Zhejiang. The remaining provinces, which did not meet this criterion, are classified as low ESG provinces.

Table 9. Sustainable development of local governments.

Variable	(1)	(2)
CPU	−0.018 ***	0.001
	(−2.96)	(0.08)
Fin	−0.004	−0.107 ***
	(−1.45)	(−5.01)
FDI	−0.004	−0.003
	(−1.43)	(−0.52)
Edu	−0.126 ***	0.484 ***
	(−4.25)	(4.26)
Sci	5.299 **	7.734 **
	(2.54)	(2.29)
Inv	0.061 ***	0.044
	(6.17)	(0.64)
Ind	0.226 ***	0.403
	(2.60)	(0.98)
Constant	1.098 ***	−6.950 ***
	(2.90)	(−5.87)
R-squared	0.630	0.975
F	25.57	65.59

(Note: t in parentheses; ** p < 0.05, *** p < 0.01).

reports the regression results. The first column shows the regression results for low local government ESG, and the second column shows the regression results for high local government ESG. It can be seen that climate policy uncertainty did not dampen new quality productivity if local government ESG scores were high, which indicates that the government’s capacity for sustainable development effectively mitigates the inhibitory effect of CPU on NQPF.

5. Conclusions and Policy Recommendations

This paper explores the impact of climate policy uncertainty on the new quality productive forces using panel data from thirty provinces in China from 2013 to 2021. The results find that climate policy uncertainty significantly suppresses new, quality productive forces. The finding remains robust after robustness and endogeneity tests. However, the development of inclusive finance and the opening of carbon markets can reduce the negative impact of climate policy uncertainty. Finally, this paper finds that the effect of climate policy uncertainty is also related to the sustainable development performance of local governments.

This paper’s result is consistent with a number of prior studies, which have also identified the negative effects of climate policy uncertainty on economic and developmental outcomes. For instance, Ren et al. [15], Dai and Zhu [60], and Sun et al. [61] provide empirical evidence that heightened climate policy uncertainty tends to hinder investment, innovation, and productivity growth. Our findings are in line with this broader understanding that policy volatility, particularly in the context of environmental regulation, can undermine long-term development objectives. This study contributes to the existing literature by focusing on the emerging concept of NQPF, offering fresh insight into how uncertainty affects not only traditional economic indicators but also the transformation toward high-quality and sustainable development.

Based on the conclusions, this paper makes the following recommendations.

Strengthening the Predictability and Transparency of Climate Policy: To reduce the adverse effects of policy uncertainty on firms’ investment behavior, it is essential for policymakers to enhance both the consistency and clarity of climate-related regulations. Establishing a transparent, stable, and long-term climate policy roadmap will enable enterprises to form stable expectations, reduce risk perceptions, and make forward-looking decisions. This is particularly critical for stimulating investment in innovation, green technologies, and other productivity-enhancing initiatives that require long-term strategic planning and stable policy signals.

Deepening Financial Inclusion to Support Enterprise Resilience: Developing a more inclusive and adaptive financial system can significantly ease the constraints faced by enterprises—especially small and medium-sized firms—under uncertain policy environments. By expanding access to credit, improving the coverage of green finance, and supporting diversified financing channels, inclusive finance can buffer the impact of external shocks. This, in turn, allows firms to maintain investment momentum, enhance technological upgrading, and sustain the growth of new, quality productive forces even in the face of regulatory fluctuations.

Advancing the Carbon Market Framework for Risk Management: A well-functioning and transparent carbon market plays a crucial role in mitigating climate-related policy uncertainty. Policymakers should prioritize the expansion and refinement of national and regional carbon trading schemes, including broader sectoral coverage, more effective allocation mechanisms, and higher market liquidity. By improving the institutional infrastructure of carbon trading, firms will be better positioned to manage emissions-related compliance costs, hedge against regulatory risk, and integrate carbon considerations into their long-term development strategies. Such market-based tools can help stabilize firms’ expectations and promote green investment behavior.

This study contributes to the growing body of literature on the intersection between environmental governance and economic productivity by introducing NQPF as a novel outcome variable. By exploring how the CPU undermines the development of NQPF—an increasingly central component in China’s high-quality growth paradigm—this paper provides a fresh empirical perspective that links policy volatility to innovation and green transformation. The results offer theoretical value for researchers interested in policy uncertainty and institutional economics, and practical guidance for policymakers seeking to build resilient innovation ecosystems.

Finally, it is necessary to clarify the limitations of this paper. While the empirical strategy adopted in this study includes instrumental variable approaches to mitigate endogeneity concerns, the direction of causality between climate policy uncertainty and the development of new quality productive forces remains open to debate. It is possible that the weak development or stagnation of NQPF in certain regions may itself increase policy ambiguity, for instance, by reducing local capacity for policy implementation or undermining confidence in long-term green innovation pathways. Future research could explore bidirectional or dynamic relationships between CPU and NQPF using methods such as structural equation modeling or panel vector autoregression.

While this study provides empirical evidence of the negative effects of CPU on the development of NQPF, there remain several avenues for future exploration. First, future research could expand the scope of analysis by incorporating more granular data at the enterprise level. Second, comparative studies across countries, particularly in other emerging economies undergoing green transitions, could assess whether similar mechanisms apply in different institutional contexts. Third, the role of innovation ecosystems and technological collaboration networks in buffering CPU could be further explored using social network analysis. These extensions will not only enrich the empirical literature but also offer more practical insights for tailoring adaptive governance strategies in the face of climate uncertainty.