How Do Stringent Climate Policies Influence Carbon Intensity and Profitability? Insights from the Chinese Market
-
Yichen Liu
and
Doojin Ryu
Abstract
This study investigates the economic and environmental impacts of China’s Green Finance Reform and Innovation Pilot Zone (GFRIPZ) program, a climate policy designed to promote sustainability by reducing carbon emissions intensity (CEI) among high-energy-consuming firms. Using a dataset of Chinese A-share listed companies, triple-difference estimations reveal that the GFRIPZ program significantly lowers CEI. This outcome is partly driven by increased investments in green innovation, as firms enhance their research and development efforts to align with environmental goals. Despite these advancements, the program presents a trade-off by exerting downward pressure on profitability. However, green finance policies mitigate financing constraints, supporting firms’ financial stability under sustainability-focused reforms. Our findings explain the interplay between environmental outcomes and financial performance, highlighting the necessity of balancing sustainable development and economic growth.
1 Introduction
As global warming intensifies, reducing carbon emissions and improving environmental, social, and governance (ESG) performance have become an urgent global priority. Governments around the world are implementing policies that integrate the principle of ESG to balance carbon emission reduction and firm sustainability (Feng et al. 2025; Goodell et al. 2024). As the world’s largest carbon emitter, China is committed to achieving carbon peaking by 2030 and carbon neutrality by 2060 (Xu et al. 2023). In order to achieve the climate target, the Chinese government established the Green Finance Reform and Innovation Pilot Zone (GFRIPZ) in 2017 to reduce carbon emissions and support sustainable economic development.
The actual effects of green finance policies often exhibit complexity and stage-specific dynamics. The green paradox suggests that stringent environmental regulation can reduce firms’ profits in the short run and, due to expectations of tighter future policies, may induce firms to accelerate production or resource extraction, leading to higher short-term emissions and undermining policy effectiveness (Zhang et al. 2017). The innovation paradox emphasizes that, although green innovation yields long-term benefits, it is often associated with high short-term costs and uncertainty, requiring firms to absorb initial pressures to achieve sustainable value (Li et al. 2024). These theories remind us that green financial reform may produce diversified economic and environmental effects at different stages of development.
High-energy-consuming industries account for 80 % of China’s carbon emissions (Wang et al. 2019). These industries face challenges in responding to environmental protection requirements and green financial constraints, including broader ESG issues. For example, improving the governance structure can strengthen compliance with environmental standards by promoting corporate accountability and risk management strategies. Firms that fail to solve sustainable development problems may face reputation risks, resulting in a decline in stakeholder trust and market competitiveness.[1]
As climate-related risks increase and ESG-driven policies tighten, high-energy-consuming firms face higher operating and financing costs, resulting in reduced productivity and profitability (Batten et al. 2019; Chen et al. 2022; Gao 2022). A decline in profitability may lower firms’ credit ratings and consequently increase borrowing costs (Hasan et al. 2023). Maintaining a high credit rating enables firms to obtain lower loan interest rates. This eases financial constraints and frees up funds for investments that meet ESG standards, thus improving environmental performance. These investments enable firms to meet strict environmental standards. They also enhance market competitiveness and financial health by improving energy efficiency and reducing long-term operating costs.
Carbon emission reduction policies will bring macroeconomic challenges and have ESG implications (Dissanayake et al. 2020). While excessively stringent regulation can suppress short-term output and economic growth in high-energy-consuming industries (Liu et al. 2021), insufficient regulatory stringency may fail to deliver the intended environmental outcomes. Policymakers therefore face a fundamental trade-off between emission reduction and economic stability.
Green financial reforms across countries and regions have shown substantial heterogeneity. In emerging economies such as India and Brazil, there are both policy-driven improvements in environmental performance (Chauhan and Thangavel 2025) and policy backlash that inhibit innovation due to industrial structure or institutional differences (Khan and Vismara 2025). Compared with these economies, China has unique characteristics in terms of institutional environment, industrial structure, and policy implementation. A systematic assessment of high-energy-consuming firms can test the effectiveness of the GFRIPZ program and provide a reference for other emerging economies.
This study investigates the impact of the GFRIPZ program on the carbon emissions intensity (CEI) of high-energy-consuming firms. CEI is measured as the natural logarithm of firm-level carbon emissions divided by total sales. We investigate whether the GFRIPZ program has effectively reduced the CEI of these firms, and explore the operating mechanisms of the policy, including green innovation, profitability, and financing constraints. Our China-specific evidence provides a theoretical and empirical basis for optimizing global green finance policy.
We use the financial data of Chinese A-share listed companies from 2013 to 2021 and apply a triple-difference model to identify the causal effect of the policy. Although the CEI of high-energy-consuming industries will increase over time, the triple interaction term, which captures time, industry, and region, shows a significant negative coefficient. This suggests that the GFRIPZ program effectively reduces the CEI of high-energy-consuming firms in pilot zones. Robustness checks, including parallel trend, instrumental variable (IV), and placebo tests, support the validity of our findings. Mechanism analysis suggests that GFRIPZ reduces CEI by enhancing firms’ green innovation capacity. We also find that the program lowers short-term profitability but boosts long-term market value and eases financing constraints for high-energy-consuming firms.
We contribute to the literature on green finance, carbon emissions, and firm sustainability in several ways. First, we analyze the GFRIPZ program from the perspective of sustainable development to clarify the conflict between economic growth and environmental goals. Although the green financial reform may reduce the short-term profitability of high-energy-consuming firms, the program has increased the long-term market value, which is consistent with the innovation paradox. Second, we use the triple-difference model to assess the impact of green financial policies on carbon emissions, which provides methodological insights for other emerging markets. Third, we identify the underlying mechanisms at the micro level. Green financial policy has reduced CEI by easing financing constraints and promoting green innovation, enriching the evidence of transmission channels. Fourth, although the current green financial instruments have played a positive role, there are still limitations in incentive design and coverage. This underscores the need for better-designed and more broadly implemented green financial mechanisms to support long-term sustainability.
This study proceeds as follows. Section 2 reviews the relevant literature and research background. Section 3 explains the data and variable constructions. Section 4 presents the empirical results. Section 5 includes robustness tests. Section 6 discusses the findings. Section 7 concludes the paper.
2 Literature and Research Background
In the context of the global promotion of green financial policies, firms need to reduce carbon emissions and improve environmental performance. In information-asymmetric markets, investors favor firms that combine strong environmental outcomes with sound corporate governance (Chuang and Huang 2018). By adopting green strategies and investing in environmental protection, firms can reduce carbon emissions in the short term and generate greater long-term business value (Wang et al. 2021).
Green financial policies affect firms’ CEI in many ways. Under financing constraints, firms tend to reduce investment in green technology (Yang 2023). Green financial policies can provide tax relief and other measures to encourage firms to invest in green technology (Yu et al. 2021). The higher the ESG score, the lower the risk of default and the smaller the financing constraint (He et al. 2023). These advantages encourage firms to adopt low-carbon technologies and clean energy projects, improving energy efficiency and reducing carbon emissions (Yang et al. 2021).
With the support of green financial policies, firms can apply for green patents more actively. Green innovation can reduce carbon emissions while reducing energy consumption (Kirat et al. 2024). Green patents can also enhance the competitiveness of firms (Li et al. 2019). The impact of technological progress on the environment is still controversial. On the one hand, technological improvement can improve resource efficiency and promote green growth (Liu et al. 2025). On the other hand, in highly polluted countries, technological progress can promote economic growth, but the impact on emissions is different (Celik et al. 2025).
An increase in management’s shareholding ratio often aligns with a firm’s long-term sustainability goals. Firms with higher insider ownership focus more on long-term interests. They are therefore more likely to implement carbon-reduction policies and adopt environmental measures (Haque 2017). Thus, firms can use green finance policies to alleviate financing constraints and increase green R&D investments, channeling more funds into environmental management and reducing carbon emissions (Xi and Jia 2025).
A growing body of research examines how the green and innovation paradoxes emerge across countries and regions. Developed economies such as the United States and the European Union usually have relatively well-established supporting policies and innovative measures, and the green financial system is relatively mature. Firms in developed economies typically manage to balance carbon-reduction pressures with profitability goals, and the paradox effects are relatively muted (Dechezleprêtre et al. 2023; Flammer 2021). Emerging economies are quite different from developed economies in terms of industrial structure and policy implementation. Thus, the green financial policy faces greater challenges. During the implementation of India’s green financial policy, carbon emissions in some industries do not decrease significantly (Oak and Bansal 2022), and some firms reduce environmental investments due to financing pressures (Ghosh and Dutta 2022). The introduction of an environmental tax in Chile can significantly reduce emissions, but higher tax rates are related to lower employment and higher poverty (Mardones and Mena 2020). The introduction of a carbon tax in Indonesia reduces emissions. While the economy declines in the short term, it recovers in the long run (Hartono et al. 2023).
According to the Porter hypothesis, moderate environmental policies can stimulate internal technological innovation of firms. Firms can offset the additional costs brought by environmental policies through technological innovation and gain a competitive advantage (Weiss et al. 2019). However, some scholars believe that strict policy constraints may increase short-term operating costs and reduce profitability (Alexander et al. 2024). Based on this, we put forward the following hypotheses:
H1:
The GFRIPZ program significantly reduces the CEI of high-energy-consuming firms.
H2:
The GFRIPZ program affects the CEI through green technological innovation.
H3:
The GFRIPZ program reduces the short-term profitability of high-energy-consuming firms.
H4:
The GFRIPZ program alleviates the financing constraints for high-energy-consuming firms.
The existing research mainly explores the relationship between green financial policies and the environmental performance of high-pollution firms (Lin and Zhang 2023). High-energy-consuming firms usually have a higher CEI, but with limited attention. Exploring the impact of green financial policies on such firms is crucial to assessing the effectiveness of policies, promoting sustainable development, and reducing emissions. This study aims to evaluate whether the GFRIPZ program can reduce the CEI of high-energy-consuming firms and contribute to sustainable development goals.
3 Sample Data and Methodology
3.1 Dataset
We analyze the companies listed in the Chinese A-share market from 2013 to 2021. The annual financial data of the firms come from the China Stock Market & Accounting Research Database (CSMAR). CEI is defined as the natural logarithm of the ratio of carbon emissions to total sales of firms (Luo and Tang 2014). Carbon emissions data are collected from the annual reports, social responsibility reports, and environmental reports released by A-share listed companies. Carbon emissions are calculated by combining a firm’s Scope 1 (direct emissions) and Scope 2 (indirect emissions from purchased electricity/heat, etc.) (Doda et al. 2016). We exclude samples with abnormal financial conditions, including irregular figures or trends in financial reports, significant deviations from industry norms in profitability or balance sheet indicators, delisting warnings, or listing dates after 2013. Our research sample consists of annual observations for 936 firms.
We identify high-energy-consuming industries based on the “National Economic Industry Classification” released by the National Bureau of Statistics of China in 2021. The industries and their corresponding codes are petroleum, coal, and other fuel processing (C25); chemical manufacturing (C26); non-metallic mineral products (C30); ferrous metal smelting and rolling processing (C31); non-ferrous metal smelting and rolling processing (C32); and electricity and heat production and supply (C44). Our sample consists of A-share listed companies with available carbon emissions data. Based on the above industry classification, we classify firms into high-energy-consuming industries (HighEnergy = 1) and other industries (HighEnergy = 0) to implement the triple-difference design across time, regions, and industry types. We focus on A-share listed companies because their environmental and financial disclosures are more complete and standardized, meeting the comparability requirements of a triple-difference design. Listed companies are typically industry leaders with higher energy consumption and greater exposure to environmental regulation, making them suitable for identifying policy impacts. In contrast, inconsistent and discontinuous disclosure among unlisted firms prevents the construction of reliable environmental efficiency measures. Therefore, our analysis focuses on listed companies, while the results remain informative for the broader population of high-energy-consuming firms.
In this analysis, we use CEI as the dependent variable. We construct three indicator variables: Post, HighEnergy, and PilotZone. Post is a dummy variable equal to 1 for observations in years at or after the policy implementation year (2017), and 0 otherwise. HighEnergy equals 1 if a firm belongs to a high-energy-consuming industry and 0 otherwise. PilotZone equals 1 if a firm is located in a policy pilot zone and 0 otherwise. We use return on assets (ROA) to measure the short-term profitability of firms and Tobin’s Q (Tobin_Q) to measure their long-term firm value. The number of green patents granted in a given year (Green) is used as a measure of green innovation. Patent data are obtained from the China National Intellectual Property Administration (CNIPA). In addition, we construct a measure of green innovation efficiency (GIE), defined as the ratio of green innovation output to innovation input, which is used in the mediation analysis. To measure financing constraints, we employ the SA index (SA) (Hadlock and Pierce 2010). When CEI is the dependent variable, we include firm-level controls for operating dynamics and governance: operating income growth rate (Growth), management shareholding proportion (Mshare), and R&D investment (Input). These variables absorb the concurrent effects of expansion, incentives, and technological effort on carbon intensity. When ROA is the dependent variable, we include firm size (LnSize), listing age (LnAge), asset tangibility (AT), current ratio (CR), and board independence (BoardIndep) to mitigate structural heterogeneity in performance. The SA index is calculated based on LnSize and the number of years the firm has been listed (Age). To isolate the effect of the GFRIPZ program from other mandatory green policies, we add two policy dummy variables as controls: Policy1 for the 2015 Environmental Protection Law and Policy2 for the 2016 Energy Conservation Law, each indicating whether a firm is affected by these policies. Table 1 explains the definition and construction of these variables.
Definition of variables.
| Variable | Definition |
|---|---|
| CEI | Defined as the carbon emissions intensity of the firm. The calculation formula is: CEI = ln(carbon emissions/total sales). |
| Post | Dummy variable equal to 1 if the year is the policy implementation year (2017) or later, and 0 otherwise. |
| PilotZone | Dummy variable equal to 1 if the firm is located in a policy implementation zone, and 0 otherwise. |
| HighEnergy | Dummy variable equal to 1 if the firm belongs to a high-energy-consuming industry, and 0 otherwise. |
| ROA | Defined as the return on assets of the firm. |
| Tobin_Q | Defined as the Tobin’s Q of the firm. |
| Green | Defined as the number of green patents obtained by the firm in that year. |
| GIE | Defined as the ratio of green innovation output to R&D investment, where green innovation output is defined as ln(1 + the total number of green inventions, utility models, and design patents). |
| SA | Defined as the firm-level financing constraint index. The calculation formula is: SA = −0.737 × LnSize + 0.043 × LnSize2 – 0.04 × Age. |
| Growth | Defined as the growth rate of the firm’s operating income. |
| Mshare | Defined as the shareholding ratio held by firm management. |
| Input | Defined as the natural logarithm of the amount of R&D investment by the firm. |
| LnSize | Defined as the natural logarithm of the total assets of the firm. |
| LnAge | Defined as the natural logarithm of the number of years the firm has been listed. |
| AT | Defined as the ratio of tangible assets of the firm. The calculation formula is: AT = (total assets – net intangible assets – net goodwill)/total assets. |
| CR | Defined as the current ratio of the firm. The calculation formula is: CR = current assets/current liabilities. |
| BoardIndep | Defined as the proportion of independent directors of the firm. |
| Age | Defined as the number of years the firm has been listed, used in the calculation of SA. |
| Policy1 | Dummy variable equal to 1 if the firm is affected by the Environmental Protection Law promulgated in 2015, and 0 otherwise. |
| Policy2 | Dummy variable equal to 1 if the firm is affected by the Energy Conservation Law of the People’s Republic of China promulgated in 2016, and 0 otherwise. |
In order to improve the analysis and reduce the extreme value effect, we divide the sample according to whether it is a high-energy-consuming firm. We winsorize the continuous variables within each group at the 1st and 99th percentiles. Table 2 reports the descriptive statistics of sub-sample variables. The average CEI value of high-energy-consuming firms (6.61) is higher than that of other firms (5.71), indicating that there is potential for improvement in the future. All variables regarding firm characteristics show significant differences between the two groups.
Descriptive statistics.
| Group | Mean | Std | Min | Q1 | Med | Q3 | Max | t-stat | |
|---|---|---|---|---|---|---|---|---|---|
| CEI | A | 6.61 | 1.20 | 4.62 | 5.71 | 6.40 | 7.25 | 10.23 | 27.02 |
| B | 5.71 | 1.16 | 3.34 | 4.94 | 5.70 | 6.36 | 9.25 | ||
| ROA | A | 0.04 | 0.05 | −0.12 | 0.01 | 0.04 | 0.07 | 0.22 | 1.84 |
| B | 0.04 | 0.06 | −0.19 | 0.01 | 0.04 | 0.07 | 0.20 | ||
| Tobin_Q | A | 1.93 | 1.04 | 0.76 | 1.22 | 1.61 | 2.28 | 9.81 | −13.35 |
| B | 2.34 | 1.44 | 0.76 | 1.42 | 1.92 | 2.74 | 14.91 | ||
| Green | A | 2.25 | 6.03 | 0.00 | 0.00 | 0.00 | 2.00 | 42.00 | −5.49 |
| B | 3.30 | 9.72 | 0.00 | 0.00 | 0.00 | 2.00 | 69.00 | ||
| GIE | A | 0.03 | 0.05 | 0.00 | 0.00 | 0.00 | 0.06 | 0.22 | −1.66 |
| B | 0.03 | 0.05 | 0.00 | 0.00 | 0.00 | 0.06 | 0.28 | ||
| SA | A | 4.95 | 1.54 | 1.63 | 3.83 | 4.70 | 5.89 | 10.21 | 7.72 |
| B | 4.62 | 1.46 | 1.43 | 3.63 | 4.37 | 5.28 | 13.64 | ||
| Growth | A | 0.18 | 0.42 | −0.41 | −0.03 | 0.10 | 0.27 | 2.79 | 2.86 |
| B | 0.15 | 0.29 | −0.44 | −0.01 | 0.11 | 0.25 | 1.53 | ||
| Mshare | A | 0.10 | 0.16 | 0.00 | 0.00 | 0.00 | 0.16 | 0.57 | −4.03 |
| B | 0.12 | 0.17 | 0.00 | 0.00 | 0.01 | 0.21 | 0.63 | ||
| Input | A | 18.09 | 1.54 | 13.58 | 17.18 | 18.05 | 19.06 | 21.62 | −5.51 |
| B | 18.32 | 1.40 | 14.94 | 17.41 | 18.24 | 19.16 | 22.37 | ||
| LnSize | A | 22.69 | 1.31 | 20.33 | 21.72 | 22.53 | 23.61 | 26.34 | 8.57 |
| B | 22.39 | 1.20 | 20.17 | 21.54 | 22.23 | 23.06 | 25.99 | ||
| LnAge | A | 2.48 | 0.55 | 1.10 | 2.08 | 2.49 | 2.94 | 3.33 | 5.57 |
| B | 2.39 | 0.56 | 1.10 | 1.95 | 2.40 | 2.89 | 3.33 | ||
| AT | A | 0.93 | 0.07 | 0.62 | 0.92 | 0.95 | 0.97 | 1.00 | 2.68 |
| B | 0.92 | 0.08 | 0.59 | 0.91 | 0.95 | 0.97 | 1.00 | ||
| CR | A | 1.71 | 1.38 | 0.26 | 0.85 | 1.32 | 2.08 | 8.43 | −16.97 |
| B | 2.44 | 2.11 | 0.66 | 1.24 | 1.72 | 2.73 | 13.50 | ||
| BoardIndep | A | 0.37 | 0.05 | 0.33 | 0.33 | 0.33 | 0.40 | 0.54 | −8.15 |
| B | 0.38 | 0.05 | 0.33 | 0.33 | 0.36 | 0.43 | 0.57 | ||
| Age | A | 13.61 | 6.56 | 2.00 | 8.00 | 12.00 | 19.00 | 30.00 | 5.48 |
| B | 12.61 | 6.49 | 2.00 | 7.00 | 11.00 | 18.00 | 30.00 | ||
| Policy1 | A | 0.66 | 0.47 | 0.00 | 0.00 | 1.00 | 1.00 | 1.00 | −0.80 |
| B | 0.67 | 0.47 | 0.00 | 0.00 | 1.00 | 1.00 | 1.00 | ||
| Policy2 | A | 0.77 | 0.42 | 0.00 | 1.00 | 1.00 | 1.00 | 1.00 | −0.56 |
| B | 0.78 | 0.42 | 0.00 | 1.00 | 1.00 | 1.00 | 1.00 |
-
Group, Mean, Std, Min, Q1, Med, Q3, and Max represent firm type, mean, standard deviation, minimum, 25th percentile, median, 75th percentile, and maximum values, respectively. Group A includes high-energy-consuming firms, with 1,609 observations, and Group B includes the remaining firms, with 6,815 observations. A t-test is conducted to compare variables between the two groups, and t-stat denotes its statistical value. CEI is the natural logarithm of the ratio of carbon emissions to total sales. ROA is return on assets. Tobin_Q is Tobin’s Q of the firm. Green is the number of green patents. GIE is a proxy for green innovation efficiency. SA is the financing constraint index. Additional variables include operating income growth rate (Growth), management shareholding proportion (Mshare), the natural logarithm of R&D investment (Input), the natural logarithm of firm size (LnSize), the natural logarithm of firm age measured by listing years (LnAge), asset tangibility (AT), current ratio (CR), board independence (BoardIndep), firm age measured as the number of years the firm has been listed (Age), and dummy variables for the 2015 Environmental Protection Law (Policy1) and the 2016 Energy Conservation Law (Policy2).
3.2 Models
We consider the GFRIPZ program a quasi-natural experiment. Compared with the difference-in-differences method, the triple-difference model introduces an additional comparison layer (Olden and Møen 2022). We estimate the impact of the policy on CEI by sequentially adding double interaction, triple interaction, and control variables. Our model uses a triple-difference approach as follows:
where CEI i,t represents the carbon emissions intensity of firm i in year t; Post t is a dummy variable that takes the value of 1 if the calendar year t is 2017 or later (after the policy implementation year), and 0 otherwise; HighEnergy i,t is a dummy variable assigned the value of 1 for firm i in year t if it belongs to a high-energy-consuming industry, characterized by substantial energy consumption and significant environmental impact in the production process, and 0 otherwise; PilotZone i,t is a dummy variable that takes the value 1 if firm i is within the policy implementation zone in year t, and 0 otherwise; Control i,t are control variables including operating income growth rate (Growth), management shareholding proportion (Mshare) and R&D investment (Input). To avoid interference from other mandatory green policies on the impact of the GFRIPZ program, we add dummy variables for the 2015 Environmental Protection Law (Policy1) and the 2016 Energy Conservation Law (Policy2) as control variables. This allows us to determine whether firms are affected by these policies. We also include industry fixed effects (IndustryFE i ) and year fixed effects (YearFE t ). ε i,t is the error term.
In our triple-difference design, the treatment group comprises firms located in GFRIPZ pilot zones and operating in high-energy-consuming industries during the post-policy period. First, M1 examines the basic differences across policy implementation year (Post), pilot zones (PilotZone), and high-energy-consuming industries (HighEnergy). Second, M2 introduces three two-way interaction terms: Post × HighEnergy captures the temporal differences in high-energy-consuming industries as the policy changes; Post × PilotZone captures the temporal differences in firms in the pilot zones as the policy changes; and PilotZone × HighEnergy captures the structural differences between geographical location and industry characteristics. Third, M3 incorporates the triple interaction term (Post × PilotZone × HighEnergy) to construct the triple-difference model, which identifies the causal effect of the GFRIPZ policy on the CEI of high-energy-consuming firms in pilot zones. Finally, M4 includes control variables to account for firm-level characteristics and to test the robustness of the results.
4 Empirical Analyses
Through the baseline regression models represented by M1, M2, M3, and M4, the impact of the GFRIPZ program on firms’ CEI is tested. Table 3 shows the estimated results of the triple-difference model. In the first column (M1), the coefficient of HighEnergy is positive and significant. This shows that high-energy-consuming firms have higher CEI. The significant negative coefficient of PilotZone shows that firms in the pilot zone exhibit lower CEI. When the pairwise interaction terms among HighEnergy, Post, and PilotZone are added to the second column (M2), the coefficient on Post × HighEnergy is significantly positive. This indicates that over time, high-energy-consuming firms maintain a higher CEI relative to other firms, even after policy enactment. The third column (M3) introduces the triple interaction term Post × PilotZone × HighEnergy, which is significantly negative (−0.490), indicating that the program reduces CEI in high-energy-consuming firms within pilot zones. Even after adding control variables in M4, the results remain unchanged. This indicates that, compared with other firms, the CEI of high-energy-consuming firms in the pilot zone is significantly lower after the policy is implemented, which confirms our hypothesis (H1).
Triple-difference estimation.
| CEI | ||||
|---|---|---|---|---|
| (1) | (2) | (3) | (4) | |
| Intercept | 5.796*** | 5.825*** | 5.827*** | 5.085*** |
| (16.38) | (16.49) | (16.49) | (13.60) | |
| HighEnergy | 1.107*** | 0.954*** | 0.942*** | 0.932*** |
| (37.65) | (22.33) | (21.82) | (22.08) | |
| PilotZone | −0.147*** | −0.152* | −0.200** | −0.265*** |
| (−2.82) | (−1.86) | (−2.32) | (−3.14) | |
| Post × HighEnergy | 0.290*** | 0.311*** | 0.284*** | |
| (5.23) | (5.49) | (5.13) | ||
| Post × PilotZone | 0.064 | 0.148 | 0.189* | |
| (0.61) | (1.29) | (1.70) | ||
| PilotZone × HighEnergy | −0.172 | 0.084 | 0.222 | |
| (−1.22) | (0.41) | (1.12) | ||
| Post × PilotZone × HighEnergy | −0.490* | −0.527* | ||
| (−1.78) | (−1.96) | |||
| Growth | 0.276*** | |||
| (8.11) | ||||
| Mshare | −1.145*** | |||
| (−17.48) | ||||
| Input | 0.044*** | |||
| (5.35) | ||||
| Policy1 | −0.008 | |||
| (−0.454) | ||||
| Policy2 | −0.036 | |||
| (−1.50) | ||||
| Observations | 8,424 | 8,424 | 8,424 | 8,424 |
| Adjusted-R2 | 0.334 | 0.336 | 0.336 | 0.368 |
| Industry-fixed | Yes | Yes | Yes | Yes |
| Time-fixed | Yes | Yes | Yes | Yes |
-
This table reports the results of the triple-difference estimator. Figures in parentheses represent t-statistics. Significance levels of 10 %, 5 %, and 1 % are indicated by *, **, and ***, respectively. CEI is the dependent variable, defined as the natural logarithm of carbon emissions divided by total sales. Post is a dummy variable equal to 1 if the year is 2017 or later, otherwise 0. HighEnergy is a dummy variable equal to 1 for high-energy-consuming industries, otherwise 0. PilotZone is a dummy variable equal to 1 if the firm is in a policy implementation zone, otherwise 0. Additional variables include operating income growth rate (Growth), management shareholding proportion (Mshare), R&D investment (Input), and dummy variables for the 2015 Environmental Protection Law (Policy1) and the 2016 Energy Conservation Law (Policy2).
We also conduct a parallel trend test, replacing Post with a set of year variables. Panels A and B in Figure 1 present the results of the parallel trend test after splitting the sample by whether industries are high-energy-consuming. Panel A presents the parallel-trend test for high-energy-consuming firms in the pilot versus non-pilot zones. The results show that there is no significant difference in CEI before the policy implementation, which supports the parallel trend assumption. Panel B shows the parallel trend results of non-high-energy-consuming firms and compares the pilot zones and non-pilot zones. This indicates that there are no significant differences between these firms during the whole sample period. This further confirms the impact of the GFRIPZ program on the CEI of high-energy-consuming firms.
Parallel trend test. Notes: This figure reports the results of the parallel trend test. The x-axis is the time point; the y-axis is the policy dynamic effect; the vertical dotted line represents the time point of policy implementation; the hollow points represent the estimated effect at each time point; the error bars represent the confidence interval of each estimated effect. Panel A shows the results for high-energy-consuming firms in pilot and non-pilot zones, while Panel B shows the results for non-high-energy-consuming firms in pilot and non-pilot zones.
Although the triple-difference approach mitigates some endogeneity problems, bidirectional causality may still exist between the GFRIPZ program and firm-level carbon emissions. For example, a city’s pilot policy may affect firm-level carbon emissions, while pre-existing emission patterns could also influence government policy decisions. To alleviate this concern, we employ an IV based on high-speed rail opening in the city where the firm is located, which equals one if the city had opened high-speed rail by 2014, and zero otherwise. Data on high-speed rail openings are collected from local government official websites and the Baidu search engine. First, the opening of high-speed rail is closely related to local economic development and government planning. Cities selected as GFRIPZ pilots are typically resource-rich and economically developed, and thus tend to open high-speed rail earlier, satisfying the relevance condition. Second, high-speed rail opening mainly reflects long-term regional infrastructure development and connectivity rather than firm-level production or emission decisions. As such, it does not directly determine firm-level carbon emissions intensity and can therefore be used as an IV. To make the procedure explicit, we estimate the model using the IV approach implemented via two-stage least squares (2SLS). In the first step, we use M5 as the first-stage regression to obtain the predicted value of the triple interaction term, E(Post × PilotZone × HighEnergy):
In the second stage, we replace Post × PilotZone × HighEnergy in M4 with the fitted value E(Post × PilotZone × HighEnergy) obtained from the first stage. Table 4 reports the results of the endogeneity test using the IV method. The significance of the IV in the first column indicates its effectiveness in predicting Post × PilotZone × HighEnergy, thereby satisfying the correlation condition. In the second stage, the coefficient on E(Post × PilotZone × HighEnergy) remains negative and statistically significant, supporting our main conclusion. It indicates that even after removing potential bidirectional causality between policy selection and firm characteristics, the GFRIPZ program still significantly reduces the CEI of high-energy-consuming firms.
Endogeneity test.
|
Post × PilotZone × HighEnergy
|
CEI
|
|
|---|---|---|
| (1) | (2) | |
| Intercept | −0.011 | 4.702*** |
| (−0.46) | (12.26) | |
| IV | 0.004** | |
| (2.54) | ||
| HighEnergy | 0.927*** | |
| (22.13) | ||
| PilotZone | −0.197** | |
| (−2.46) | ||
| Post × HighEnergy | 0.261*** | |
| (4.81) | ||
| Post × PilotZone | 0.098 | |
| (0.97) | ||
| PilotZone × HighEnergy | −0.033 | |
| (−0.24) | ||
| E(Post × PilotZone × HighEnergy) | −27.62*** | |
| (−4.42) | ||
| Growth | 0.000 | 0.274*** |
| (0.01) | (8.07) | |
| Mshare | 0.007* | −0.921*** |
| (1.76) | (−11.12) | |
| Input | 0.000 | 0.059*** |
| (0.73) | (6.63) | |
| Policy1 | 0.036*** | 0.093*** |
| (3.55) | (3.32) | |
| Policy2 | 0.017 | 0.011 |
| (1.12) | (0.42) | |
| Observations | 8,424 | 8,424 |
| Industry-fixed | Yes | Yes |
| Time-fixed | Yes | Yes |
| First-stage | 17.31 | |
| Adjusted-R2 | 0.046 |
-
This table reports the results of the endogeneity test. Figures in parentheses represent t-statistics. Significance levels of 10 %, 5 %, and 1 % are indicated by *, **, and ***, respectively. The dependent variable in the first stage is Post × PilotZone × HighEnergy. CEI is the dependent variable in the second stage, defined as the natural logarithm of carbon emissions divided by total sales. The instrumental variable, IV, equals one if the city in which the firm is located had opened high-speed rail by 2014, and zero otherwise. Post is a dummy variable equal to 1 if the year is 2017 or later, otherwise 0. HighEnergy is a dummy variable equal to 1 for high-energy-consuming industries, otherwise 0. PilotZone is a dummy variable equal to 1 if the firm is in a policy implementation zone, otherwise 0. The E(Post × PilotZone × HighEnergy) variable is the predicted value obtained in the first stage regression using the selected instrumental variable. First-stage represents the first-stage F-statistic. Additional variables include operating income growth rate (Growth), management shareholding proportion (Mshare), R&D investment (Input), and dummy variables for the 2015 Environmental Protection Law (Policy1) and the 2016 Energy Conservation Law (Policy2).
5 Robustness Tests and Additional Analyses
5.1 Placebo Tests
We further test the credibility and effectiveness of the GFRIPZ program’s impact on the CEI of high-energy-consuming firms through a placebo test. The procedure involves randomly assigning 191 firms to the high-energy-consuming group and 44 to the pilot-zone group, reflecting the proportions in the actual sample. These firms constitute the treatment group, while the remaining firms form the control group. Under this setup, we perform 500 iterations of the regression analysis using the same model as in the main analysis (M4). Figure 2 shows the results of the simulation tests. Most coefficients of the triple interaction term cluster near zero and lack statistical significance. This indicates that the effects observed in the original model are likely attributable to the actual policy rather than to unobserved factors.
Placebo test. Notes: This figure reports the results of the placebo test. The x-axis is the coefficient of the triple interaction term; the y-axis is the p-value corresponding to the triple interaction term; the curve is the kernel density of the estimated coefficients; scatter points are the p-values of the corresponding estimated coefficients. Panel A reports the placebo test results of 500 regressions simulating “HighEnergy”. Panel B reports the placebo test results of 500 regressions simulating “PilotZone”.
5.2 Sensitivity and Heterogeneity
To further assess robustness, we re-estimate the triple-difference model using the full sample without winsorization and test for industry heterogeneity. We sequentially relabel the manufacturing, electricity, real estate, technology, and mining industries as high-energy-consuming industries and re-estimate the model. If the results stem from industry trends or model misspecification, these placebo treatments would show significant effects. Table 5 presents the sensitivity and heterogeneity results. In the non-winsorized regression (Column 1), the coefficient on Post × PilotZone × HighEnergy remains significant and consistent with the baseline, indicating that outliers do not drive the findings. In the heterogeneity regressions (Columns 2–6), the triple interaction coefficients are insignificant, suggesting that the policy effect is specific to high-energy-consuming industries. The evidence supports the validity of our identification strategy and confirms the robustness of the main results.
Sensitivity and heterogeneity tests.
| CEI | ||||||
|---|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Intercept | 5.171*** | 5.383*** | 5.370*** | 5.375*** | 5.375*** | 5.375*** |
| (13.52) | (13.25) | (13.27) | (13.28) | (13.28) | (13.28) | |
| HighEnergy | 0.930*** | −0.208 | 0.884*** | 0.191 | 0.586*** | 0.691*** |
| (21.41) | (−0.69) | (4.40) | (0.91) | (2.70) | (3.54) | |
| PilotZone | −0.256*** | −0.539* | −0.262*** | −0.279*** | −0.280*** | −0.281*** |
| (−2.96) | (−1.93) | (−3.08) | (−3.35) | (−3.39) | (−3.40) | |
| Post × PilotZone | 0.177 | −0.009 | 0.088 | 0.072 | 0.072 | 0.074 |
| (1.54) | (−0.02) | (0.78) | (0.65) | (0.66) | (0.67) | |
| PilotZone × HighEnergy | 0.170 | 0.281 | −0.202 | −0.001 | −0.000** | 0.000*** |
| (0.84) | (0.96) | (−0.58) | (−0.00) | (−2.14) | (3.30) | |
| Post × HighEnergy | 0.306*** | 0.019 | 0.244 | 0.099 | −0.169 | 0.048 |
| (5.38) | (0.23) | (1.22) | (0.39) | (−0.49) | (0.29) | |
| Post × PilotZone × HighEnergy | −0.507* | 0.090 | −0.418 | −0.006 | −0.000 | 0.000 |
| (−1.83) | (0.24) | (−0.89) | (−0.01) | (−1.47) | (0.29) | |
| Growth | 0.050*** | 0.336*** | 0.336*** | 0.337*** | 0.336*** | 0.336*** |
| (4.13) | (9.14) | (9.15) | (9.15) | (9.15) | (9.14) | |
| Mshare | −1.100*** | −1.272*** | −1.272*** | −1.277*** | −1.276*** | −1.276*** |
| (−16.55) | (−17.95) | (−17.95) | (−18.03) | (−18.03) | (−18.03) | |
| Input | 0.041*** | 0.026*** | 0.026*** | 0.026*** | 0.026*** | 0.026*** |
| (5.15) | (2.93) | (2.96) | (2.92) | (2.92) | (2.92) | |
| Policy1 | −0.006 | 0.010 | 0.017 | 0.019 | 0.019 | 0.019 |
| (−0.32) | (0.25) | (0.99) | (1.07) | (1.11) | (1.06) | |
| Policy2 | −0.044* | −0.025 | −0.021 | −0.021 | −0.020 | −0.021 |
| (−1.76) | (−0.78) | (−0.83) | (−0.80) | (−0.78) | (−0.80) | |
| Observations | 8,424 | 8,424 | 8,424 | 8,424 | 8,424 | 8,424 |
| Adjusted-R2 | 0.364 | 0.258 | 0.259 | 0.258 | 0.258 | 0.258 |
| Industry-fixed | Yes | Yes | Yes | Yes | Yes | Yes |
| Time-fixed | Yes | Yes | Yes | Yes | Yes | Yes |
-
This table reports the results of the sensitivity and heterogeneity tests. Column (1) reports the baseline regression using the non-winsorized full sample. Columns (2)–(6) sequentially treat manufacturing, electricity, real estate, technology, and mining as high-energy-consuming industries in heterogeneity tests. Figures in parentheses represent t-statistics. Significance levels of 10 %, 5 %, and 1 % are indicated by *, **, and ***, respectively. CEI is the dependent variable, defined as the natural logarithm of carbon emissions divided by total sales. Post is a dummy variable equal to 1 if the year is 2017 or later, otherwise 0. HighEnergy is a dummy variable equal to 1 for high-energy-consuming industries, otherwise 0. PilotZone is a dummy variable equal to 1 if the firm is in a policy implementation zone, otherwise 0. Additional variables include operating income growth rate (Growth), management shareholding proportion (Mshare), R&D investment (Input), and dummy variables for the 2015 Environmental Protection Law (Policy1) and the 2016 Energy Conservation Law (Policy2).
5.3 Mediation Effects
To examine the role of green innovation in reducing CEI, we conduct a mediation analysis, using the number of green patents (Green) as a proxy for firms’ green innovation capability. As a complementary analysis, we further use the ratio of green innovation output to innovation input as a proxy for green innovation efficiency (GIE). Innovation input is measured by R&D investment, as firm-level data on innovation input are not publicly available. Green innovation output is measured as the logarithm of one plus the total number of green inventions, utility models, and design patents. Table 6 reports the mediating role of green innovation capabilities. The regression coefficients on Post × PilotZone × HighEnergy in the first and fourth columns are significantly positive. This suggests that the GFRIPZ program enhances the green innovation capabilities of high-energy-consuming firms in the pilot zones. Combined with the results in the remaining columns, it indicates that the GFRIPZ program reduces CEI by improving the green innovation capabilities of high-energy-consuming firms, thus confirming our hypothesis (H2).
Mediation effect analysis.
|
Green |
CEI |
CEI |
GIE |
CEI |
CEI |
|
|---|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Intercept | −40.77*** | 4.870*** | 4.869*** | −0.208*** | 4.991*** | 4.991*** |
| (−12.54) | (12.92) | (12.91) | (−11.59) | (13.24) | (13.25) | |
| HighEnergy | 0.020 | 0.944*** | 0.932*** | 0.004** | 0.946*** | 0.934*** |
| (0.05) | (22.63) | (22.10) | (2.19) | (22.65) | (22.12) | |
| PilotZone | −1.040 | −0.221*** | −0.270*** | 0.000 | −0.215*** | −0.265*** |
| (−1.42) | (−2.76) | (−3.21) | (0.06) | (−2.68) | (−3.14) | |
| Post × HighEnergy | −0.489 | 0.260*** | 0.281*** | −0.001 | 0.262*** | 0.283*** |
| (−1.01) | (4.80) | (5.08) | (−0.55) | (4.83) | (5.12) | |
| Post × PilotZone | −0.652 | 0.099 | 0.186* | −0.005 | 0.100 | 0.187* |
| (−0.67) | (0.97) | (1.67) | (−0.94) | (0.98) | (1.68) | |
| PilotZone × HighEnergy | 0.296 | −0.041 | 0.223 | −0.007 | −0.046 | 0.218 |
| (0.17) | (−0.30) | (1.13) | (−0.76) | (−0.33) | (1.11) | |
| Post × PilotZone × HighEnergy | 4.072* | −0.505* | 0.045*** | −0.506* | ||
| (1.74) | (−1.88) | (3.48) | (−1.88) | |||
| Green | −0.005*** | −0.005*** | ||||
| (−4.26) | (−4.23) | |||||
| GIE | −0.466** | −0.450** | ||||
| (−2.05) | (−1.98) | |||||
| Growth | −0.574* | 0.272*** | 0.273*** | −0.003 | 0.274*** | 0.275*** |
| (−1.94) | (8.00) | (8.02) | (−1.57) | (8.05) | (8.07) | |
| Mshare | −0.121 | −1.146*** | −1.146*** | 0.001 | −1.145*** | −1.145*** |
| (−0.21) | (−17.51) | (−17.51) | (0.35) | (−17.48) | (−17.48) | |
| Input | 2.255*** | 0.055*** | 0.055*** | 0.012*** | 0.049*** | 0.049*** |
| (31.89) | (6.43) | (6.45) | (29.91) | (5.69) | (5.70) | |
| Policy1 | −0.432*** | −0.008 | −0.010 | −0.003*** | −0.007 | −0.009 |
| (−2.93) | (−0.46) | (−0.59) | (−3.13) | (−0.40) | (−0.52) | |
| Policy2 | −0.087 | −0.036 | −0.037 | −0.000 | −0.035 | −0.036 |
| (−0.42) | (−1.48) | (−1.52) | (−0.15) | (−1.46) | (−1.51) | |
| Observations | 8,424 | 8,424 | 8,424 | 8,424 | 8,424 | 8,424 |
| Adjusted-R2 | 0.146 | 0.369 | 0.369 | 0.176 | 0.370 | 0.370 |
| Industry-fixed | Yes | Yes | Yes | Yes | Yes | Yes |
| Time-fixed | Yes | Yes | Yes | Yes | Yes | Yes |
-
This table reports the results of the mediation effect analysis. Columns (1)–(3) report the mediation results using Green. Columns (4)–(6) report the mediation results using GIE. In columns (1) and (4), the dependent variables are Green and GIE, respectively. In all other columns, the dependent variable is CEI. Figures in parentheses represent t-statistics. Significance levels of 10 %, 5 %, and 1 % are indicated by *, **, and ***, respectively. Green is the number of green patents, and GIE is a proxy for green innovation efficiency. CEI is the natural logarithm of the ratio of carbon emissions to total sales. Post is a dummy variable equal to 1 if the year is 2017 or later, otherwise 0. HighEnergy is a dummy variable equal to 1 for high-energy-consuming industries, otherwise 0. PilotZone is a dummy variable equal to 1 if the firm is in a policy implementation zone, otherwise 0. Additional variables include operating income growth rate (Growth), management shareholding proportion (Mshare), R&D investment (Input), and dummy variables for the 2015 Environmental Protection Law (Policy1) and the 2016 Energy Conservation Law (Policy2).
5.4 Moderating Effects
The relationship between CEI and firm profitability is also examined to understand the broader impact of the GFRIPZ program. Following Zhang et al. (2023), we use ROA as the dependent variable in a new model that includes a four-way interaction term, Post × PilotZone × HighEnergy × CEI. This model examines how policy implementation timing, pilot zones, industry type, and carbon emissions intensity interact to influence firm profitability. We also use ROA one and two periods ahead to examine its forward-looking effects. To better capture the differences between high-energy-consuming firms and the control group, we include firm size (LnSize), listing age (LnAge), asset tangibility (AT), current ratio (CR), and board independence (BoardIndep) as control variables in M6. The model is as follows:
Table 7 shows that there is a significant negative correlation between CEI and ROA, which is consistent with the research results of Homroy (2023), who shows that reducing CEI improves profitability. The coefficient of Post × PilotZone × HighEnergy is significantly negative. This indicates that high-energy-consuming firms constrained by the GFRIPZ program have lower profitability than other firms. The positive interaction coefficient for Post × PilotZone × HighEnergy × CEI suggests that as these firms increase their CEI, their profitability also rises. This pattern persists when ROA is measured one and two periods ahead. This reflects their profit model, which relies heavily on raw material consumption and leads to higher carbon emissions. This does not contradict the negative average association; it indicates that the marginal profitability effect of CEI differs for treated firms after the policy. The results support H3: the GFRIPZ program reduces CEI but also lowers short-term profitability. The relationship between CEI and profitability is in line with prior research suggesting that ESG factors, including climate risk, influence corporate performance and capital allocation decisions (Ali et al. 2024).
Profitability analysis.
|
ROA t |
ROA t+1 |
ROA t+2 |
|
|---|---|---|---|
| (1) | (2) | (3) | |
| Intercept | −0.156*** | −0.215*** | −0.238*** |
| (−6.90) | (−7.66) | (−8.45) | |
| HighEnergy | −0.012*** | −0.008*** | 0.006 |
| (−5.04) | (−2.73) | (0.21) | |
| PilotZone | 0.005 | 0.002 | −0.000 |
| (1.12) | (0.34) | (−0.01) | |
| Post × HighEnergy | 0.024*** | 0.025*** | 0.015*** |
| (8.28) | (6.85) | (4.14) | |
| Post × PilotZone | −0.001 | 0.005 | 0.007 |
| (−0.21) | (0.76) | (0.96) | |
| PilotZone × HighEnergy | 0.002 | 0.004 | 0.012 |
| (0.23) | (0.31) | (0.96) | |
| Post × PilotZone × HighEnergy | −0.168*** | −0.207*** | −0.246*** |
| (−2.97) | (−2.94) | (−3.47) | |
| CEI | −0.002*** | −0.003*** | −0.003*** |
| (−2.96) | (−3.47) | (−4.71) | |
| Post × PilotZone × HighEnergy × CEI | 0.025*** | 0.028*** | 0.031*** |
| (3.14) | (2.83) | (3.14) | |
| Growth | 0.047*** | 0.040*** | 0.021*** |
| (25.81) | (17.85) | (9.26) | |
| Mshare | 0.023*** | 0.023*** | 0.018*** |
| (5.52) | (4.55) | (3.59) | |
| Input | 0.007*** | 0.008*** | 0.007*** |
| (12.17) | (10.14) | (9.12) | |
| LnSize | 0.000 | −0.001 | −0.001 |
| (0.33) | (−1.21) | (−0.73) | |
| LnAge | 0.001 | 0.004** | 0.005*** |
| (1.05) | (2.37) | (2.70) | |
| AT | 0.044*** | 0.123*** | 0.143*** |
| (6.11) | (13.63) | (15.68) | |
| CR | 0.002*** | 0.002*** | 0.002*** |
| (10.60) | (7.51) | (7.54) | |
| BoardIndep | −0.043*** | −0.026** | −0.028** |
| (−4.27) | (−2.11) | (−2.26) | |
| Policy1 | −0.008*** | −0.011*** | −0.021*** |
| (−8.43) | (−9.13) | (−17.70) | |
| Policy2 | −0.004*** | −0.005*** | −0.005*** |
| (−4.81) | (−3.01) | (−3.19) | |
| Observations | 8,424 | 8,424 | 8,424 |
| Adjusted-R2 | 0.166 | 0.118 | 0.300 |
| Industry-fixed | Yes | Yes | Yes |
| Time-fixed | Yes | Yes | Yes |
-
This table reports the results of the profitability analysis examining the moderating effects of CEI on firm profitability. Figures in parentheses represent t-statistics. Significance levels of 5 % and 1 % are indicated by ** and ***, respectively. In columns (1)–(3), the dependent variables are ROA, one-period-ahead ROA, and two-period-ahead ROA, respectively. ROA is the return on assets. Post is a dummy variable equal to 1 if the year is 2017 or later, otherwise 0. HighEnergy is a dummy variable equal to 1 for high-energy-consuming industries, otherwise 0. PilotZone is a dummy variable equal to 1 if the firm is in a policy implementation zone, otherwise 0. Additional variables include operating income growth rate (Growth), management shareholding proportion (Mshare), R&D investment (Input), firm size (LnSize), listing age (LnAge), asset tangibility (AT), current ratio (CR), board independence (BoardIndep), and dummy variables for the 2015 Environmental Protection Law (Policy1) and the 2016 Energy Conservation Law (Policy2).
5.5 Financing Constraints and Long-Term Firm Value
Under the influence of the GFRIPZ program, high-energy-consuming firms experience a decrease in ROA as they reduce CEI. To examine the program’s effects on long-term firm value and financing conditions, we further use Tobin’s Q (Tobin_Q) as a proxy for long-term firm value and the SA index (SA) as a firm-level measure of financing constraints. The SA index is calculated based on firm size (LnSize) and age (Age). We replace the dependent variables in the main analysis model (M4) with SA and Tobin_Q for further testing. The calculation formula for SA is as follows:
where Age is the number of years a firm has been listed.
Table 8 presents the results of the impact of the GFRIPZ program on the financing constraints and long-term value of high-energy-consuming firms in pilot zones. In the SA regression, the coefficient (Post × PilotZone × HighEnergy) is negative and significant. This indicates that the GFRIPZ program effectively alleviates financing constraints for high-energy-consuming firms in pilot zones, confirming our hypothesis (H4). Erhemjamts et al. (2024) also show that climate risk and ESG performance significantly affect capital acquisition and financial stability in the financial sector. In the Tobin_Q regression, the coefficient of Post × PilotZone × HighEnergy is positive and significant. Although the GFRIPZ program may depress firms’ short-term accounting profitability (ROA), it raises market expectations of long-term firm value and growth. Taken together, the two sets of results suggest a compensation mechanism: the program induces firms to undertake green upgrades and capital investments, which lower ROA, while easing financing constraints and improving transition expectations, which lift market valuation and thus offset the short-term decline in accounting profitability in the capital market dimension.
Financing constraints and long-term firm value analysis.
|
SA |
Tobin_Q |
|
|---|---|---|
| (1) | (2) | |
| Intercept | −8.730*** | 4.755*** |
| (−24.62) | (10.04) | |
| HighEnergy | 0.640*** | −0.681*** |
| (16.00) | (−12.74) | |
| PilotZone | −0.103 | 0.334*** |
| (−1.28) | (3.13) | |
| Post × HighEnergy | −0.069 | 0.157** |
| (−1.32) | (2.23) | |
| Post × PilotZone | 0.087 | −0.421*** |
| (0.83) | (−2.98) | |
| PilotZone × HighEnergy | −0.481** | −0.392 |
| (−2.57) | (−1.57) | |
| Post × PilotZone × HighEnergy | −0.614** | 0.578* |
| (−2.41) | (1.70) | |
| Growth | 0.069** | 0.584*** |
| (2.12) | (13.55) | |
| Mshare | −0.913*** | 0.147* |
| (−14.70) | (1.77) | |
| Input | 0.717*** | −0.193*** |
| (93.00) | (−18.70) | |
| Policy1 | −0.027* | −0.115*** |
| (−1.70) | (−5.34) | |
| Policy2 | 0.024 | 0.644*** |
| (1.03) | (21.06) | |
| Observations | 8,424 | 8,424 |
| Adjusted-R2 | 0.613 | 0.203 |
| Industry-fixed | Yes | Yes |
| Time-fixed | Yes | Yes |
-
This table reports the results of the analysis of financing constraints and long-term firm value. Figures in parentheses represent t-statistics. Significance levels of 10 %, 5 %, and 1 % are indicated by *, **, and ***, respectively. In columns (1) and (2), the dependent variables are SA and Tobin_Q, respectively. SA is the firm-level financing constraint index. The calculation formula is: SA = −0.737 × LnSize + 0.043 × LnSize 2 – 0.04 × Age. Tobin_Q serves as a proxy for long-term firm value. Post is a dummy variable equal to 1 if the year is 2017 or later, otherwise 0. HighEnergy is a dummy variable equal to 1 for high-energy-consuming industries, otherwise 0. PilotZone is a dummy variable equal to 1 if the firm is in a policy implementation zone, otherwise 0. Additional variables include the operating income growth rate (Growth), management shareholding proportion (Mshare), R&D investment (Input), and dummy variables for the 2015 Environmental Protection Law (Policy1) and the 2016 Energy Conservation Law (Policy2).
Our results show that the GFRIPZ program reduces the CEI of high-energy-consuming firms in the pilot zone. We further provide evidence that green innovation plays a mediating role in this process. The causal effect of the GFRIPZ program is supported by a series of placebo tests, which rule out alternative explanations. Although lowering CEI may reduce the short-term profitability of high-energy-consuming firms, it will increase the long-term market valuation. In addition, the GFRIPZ program eases these firms’ financing constraints. Taken together, these findings suggest that high-energy-consuming firms can achieve sustainable economic and environmental outcomes by developing green technologies while reducing carbon emissions.
6 Discussion
Although the GFRIPZ program can significantly reduce the CEI of high-energy-consuming firms, it reduces short-term profitability. This paradox of green transformation deserves further discussion.
6.1 Profitability Analysis
The GFRIPZ program reduces CEI primarily by enhancing firms’ capacity for green innovation. Green innovation often entails substantial upfront investment (López Pérez et al. 2024), and these expenditures are difficult to translate into operating profits in the short run and typically require the medium to long term to materialize. If firms lack the capacity to effectively absorb green technologies, or if market and policy mechanisms fail to provide timely, sufficient positive incentives (e.g., green credit, tax relief, or brand effects from green certification), such upfront investments are more likely to evolve into a sustained decline in profitability.
This decline in profitability is not irreversible, as we can see in subsection 5.5. According to the Porter Hypothesis, appropriate environmental regulation can stimulate innovation, improve resource utilization efficiency, and thus enhance the long-term firm value. If firms turn green investments into efficiency gains, product differentiation, or financing advantages, then profitability can recover and even exceed pre-policy levels.
6.2 Industry Heterogeneity Analysis
To further examine whether the abatement trade-off varies by industry, we conduct a heterogeneity robustness test in Subsection 5.2. We sequentially relabel non-target industries (e.g., manufacturing, electricity, real estate, technology, and mining) as high-energy-consuming industries and re-estimate the triple-difference model. Across all pseudo-settings, the triple-interaction coefficients remain insignificant. The effect of the green policy thus focuses on those truly high-energy-consuming industries.
The results suggest important heterogeneous policy effects across industries. Industries with high energy consumption are highly dependent on energy and resource usage during the production process, have relatively higher baseline levels of carbon emissions, and face much greater difficulty in pollution control and green upgrading compared with other sectors. For such sectors, green finance policy imposes the most stringent restrictions. For instance, differential interest rates and environmental licensing procedures have material impacts on credit and capital flows and day-to-day operations.
In contrast, while manufacturing is also an important part of China’s real economy, there are large variations within the sector. Light manufacturing (e.g., food processing and clothing) and heavy industry (e.g., steel production and shipbuilding) have different energy intensities and emission profiles. Service- or capital-intensive industries, such as technology and financial businesses, tend to be less energy-intensive and generally do not consume large quantities of energy. As a result, the policy has limited effects on these industries.
6.3 Balancing Abatement and Profitability
According to previous research, firms’ trade-off between environmental mitigation and economic profits is influenced by multiple factors. One important factor is firm size, which affects a firm’s ability to bear the costs of green transformation. Larger firms have stronger capital accumulation capacity and better access to external funding (Newton et al. 2024), which enables them to invest in greening capital stock, retrofitting technologies, and meeting environmental regulatory requirements (Jia and Azevedo 2025), thereby weakening the negative impact of green investment on profitability. By contrast, small and medium-sized enterprises (SMEs) often face the dilemma of “efficient mitigation but poor profitability” in response to policy pressure due to limited capital and restricted access to green finance (Amoozad Mahdiraji et al. 2024; Shahin et al. 2024).
Market orientation and export intensity are also important for balancing emissions reduction and profitability. Firms that rely heavily on foreign markets place greater weight on environmental reputation and ESG performance (Aray et al. 2021). Their green transformation may generate exogenous positive stimuli, such as international orders, which help convert environmental responsibility into profit opportunities (Bıçakcıoğlu et al. 2020). In contrast, when external market pressure is weak, firms depend more on policy incentives to undertake green investment (Aragòn-Correa et al. 2020).
Building on this literature, our study highlights the role of financing capacity. As shown in Table 8, the GFRIPZ program reduces financing constraints (lower SA) and increases firm valuation (higher Tobin’s Q) for high-energy-consuming firms in the treatment zones. These findings show that the green financial policy has also enhanced the ability of firms to obtain funds and long-term growth expectations. Therefore, firms need to weigh many factors and balance the relationship between emission reduction and profitability through the support of green financial policies to achieve sustainable development.
6.4 Pathways of Green Innovation
In Subsection 5.3, we use the number of green patents to proxy green innovation and show that the GFRIPZ program, by enhancing green-innovation capacity among high-energy-consuming firms, reduces their CEI. Extending this perspective, it is useful to explore how such innovation translates into concrete actions. Beyond patent counts, focusing on the quality and practical impact of innovation helps clarify the pathways through which firms transform green concepts into real abatement outcomes.
The first manifestation of green innovation is increased R&D investment. Driven by policy, high-energy-consuming firms have raised R&D spending on energy-saving and emission-reduction technologies, including clean technologies, waste-gas control and recovery, and carbon capture, utilization, and storage (CCUS) (Fikru et al. 2024; Khan et al. 2021). Although such investment increases costs in the short term, it helps strengthen competitive advantage over the long run (Hang et al. 2019). Firms can also promote the innovation and implementation of green technologies through Environmental Management Control Systems (EMCS) (Hennig et al. 2023).
Firms can further advance green innovation by expanding their business models. By offering green solutions and energy-efficient operational services, firms can convert green capabilities into both economic and environmental performance (Johl et al. 2024). They can also translate green innovation into economic benefits and achieve sustainability goals by improving technological efficiency to obtain green certifications (e.g., ISO 14001) and by building green supply chains (Song et al. 2026; Wang et al. 2023).
7 Conclusion and Policy Implications
This study takes the GFRIPZ program as a quasi-natural experiment and uses the triple-difference model to evaluate the impact of the policy on the CEI of high-energy-consuming firms from an ESG perspective. Our results show that the reform effectively reduces the CEI of high-energy-consuming firms in the pilot zones. Green innovation plays a mediating role in this process. In addition, it eases the financing constraints of high-energy-consuming firms in the pilot zone. We also find that when high-energy-consuming firms reduce CEI, their short-term profitability is negatively affected, while their long-term market value increases. These firms rely on energy- and resource-intensive production (e.g., coal-based inputs), which is associated with higher emissions intensity. It is more necessary to introduce green technology and optimize the production process. They may also resist ESG-oriented policies for reasons of profitability. Policymakers and financial institutions, such as banks, also need to strengthen regulation and support to help firms achieve a green transition.
Policymakers should strengthen green financial products and investor confidence by establishing a unified green project certification and information disclosure system. We also suggest that policymakers set up government-led green funds to attract investors’ attention to the green economy and improve the ability of firms to obtain green financing. The government can improve the efficiency of green capital allocation and the coverage of green finance through these measures. Policymakers should also support green-oriented firms facing transformation difficulties by setting up green transformation funds. They can provide flexible support for firms in different industries and at different stages of transformation, and guide firms to gradually achieve the goals of technological upgrading and sustainable development. Banks and financial institutions can incorporate firms’ green performance into the assessment standards for green credit ratings and risk evaluation to promote the development of green credit. They can also provide more preferential loan terms for firms that successfully improve their ESG performance. In addition, financial instruments such as green bonds and sustainability-linked loans can help firms achieve long-term environmental goals.
Although this study shows that the GFRIPZ program can significantly reduce the CEI of high-energy-consuming firms in the pilot zone, we acknowledge several limitations. First, due to constraints on data access, our sample does not include unlisted companies. Second, there is a time lag in the transformation of green patents into production technologies, and green innovation may also be subject to measurement error. However, we control for relevant policy factors and conduct parallel-trends tests and multiple robustness checks, which further reduce the influence of unobserved factors on our results.
Our study expands the research on how green finance and ESG strategies can promote environmental and economic sustainability at the same time. Our findings suggest that while green financial policies are effective in reducing corporate carbon emissions, they may impose short-term profitability costs on firms, even as they enhance market valuation and long-term growth prospects. We emphasize that policymakers need to weigh trade-offs and improve economic performance as much as possible while ensuring environmental performance. Therefore, policymakers and financial institutions such as banks can adjust their green-oriented policies accordingly to reduce the economic cost of ecological transformation. Future research can expand the scope of the data, include unlisted companies, and add more indicators to measure green innovation, so as to draw more robust conclusions.
Acknowledgment
The authors express their gratitude to Katharine Rocket (Editor-in-Chief), Małgorzata Milewska (Managing Editor), and anonymous reviewers for their insightful suggestions.
-
Research funding: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) [RS-2025-00518388].
-
Author contributions: All authors accept the responsibility for the entire content of this manuscript and consent to its submission to the journal, review all the results, and approve the final version of the manuscript. YL (First author): Methodology, Software, Formal analysis, Data Curation, and Visualization; HK (Corresponding author): Conceptualization, and Writing – Original Draft; DR (Corresponding author): Conceptualization, Validation, Investigation, Resources, Writing – Review & Editing, Supervision, and Project administration.
-
Conflict of interest: There is no conflict of interest. There is no competing interest.
-
Data availability statement: The data supporting the findings of this study are available from the corresponding authors upon reasonable request.
References
Alexander, Anna, Antonio De Vito, and Luca Menicacci. 2024. “At What Cost? Environmental Regulation and Corporate Cash Holdings.” Finance Research Letters 61: 104960. https://doi.org/10.1016/j.frl.2023.104960.Search in Google Scholar
Ali, Sara, Ihsan Badshah, Riza Demirer, Prasad Hegde, and Lavinia Rognone. 2024. “Climate Risk, ESG Ratings, and the Flow-Performance Relationship in Mutual Funds.” Global Finance Journal 63: 101041. https://doi.org/10.1016/j.gfj.2024.101041.Search in Google Scholar
Amoozad Mahdiraji, Hannan, Seyed Hossein Razavi Hajiagha, Vahid Jafari-Sadeghi, Donatella Busso, and Alain Devalle. 2024. “Towards Financing the Entrepreneurial SMEs: Exploring the Innovation Drivers of Successful Crowdfunding via a Multi-Layer Decision-Making Approach.” European Journal of Innovation Management 27 (7): 2275–301. https://doi.org/10.1108/EJIM-12-2021-0618.Search in Google Scholar
Aragòn-Correa, J. Alberto, Alfred A. Marcus, and David Vogel. 2020. “The Effects of Mandatory and Voluntary Regulatory Pressures on Firms’ Environmental Strategies: A Review and Recommendations for Future Research.” Academy of Management Annals 14 (1): 339–65. https://doi.org/10.5465/annals.2018.0014.Search in Google Scholar
Aray, Yulia, Desislava Dikova, Tatiana Garanina, and Anna Veselova. 2021. “The Hunt for International Legitimacy: Examining the Relationship Between Internationalization, State Ownership, Location and CSR Reporting of Russian Firms.” International Business Review 30 (5): 101858. https://doi.org/10.1016/j.ibusrev.2021.101858.Search in Google Scholar
Bang, Jeongseok, and Doojin Ryu. 2024. “ESG Factors and the Cross-Section of Expected Stock Returns: A LASSO-Based Approach.” Finance Research Letters 65: 105482. https://doi.org/10.1016/j.frl.2024.105482.Search in Google Scholar
Bang, Jeongseok, Doojin Ryu, and Jinyoung Yu. 2023a. “ESG Controversies and Investor Trading Behavior in the Korean Market.” Finance Research Letters 54: 103750. https://doi.org/10.1016/j.frl.2023.103750.Search in Google Scholar
Bang, Jeongseok, Doojin Ryu, and Robert I. Webb. 2023b. “ESG Controversy as a Potential Asset-Pricing Factor.” Finance Research Letters 58 (A): 104315. https://doi.org/10.1016/j.frl.2023.104315.Search in Google Scholar
Bang, Jeongseok, Hankil Kang, and Doojin Ryu. 2024. “ESG, Firm Image, and Explanatory Power for Stock Returns.” Applied Economics Letters 32 (16): 2315–21. https://doi.org/10.1080/13504851.2024.2332579.Search in Google Scholar
Batten, Jonathan A., Harald Kinateder, Peter G. Szilagyi, and Niklas F. Wagner. 2019. “Time-Varying Energy and Stock Market Integration in Asia.” Energy Economics 80: 777–92. https://doi.org/10.1016/j.eneco.2019.01.008.Search in Google Scholar
Bıçakcıoğlu, Nilay, Vasilis Theoharakis, and Mustafa Tanyeri. 2020. “Green Business Strategy and Export Performance: An Examination of Boundary Conditions from an Emerging Economy.” International Marketing Review 37 (1): 56–75. https://doi.org/10.1108/IMR-11-2018-0317.Search in Google Scholar
Celik, Ali, Ebru Gul Yilmaz, Sukran Kahveci, and Gonca Yilmaz. 2025. “Do Technological Advances Promote Ecological Sustainability? Evidence from Panel Causality Analysis for the Dirtiest and Cleanest Countries.” Journal of the Knowledge Economy 16 (5): 17292–327. https://doi.org/10.1007/s13132-024-02423-9.Search in Google Scholar
Chauhan, Anushka, and Mohanasundari Thangavel. 2025. “Assessing the Impact of the Perform, Achieve, and Trade Scheme on Energy Efficiency in India’s Iron and Steel Sector.” Energy Policy 198: 114480. https://doi.org/10.1016/j.enpol.2024.114480.Search in Google Scholar
Chen, Shaojian, Hui Mao, and Junqin Sun. 2022. “Low-Carbon City Construction and Corporate Carbon Reduction Performance: Evidence from a Quasi-Natural Experiment in China.” Journal of Business Ethics 180 (1): 125–43. https://doi.org/10.1007/s10551-021-04886-1.Search in Google Scholar
Cheng, Rui, Bart Frijns, Hyeongjun Kim, and Doojin Ryu. 2024a. “Effects of Option Incentive Compensation on Corporate Innovation: The Case of China.” Economic Systems 48 (1): 101171. https://doi.org/10.1016/j.ecosys.2023.101171.Search in Google Scholar
Cheng, Rui, Hyeongjun Kim, and Doojin Ryu. 2024b. “ESG Performance and Firm Value in the Chinese Market.” Investment Analysts Journal 53 (1): 1–15. https://doi.org/10.1080/10293523.2023.2218124.Search in Google Scholar
Choi, Seo Yun, Doojin Ryu, and Wonhee You. 2024. “ESG Activities and Financial Stability: The Case of Korean Financial Firms.” Borsa Istanbul Review 24 (5): 945–51. https://doi.org/10.1016/j.bir.2024.05.007.Search in Google Scholar
Chuang, Shun-Pin, and Sun-Jen Huang. 2018. “The Effect of Environmental Corporate Social Responsibility on Environmental Performance and Business Competitiveness: The Mediation of Green Information Technology Capital.” Journal of Business Ethics 150: 991–1009. https://doi.org/10.1007/s10551-016-3167-x.Search in Google Scholar
Dechezleprêtre, Antoine, Daniel Nachtigall, and Frank Venmans. 2023. “The Joint Impact of the European Union Emissions Trading System on Carbon Emissions and Economic Performance.” Journal of Environmental Economics and Management 118: 102758. https://doi.org/10.1016/j.jeem.2022.102758.Search in Google Scholar
Dissanayake, Sumali, Renuka Mahadevan, and John Asafu-Adjaye. 2020. “Evaluating the Efficiency of Carbon Emissions Policies in a Large Emitting Developing Country.” Energy Policy 136: 111080. https://doi.org/10.1016/j.enpol.2019.111080.Search in Google Scholar
Doda, Baran, Caterina Gennaioli, Andy Gouldson, David Grover, and Rory Sullivan. 2016. “Are Corporate Carbon Management Practices Reducing Corporate Carbon Emissions?” Corporate Social Responsibility and Environmental Management 23 (5): 257–70. https://doi.org/10.1002/csr.1369.Search in Google Scholar
Erhemjamts, Otgontsetseg, Kershen Huang, and Hassan Tehranian. 2024. “Climate Risk, ESG Performance, and ESG Sentiment in US Commercial Banks.” Global Finance Journal 59: 100924. https://doi.org/10.1016/j.gfj.2023.100924.Search in Google Scholar
Feng, Langang, Wenli Yang, Jin Hu, Kaiya Wu, and Huiyang Li. 2025. “Exploring the Nexus Between Rural Economic Digitalization and Agricultural Carbon Emissions: A Multi-Scale Analysis Across 1607 Counties in China.” Journal of Environmental Management 373: 123497. https://doi.org/10.1016/j.jenvman.2024.123497.Search in Google Scholar
Fikru, Mahelet G., Fateh Belaïd, and Hongyan Ma. 2024. “Carbon Capture and Renewable Energy Policies: Could Policy Harmonization be a Puzzle Piece to Solve the Electricity Crisis?” Energy Economics 136: 107753. https://doi.org/10.1016/j.eneco.2024.107753.Search in Google Scholar
Flammer, Caroline. 2021. “Corporate Green Bonds.” Journal of Financial Economics 142 (2): 499–516. https://doi.org/10.1016/j.jfineco.2021.01.010.Search in Google Scholar
Gao, Yihong. 2022. “Green Credit Policy and Trade Credit: Evidence from a Quasi-Natural Experiment.” Finance Research Letters 50: 103301. https://doi.org/10.1016/j.frl.2022.103301.Search in Google Scholar
Ghosh, Debarati, and Meghna Dutta. 2022. “Environmental Behaviour Under Credit Constraints–Evidence from Panel of Indian Manufacturing Firms.” Structural Change and Economic Dynamics 63: 490–500. https://doi.org/10.1016/j.strueco.2022.07.004.Search in Google Scholar
Goodell, John W., Constantin Gurdgiev, Sitara Karim, and Alessia Palma. 2024. “Carbon Emissions and Liquidity Management.” International Review of Financial Analysis 95 (A): 103367. https://doi.org/10.1016/j.irfa.2024.103367.Search in Google Scholar
Hadlock, Charles J., and Joshua R. Pierce. 2010. “New Evidence on Measuring Financial Constraints: Moving Beyond the KZ Index.” Review of Financial Studies 23 (5): 1909–40. https://doi.org/10.1093/rfs/hhq009.Search in Google Scholar
Hang, Markus, Jerome Geyer‐Klingeberg, and Andreas W. Rathgeber. 2019. “It is Merely a Matter of Time: A Meta-Analysis of the Causality Between Environmental Performance and Financial Performance.” Business Strategy and the Environment 28 (2): 257–73. https://doi.org/10.1002/bse.2215.Search in Google Scholar
Haque, Faizul. 2017. “The Effects of Board Characteristics and Sustainable Compensation Policy on Carbon Performance of UK Firms.” British Accounting Review 49 (3): 347–64. https://doi.org/10.1016/j.bar.2017.01.001.Search in Google Scholar
Hartono, Djoni, Witri Indriyani, Beta Septi Iryani, Ahmad Komarulzaman, Anda Nugroho, and Robi Kurniawan. 2023. “Carbon Tax, Energy Policy, and Sustainable Development in Indonesia.” Sustainable Development 31 (4): 2332–46. https://doi.org/10.1002/sd.2511.Search in Google Scholar
Hasan, Iftekhar, Suk-Joong Kim, Panagiotis N. Politsidis, and Eliza Wu. 2023. “Regulation and Information Costs of Sovereign Distress: Evidence from Corporate Lending Markets.” Journal of Corporate Finance 82: 102468. https://doi.org/10.1016/j.jcorpfin.2023.102468.Search in Google Scholar
He, Guosheng, Yiting Liu, and Fu Chen. 2023. “Research on the Impact of Environment, Society, and Governance (ESG) on Firm Risk: An Explanation from a Financing Constraints Perspective.” Finance Research Letters 58 (A): 104038. https://doi.org/10.1016/j.frl.2023.104038.Search in Google Scholar
Hennig, Jan C., Sebastian Firk, Michael Wolff, and Hülgen Coskun. 2023. “Environmental Management Control Systems: Exploring the Economic Motivation Behind their Implementation.” Journal of Business Research 169: 114283. https://doi.org/10.1016/j.jbusres.2023.114283.Search in Google Scholar
Homroy, Swarnodeep. 2023. “GHG Emissions and Firm Performance: The Role of CEO Gender Socialization.” Journal of Banking & Finance 148: 106721. https://doi.org/10.1016/j.jbankfin.2022.106721.Search in Google Scholar
Jia, Haiying, and Paulo Azevedo. 2025. “Sustainable Investing and Ownership Structure in International Shipping Companies.” Maritime Policy & Management 52 (5): 691–708. https://doi.org/10.1080/03088839.2024.2402790.Search in Google Scholar
Johl, Satirenjit Kaur, Kashif Ali, Kunio Shirahada, and Oyelakin Idris Oyewale. 2024. “Green Servitization, Circular Economy, and Sustainability: A Winning Combination Analysis Through Hybrid SEM-ANN Approach.” Business Strategy and the Environment 33 (8): 8978–93. https://doi.org/10.1002/bse.3950.Search in Google Scholar
Kang, Sungchang, Jeongseok Bang, and Doojin Ryu. 2024. “Female CEOs’ Risk Management and Earnings Performance During the Financial Crisis.” Asian Business & Management 23 (1): 110–38. https://doi.org/10.1057/s41291-023-00233-6.Search in Google Scholar
Khan, Muhammad Arif, and Silvio Vismara. 2025. “Green Bond Issuance and Corporate Environmental and Financial Performance: A Meta-Analysis.” International Review of Economics & Finance 102: 104313. https://doi.org/10.1016/j.iref.2025.104313.Search in Google Scholar
Khan, Sher Jahan, Puneet Kaur, Fauzia Jabeen, and Amandeep Dhir. 2021. “Green Process Innovation: Where We are and Where We are Going.” Business Strategy and the Environment 30 (7): 3273–96. https://doi.org/10.1002/bse.2802.Search in Google Scholar
Kim, Hakkon, Kwangwoo Park, and Doojin Ryu. 2017. “Corporate Environmental Responsibility: A Legal Origins Perspective.” Journal of Business Ethics 140: 381–402. https://doi.org/10.1007/s10551-015-2641-1.Search in Google Scholar
Kirat, Yassine, Tina Prodromou, and Sandy Suardi. 2024. “Unveiling the Nexus: Climate Change, Green Innovation, and the Pendulum of Energy Consumption and Carbon Emissions.” Energy Economics 138: 107727. https://doi.org/10.1016/j.eneco.2024.107727.Search in Google Scholar
Li, Guangpei, Xiaoyu Wang, Shibin Su, and Yuan Su. 2019. “How Green Technological Innovation Ability Influences Enterprise Competitiveness.” Technology in Society 59: 101136. https://doi.org/10.1016/j.techsoc.2019.04.012.Search in Google Scholar
Lin, Boqiang, and Aoxiang Zhang. 2023. “Can Government Environmental Regulation Promote Low-Carbon Development in Heavy Polluting Industries? Evidence from China’s New Environmental Protection Law.” Environmental Impact Assessment Review 99: 106991. https://doi.org/10.1016/j.eiar.2022.106991.Search in Google Scholar
Li, Haichao, Yuqi Su, Chante Jian Ding, Gary Gang Tian, and Zhan Wu. 2024. “Unveiling the Green Innovation Paradox: Exploring the Impact of Carbon Emission Reduction on Corporate Green Technology Innovation.” Technological Forecasting and Social Change 207: 123562. https://doi.org/10.1016/j.techfore.2024.123562.Search in Google Scholar
Liu, Cenjie, Jiayu Fang, and Rui Xie. 2021. “Energy Policy and Corporate Financial Performance: Evidence from China’s 11th Five-Year Plan.” Energy Economics 93: 105030. https://doi.org/10.1016/j.eneco.2020.105030.Search in Google Scholar
Liu, Cheng, Hemachandra Padhan, Vijaya Rekha, and Giray Gozgor. 2025. “The Path to Green Growth in OECD Economies: The Role of Energy Transition and Education.” Energy Economics 145: 108419. https://doi.org/10.1016/j.eneco.2025.108419.Search in Google Scholar
López Pérez, Germán, Isabel María García Sánchez, and José Luis Zafra Gómez. 2024. “A Systematic Literature Review and Bibliometric Analysis of Eco‐Innovation on Financial Performance: Identifying Barriers and Drivers.” Business Strategy and the Environment 33 (2): 1321–40. https://doi.org/10.1002/bse.3550.Search in Google Scholar
Luo, Le, and Qingliang Tang. 2014. “Does Voluntary Carbon Disclosure Reflect Underlying Carbon Performance?” Journal of Contemporary Accounting & Economics 10 (3): 191–205. https://doi.org/10.1016/j.jcae.2014.08.003.Search in Google Scholar
Mardones, Cristian, and Camilo Mena. 2020. “Effects of the Internalization of the Social Cost of Global and Local Air Pollutants in Chile.” Energy Policy 147: 111875. https://doi.org/10.1016/j.enpol.2020.111875.Search in Google Scholar
Nam, Hyun-Jung, and Doojin Ryu. 2024. “Impacts of Trade and Institutional Quality on Carbon Emissions in Transition Economies.” Finance Research Letters 67 (B): 105872. https://doi.org/10.1016/j.frl.2024.105872.Search in Google Scholar
Nam, Hyun-Jung, Mehmet Huseyin Bilgin, and Doojin Ryu. 2024. “Firm Value, Ownership Structure, and Strategic Approaches to ESG Activities.” Eurasian Business Review 14 (1): 187–226. https://doi.org/10.1007/s40821-024-00252-z.Search in Google Scholar
Nam, Hyun-Jung, Doojin Ryu, and Peter G. Szilagyi. 2025a. “Gender Inequality, Institutional Quality and Economic Outcomes in the European Union.” European Financial Management 31 (1): 463–92. https://doi.org/10.1111/eufm.12508.Search in Google Scholar
Nam, Hyun-Jung, Jonathan A. Batten, and Doojin Ryu. 2025b. “Do Institutional Quality and Trade Openness Enhance the Role of Financial Openness in Eastern European Financial Development?” Global Finance Journal 64: 101071. https://doi.org/10.1016/j.gfj.2024.101071.Search in Google Scholar
Newton, David P., Steven Ongena, Ru Xie, and Binru Zhao. 2024. “Firm ESG Reputation Risk and Debt Choice.” European Financial Management 30 (4): 2071–94. https://doi.org/10.1111/eufm.12468.Search in Google Scholar
Oak, Hena, and Sangeeta Bansal. 2022. “Enhancing Energy Efficiency of Indian Industries: Effectiveness of PAT Scheme.” Energy Economics 113: 106220. https://doi.org/10.1016/j.eneco.2022.106220.Search in Google Scholar
Olden, Andreas, and Jarle Møen. 2022. “The Triple Difference Estimator.” Econometrics Journal 25 (3): 531–53. https://doi.org/10.1093/ectj/utac010.Search in Google Scholar
Shahin, Vahideh, Moslem Alimohammadlou, and Abbas Abbasi. 2024. “Identifying and Prioritizing the Barriers to Green Innovation in SMEs and the Strategies to Counteract the Barriers: An Interval-Valued Intuitionistic Fuzzy Approach.” Technological Forecasting and Social Change 204: 123408. https://doi.org/10.1016/j.techfore.2024.123408.Search in Google Scholar
Song, Danbee, Jong Hoon Shin, and Abdoul G. Sam. 2026. “Corporate Environmentalism and Economic Performance: Examining the Effects of ISO 14001 Certification on Technical Efficiency.” Journal of Environmental Planning and Management 69 (2): 404–24. https://doi.org/10.1080/09640568.2024.2371567.Search in Google Scholar
Wang, Juan, Mingming Hu, Arnold Tukker, and João F. D. Rodrigues. 2019. “The Impact of Regional Convergence in Energy-Intensive Industries on China’s CO2 Emissions and Emission Goals.” Energy Economics 80: 512–23. https://doi.org/10.1016/j.eneco.2019.01.024.Search in Google Scholar
Wang, Juan, Jing Li, and Qingjun Zhang. 2021. “Does Carbon Efficiency Improve Financial Performance? Evidence from Chinese Firms.” Energy Economics 104: 105658. https://doi.org/10.1016/j.eneco.2021.105658.Search in Google Scholar
Wang, Jinfeng, Lei Zhu, Lijie Feng, and Jian Feng. 2023. “A Meta-Analysis of Sustainable Supply Chain Management and Firm Performance: Some New Findings on Sustainable Supply Chain Management.” Sustainable Production and Consumption 38: 312–30. https://doi.org/10.1016/j.spc.2023.04.015.Search in Google Scholar
Weiss, Jan, Andreas Stephan, and Tatiana Anisimova. 2019. “Well-Designed Environmental Regulation and Firm Performance: Swedish Evidence on the Porter Hypothesis and the Effect of Regulatory Time Strategies.” Journal of Environmental Planning and Management 62 (2): 342–63. https://doi.org/10.1080/09640568.2017.1419940.Search in Google Scholar
Xi, Bin, and Weifang Jia. 2025. “Research on the Impact of Carbon Trading on Enterprises’ Green Technology Innovation.” Energy Policy 197: 114436. https://doi.org/10.1016/j.enpol.2024.114436.Search in Google Scholar
Xu, Jie, Fu Chen, Weiwei Zhang, Yiting Liu, and Tingting Li. 2023. “Analysis of the Carbon Emission Reduction Effect of Fintech and the Transmission Channel of Green Finance.” Finance Research Letters 56: 104127. https://doi.org/10.1016/j.frl.2023.104127.Search in Google Scholar
Yang, Shengyi. 2023. “Carbon Emission Trading Policy and Firm’s Environmental Investment.” Finance Research Letters 54: 103695. https://doi.org/10.1016/j.frl.2023.103695.Search in Google Scholar
Yang, Jun, Yun Hao, and Chao Feng. 2021. “A Race Between Economic Growth and Carbon Emissions: What Play Important Roles Towards Global Low-Carbon Development?” Energy Economics 100: 105327. https://doi.org/10.1016/j.eneco.2021.105327.Search in Google Scholar
Yang, Eunmo, Hojoong Bae, and Doojin Ryu. 2024. “Air Pollution and Daily Public Transportation Ridership: The Case of Seoul City.” Energy & Environment 35 (8): 3940–60. https://doi.org/10.1177/0958305X231174011.Search in Google Scholar
You, Wonhee, Jens Müller-Merbach, and Doojin Ryu. 2024. “Corporate Social Responsibility and Firm Performance: Evidence from Korea.” Investment Analysts Journal 54 (2): 234–52. https://doi.org/10.1080/10293523.2024.2412451.Search in Google Scholar
Yu, Chin-Hsien, Xiuqin Wu, Dayong Zhang, Shi Chen, and Jinsong Zhao. 2021. “Demand for Green Finance: Resolving Financing Constraints on Green Innovation in China.” Energy Policy 153: 112255. https://doi.org/10.1016/j.enpol.2021.112255.Search in Google Scholar
Zhang, Kun, Zong-Yong Zhang, and Qiao-Mei Liang. 2017. “An Empirical Analysis of the Green Paradox in China: From the Perspective of Fiscal Decentralization.” Energy Policy 103: 203–11. https://doi.org/10.1016/j.enpol.2017.01.023.Search in Google Scholar
Zhang, Rongwu, Wenqiang Fu, and Tong Lu. 2023. “Capital Market Opening and Corporate Environmental Performance: Empirical Evidence from China.” Finance Research Letters 53: 103587. https://doi.org/10.1016/j.frl.2022.103587.Search in Google Scholar
Supplementary Material
This article contains supplementary material (https://doi.org/10.1515/econ-2025-0189).
© 2026 the author(s), published by De Gruyter, Berlin/Boston
This work is licensed under the Creative Commons Attribution 4.0 International License.
Articles in the same Issue
- Review Articles
- Professional Development of ESP Teachers for Digital Transformation: Approaches, Challenges, and Policy Implications
- How Does the Adoption of Generative AI Affect the Speed of Strategic Decision-Making and Innovation Performance in SMEs?
- Research Articles
- Family vs. Non-Family Firms: Linking Future Environmental Expenditures to Firm Value
- The Dynamic Association of Economic Policy Uncertainty, Short-Term Cross-Border Capital Flows and Systemic Financial Risk
- On- and Off-Chain Demand and Supply Drivers of Bitcoin Price
- Internet Use and Child Malnutrition: Evidence from Rural Areas in China
- Public Services, Urban Innovation and Green Total Factor Productivity
- Convergence and Paths of Economic Growth
- Digital Financial Inclusion and Multidimensional Poverty Alleviation: Empowering Young Migrant Workers in China
- The Impact of Digital Economy on Urban-Rural Integrated Development: Taking the 19 City Clusters in China as Examples
- Digital Economy and Urban Green Development: Regional Heterogeneity and Empirical Evidence from Chinese Cities
- “Kazakhstan’s Dairy Sector: Greening the Economy Through Innovation and Clean Production”
- How Do Stringent Climate Policies Influence Carbon Intensity and Profitability? Insights from the Chinese Market
- Special Issue: Shapes of Performance Evaluation – 2nd Edition
- Coupling Tourism Carbon Emission Efficiency and Economic Resilience in China’s Yellow River Basin
Articles in the same Issue
- Review Articles
- Professional Development of ESP Teachers for Digital Transformation: Approaches, Challenges, and Policy Implications
- How Does the Adoption of Generative AI Affect the Speed of Strategic Decision-Making and Innovation Performance in SMEs?
- Research Articles
- Family vs. Non-Family Firms: Linking Future Environmental Expenditures to Firm Value
- The Dynamic Association of Economic Policy Uncertainty, Short-Term Cross-Border Capital Flows and Systemic Financial Risk
- On- and Off-Chain Demand and Supply Drivers of Bitcoin Price
- Internet Use and Child Malnutrition: Evidence from Rural Areas in China
- Public Services, Urban Innovation and Green Total Factor Productivity
- Convergence and Paths of Economic Growth
- Digital Financial Inclusion and Multidimensional Poverty Alleviation: Empowering Young Migrant Workers in China
- The Impact of Digital Economy on Urban-Rural Integrated Development: Taking the 19 City Clusters in China as Examples
- Digital Economy and Urban Green Development: Regional Heterogeneity and Empirical Evidence from Chinese Cities
- “Kazakhstan’s Dairy Sector: Greening the Economy Through Innovation and Clean Production”
- How Do Stringent Climate Policies Influence Carbon Intensity and Profitability? Insights from the Chinese Market
- Special Issue: Shapes of Performance Evaluation – 2nd Edition
- Coupling Tourism Carbon Emission Efficiency and Economic Resilience in China’s Yellow River Basin
