How does voluntary environmental regulation affect firm export performance? Evidence from the green factory identification in China

The manufacturing sector plays a crucial role in driving global economic growth. However, its deep reliance on high carbon emissions and intensive resource consumption renders it a structural cause of the climate crisis (Liu et al., 2022). The Sustainable Development Solutions Network (SDSN) releases the Sustainable Development Report 2025, which shows that fewer than 20% of the SDG targets are on track to be achieved by 2030, while the climate crisis continues to worsen. With this background, green manufacturing has garnered growing attention from governments and enterprises worldwide for its potential to reconcile environmental sustainability with economic growth (D’Angelo et al., 2023). It increasingly emerges as a critical pathway for advancing low-carbon transformation in the global manufacturing industry (He et al., 2024).

The balance between trade and the environment has long been a key concern for countries worldwide. As a leading global manufacturing power, China has advanced its industrial green transition, yet entrenched high-pollution, high-emission, high-energy-consumption practices persist, resulting in chronic resource inefficiency and environmental degradation (Shen and Zhang, 2023). At the same time, with the accelerated revolution of global green trade rules, such as the launch of the EU’s Carbon Border Adjustment Mechanism, China’s manufacturing industry is facing increasingly fierce challenges from “green barriers” in the international market (Lin and Zhao, 2023). To this end, the Chinese government actively promotes the development of a green manufacturing system as a strategic lever to strengthen its green and low-carbon international competitiveness (Li, 2019). In 2016, the Ministry of Industry and Information Technology (MIIT) launched the Green Factory Identification (GFI), a typical form of voluntary environmental regulation. The GFI encourages firms to proactively commit to environmental improvements by promoting cleaner production, low-carbon energy use and more efficient resource utilization (Chen et al., 2025).

As key actors in green transformation, firms play an increasingly important role in GFI practices. In response to GFI policies, firms adjust their production strategies to meet environmental regulatory requirements while maximizing profitability (Wei et al., 2024). Such policy-induced resource reallocation promotes energy conservation, cleaner production and carbon reduction (Hu et al., 2025). It also signals superior environmental performance, easing financing constraints and promoting green innovation through investment and Environmental, Social and Governance (ESG) improvement (Zeng et al., 2024). However, under conditions of sluggish global recovery, rising protectionism and stricter green trade barriers, doubts remain as to whether GFI can translate into sustained export gains. For example, weak enforcement may lead some firms to pursue symbolic compliance or greenwashing, while poor governance or inadequate incentives risk undermining policy effectiveness (Flankova et al., 2024). Moreover, high fixed costs and transformation risks may impose resource misallocation or operational burdens that constrain export capacity (Acquah et al., 2021). Therefore, this paper aims to explore whether GFI enhances firm export performance (FEP) and to identify the underlying mechanisms.

Leveraging panel data from Chinese A-share listed manufacturing firms from 2008 to 2023, this study examines the implementation of GFI impact on FEP and its transmission channels. Specifically, we use a quasi-natural experiment design with a multi-period difference-in-differences (multi-period DID) model to identify causal effects. Results indicate that GFI implementation significantly enhances FEP. Mechanistically, the policy operates through three channels: fostering technological innovation, optimizing resource allocation and enhancing corporate reputation. Furthermore, common institutional ownership has a positive moderating effect. Heterogeneity analysis indicates that the effect of the GFI policy on FEP is markedly stronger for domestic enterprises, nonheavily polluting industries and capital-intensive industries. Extended analysis confirms significant spillover effects at the industry level.

This study makes three main contributions to the literature. First, existing studies primarily focus on command-and-control or market-based environmental regulations in examining the impact of green transition on firm exports. This study is the first to investigate the GFI as a uniquely Chinese form of voluntary environmental regulation, empirically demonstrating that the GFI significantly promotes FEP. This evidence not only provides new empirical support for research on the relationship between environmental regulation and exports but also contributes to the ongoing debate on whether such regulation promotes or hinders export competitiveness. Second, drawing on Porter Hypothesis (Porter and Linde, 1995), Stakeholder Theory (Freeman, 2010) and Signaling Theory (Spence, 1973), this study identifies three key pathways: technological innovation, resource allocation and corporate reputation, through which GFI improves FEP. These results underscore the heterogeneous effects of policy across governance environments, thereby deepening the understanding of the interaction between corporate governance and green strategies. Third, this study provides important practical implications for policymakers, particularly in developing countries seeking to balance environmental protection with export competitiveness. This study reveals the significant moderating role of common institutional ownership, demonstrating that green transformation driven by voluntary environmental regulation generates environmental spillover effects at the industry level. These findings offer direct evidence of the long-term benefits of firms’ voluntary engagement in green manufacturing transformation.

The remainder of this paper is organized as follows. Section 2 presents the policy context and literature review. Section 3 includes a theoretical hypothesis. Section 4 describes the data, model specification and variable definitions. Section 5 reports the empirical findings. Section 6 offers additional analyses. Finally, Section 7 concludes and discusses policy implications.

In recent years, the Chinese government has prioritized the advancement of green manufacturing, accelerating strategic planning through comprehensive policy frameworks that provide robust support for corporate green transformation. According to the announcement released by the MIIT on its official website, after compiling the data, by December 2023, China had a total of 5145 green factories. Figure 1 displays the number of green factories certified in batches 1–8 from 2017 to 2023, along with the year-on-year growth rates. This figure shows the number of green factories on the left y-axis and the year-on-year growth rate on the right y-axis. Figure 2 reports the total number of green factories certified from the first to the eighth batch, categorized by provincial-level administrative divisions.

Environmental regulation is a key policy instrument for mitigating the negative environmental externalities of economic growth. Based on policy tools, environmental regulation can be classified into command-and-control, market-based and voluntary environmental regulation (Ren et al., 2018; Lu et al., 2025). Existing studies on the impact of environmental regulation on firm export yield two main conclusions. On the one hand, environmental regulation drives the upgrading of export product quality through the “Porter effect” (Porter and Linde, 1995; Jiang et al., 2023), while also boosting firms’ export scale by enhancing total factor productivity (Chen et al., 2022). It also increases the domestic value-added in exports via resource allocation effects (Sun et al., 2023). On the other hand, some studies show that stringent environmental regulations raise production costs and narrow profit margins, thereby weakening firms’ international competitiveness and significantly suppressing exports (Shi and Xu, 2018; Zhang et al., 2020).

Voluntary environmental regulation lies between command-and-control and market-based policies, emphasizing firms’ proactive provision of public environmental goods in line with their own circumstances. Firms may adopt such regulation in response to market or government pressure (Aragòn-Correa et al., 2020). GFI represents a typical form of voluntary environmental regulation, and a small but growing body of research investigates its economic effects. As a key dimension of the green transition, Liu et al. (2025) show that GFI stimulates green technological innovation, thereby making green dividends more inclusive. Recognizing ESG as a critical driver of corporate sustainability, Zeng et al. (2024) find that GFI enhances firms’ ESG performance. From the perspective of corporate debt, Chen et al. (2025) demonstrate that GFI strengthens environmental reputation, reduces risk perception and lowers debt costs. Furthermore, Wei et al. (2024) examine whether GFI benefits workers and find that while it increases labor productivity, it suppresses the labor income share.

A review of the existing literature reveals that research on environmental regulation and FEP is relatively extensive. However, few studies integrate GFI and FEP within a unified analytical framework or investigate the moderating role of common institutional ownership in their causal relationship.

As a key element of green manufacturing, GFI demonstrates three distinctive features that set it apart from earlier green evaluation programs. First, its rigorous accreditation standards not only raise entry thresholds but also effectively identify firms with both the motivation and capability for green innovation. Second, GFI offers broad subsidies and incentive policies that enhance firms’ competitive advantages. Third, consistent with Signaling Theory, GFI enables firms to project an environmentally responsible image, thereby improving the quality of environmental information disclosure and strengthening corporate reputation. More importantly, the green productivity generated through this transformation allows firms to bypass green trade barriers while reinforcing technological leadership and brand value, ultimately enhancing FEP (Huang, 2023; Lu et al., 2025). Based on this analysis, this study proposes the following hypothesis:

According to the Porter Hypothesis, flexible environmental regulation policies can stimulate firms to engage in technological innovation, thereby reducing compliance costs and improving export competitiveness (Ramanathan et al., 2017; Lin and Chen, 2020). As a representative form of voluntary environmental regulation, GFI exemplifies this innovation mechanism. Specifically, it enforces strict environmental performance standards, compelling firms to engage in continuous technological upgrading (Hu et al., 2025). Moreover, the construction of GFI requires the adoption of eco-friendly materials and processes, while encouraging firms to introduce advanced green technologies. Such sustained technological upgrading fosters innovation capacity and enables firms to better align with international technical and environmental standards, thereby enhancing long-term export competitiveness. Therefore, we propose the following hypothesis:

Firms certified as green factories not only engage actively in green transformation but also benefit from preferential access to policy resources (Chen et al., 2025). These include financial subsidies, green credit and tax incentives (Wei et al., 2024). At the same time, GFI creates positive linkages with capital markets, enabling firms to secure dedicated green financing from financial institutions (Zeng et al., 2024). These funds are used to support carbon-reduction projects, resource-efficient production and digital upgrades (Hu et al., 2025). Such incentive mechanisms alleviate the financial constraints associated with green transformation, optimize the financing environment and provide vital support for capacity expansion and technological development. This improves firms’ ability to compete in international markets. Therefore, we propose the following hypothesis:

Corporate reputation is a multidimensional construct that includes corporate image, social responsibility and business ethics (Agarwal et al., 2018). From the perspective of internal incentives, GFI acts as a “green signal” of superior environmental performance (Wei et al., 2024). It reduces the perceived risk of Chinese exports in overseas markets and builds trust in both the firm and its products. This trust generates export premiums, enhances recognition in high-end markets and strengthens export competitiveness (Xu et al., 2018). From the perspective of external supervision, in the information age, the media serves as a key channel for transmitting market information and a major force driving firms to adopt eco-innovation and improve their environmental impact (Liu et al., 2025). Under such reputational pressure, firms improve product quality and upgrade services to maintain market standing, which further enhances export performance (Nguyen et al., 2023). Therefore, we propose the following hypothesis:

Against the backdrop of global sustainable development, common institutional ownership is increasingly paying attention to corporate environmental performance (Wu et al., 2025). First, through the information transmission mechanism, institutional investors typically possess strong information-gathering and analytical capabilities, thereby reducing information asymmetry (Hu and Yang, 2025). When institutional investors increase their holdings in firms with green factory certifications, other market participants may interpret this as an endorsement of the firm’s green transformation and sustainable development strategy. Second, through the risk governance mechanism, institutional investors act as external monitors within corporate governance structures. Their presence incentivizes firms to prioritize environmental compliance, discouraging short-term profit-seeking behavior. This governance pressure drives firms to invest in green technological innovation (Shi and Wang, 2024), thereby strengthening their long-term competitiveness in green manufacturing. Therefore, we propose the following hypothesis:

Based on the above theoretical analysis and hypothesis, Figure 3 presents the research framework and the impact mechanism of the GFI on FEP.

To guarantee the comparability of the research sample, this study focuses on all A-share listed manufacturing firms in China from 2008 to 2023, as green factories are exclusively distributed within the manufacturing sector. The sample is refined based on the following criteria:

• firms in the financial and real estate sectors are excluded;

• firms under special treatment (ST, *ST and PT) due to abnormal trading status are removed;

• firms listed for less than one year and firms with negative net assets are excluded; and

• all key continuous variables are winsorized at the 1% level to reduce the impact of outliers.

The data for this study are mainly from CNRDS, CSMAR and Wind databases.

On the basis of the theoretical analysis, this paper uses GFI as a quasi-natural experiment and applies the multi-period DID to identify the net effect of the policy on FEP. We analyze a DID model:

In the model, the index i represents the individual listed firms, while the index t represents the year. The dependent variable, Export_it represents FEP. We use Ene_it and Osa_it as two complementary measures of Export_it. Ene_it represents the export propensity of listed firm i in year t, and Osa_it represents the export size of listed firm i in year t. Gplant_it is a dummy variable for whether the sample firm is selected for the green factory demonstration list. Control_it represents the vector of the control variables. μ_i denotes individual fixed effects, θ_t denotes year-fixed effects and ε_it is a random perturbation term.

Based on the theoretical discussion of the moderating effect in the previous section, to verify the moderating effect of common institutional ownership in a more in-depth way, this paper constructs a moderating effect model:

where Coz_it represents common institutional ownership, and the coefficient β₃ of the interaction term between GFI and common institutional ownership is the key coefficient in this paper. If the coefficient β₃ is significant and has the same sign as the coefficient β₁, it indicates that there is a positive moderating effect. The other variables and parameter descriptions in the model are the same as in equation (1).

The dependent variable is Export. According to the new-new trade theory, corporate export growth can be decomposed into the extensive margin and the intensive margin, jointly referred to as the dual margins of export performance (Melitz, 2003). This study constructs two measures of export performance using data on listed firms. First, Ene is a binary variable indicating the firm’s export status; it equals 1 if the firm reports any overseas sales revenue in a given year, and 0 otherwise. Second, Osa measures the scale of exports, defined as the natural logarithm of overseas sales revenue plus one. The core explanatory variable is Gplant, which indicates whether a firm is certified as a green factory. According to the MIIT, green factory demonstration lists are released in eight separate batches between 2017 and 2023, reflecting heterogeneity in certification timing. To capture the policy’s dynamic effect, Gplant is defined as follows: if listed firm i is first included in any batch of the green factory list in year t, the variable takes the value of 1 from year t onward; otherwise, it is 0. A range of factors that may affect Export should also be controlled. This study follows the approach of Hu et al. (2025) and incorporates several firm-level characteristics as control variables. These include firm age (Age), profitability (ROA), leverage (Lev), capital intensity (Capital), Tobin’s Q value (TobinQ), ownership concentration (Mshare), board independence (Indep) and CEO duality (Dual), as detailed in Table 1.

The descriptive statistics presented in Table 2 reveal that the mean values of Ene and Osa are 0.6179 and 12.0305, respectively, with minimum values of 0 and maximum values of 1 and 24.3289. These results indicate considerable heterogeneity in FEP, providing a suitable empirical context for this study. The core explanatory variable, Gplant, has a mean value of 0.1207, indicating that 12.07% of the observations correspond to firms that have been selected for the green factory demonstration program.

Table 3 reports the regression results of equation (1). Columns (1)–(2) include only the core explanatory variable GPlant, without any control variables or firm-fixed and year-fixed effects from the model. Columns (3)–(4) add firm-fixed and year-fixed effects to the baseline specification, and columns (5)–(6) further incorporate control variables. Across all specifications, the coefficient of GPlant is significantly positive at least at the 5% level, indicating that participation in GFI significantly enhances FEP. Taking columns (5) and (6) as examples, the regression coefficient of GPlant on the firm’s export propensity (Ene) is 0.0229, and its coefficient on the export scale (Osa) is 0.7795, both significant at the 1% level. Judging from the magnitude of the coefficients, GFI has a stronger positive effect on firms’ export scale than on their export propensity.

To ensure the validity of the DID estimation results, the common trend assumption must be satisfied. Following the approach of Jacobson et al. (1993), this study constructs equation (3) to conduct a common trend test, where k denotes the k-th year, and the remaining parameters are consistent with those in equation (1):

Taking 2016 as the base period for the construction of green factories, and considering that the data for the first seven years and the last four years of the implementation of the policy are relatively small, the constructed equation (3) is used to conduct a common trend test of “the first seven years and the last four years”, and the results are shown in Figure 4. Before the implementation of GFI, there was no significant difference in the FEP between the experimental and control groups, proving that the study sample satisfied the common trend hypothesis. After the policy took effect, a marked increase in FEP was observed in the implementation year. Two years following the launch of GFI, both the magnitude and statistical significance of the coefficients increased, indicating a cumulative effect of the policy over time. As shown in Figure 4(a), in the common trend test with firm’s propensity to export as the dependent variable, the coefficient in the fourth period becomes statistically insignificant, possibly suggesting that the positive impact of the green factory initiative on firm’s export propensity begins to taper off beyond the fourth year. As shown in Figure 4(b), in the common trend test with firm’s export size as the dependent variable, the coefficient of the core explanatory variable remains significantly positive following the implementation of the GFI policy, peaking in the second period. This suggests the policy exerts a persistent and strengthening effect on firm’s export size.

To ensure that the observed effects of GFI on FEP are not the result of random factors, this study conducts a placebo test to identify the potential endogeneity of the policy intervention. We generate 500 pseudo-policy treatment variables through random sampling. Equation (1) is then re-estimated using these pseudo-variables. The distribution of the estimated coefficients and p-values is illustrated in Figure 5. As shown in Figure 5(a), when using firm’s propensity to export as the dependent variable, the placebo-estimated coefficients follow an approximately normal distribution and most coefficients are smaller than the baseline estimate of 0.0229. As shown in Figure 5(b), when using firm’s export size as the dependent variable, the placebo-estimated coefficients follow an approximately normal distribution, and all coefficients are smaller than the baseline estimate of 0.7795. These findings indicate that the observed impact of GFI on FEP is not due to random chance, thus confirming the robustness of the main results.

The study adopts a two-stage least squares (2SLS) instrumental variable (IV) approach to deal with potential endogeneity further. We use the ventilation coefficient of cities (VGC) as an instrument, which captures a city’s capacity for atmospheric self-purification. Higher VGC promotes more effective dispersion and dilution of pollutants, thereby reducing environmental governance costs and improving environmental carrying capacity, ultimately raising the probability that firms achieve GFI certification. Following the methodology of Hering and Poncet (2014), we argue that VGC, determined by exogenous geographic and meteorological factors, is strongly correlated with the intensity of local environmental regulation, and hence with the likelihood of GFI, while remaining exogenous to FEP.

Table 4 presents the 2SLS regression results. In the first-stage regression, the coefficient of the instrumental variable VGC is significantly positive at the 1% level, indicating that cities with higher VGC are more likely to promote GFI, consistent with theoretical expectations. The second-stage results in columns (2)–(3) confirm that the coefficient of the Gplant remains significantly positive, reaffirming the main findings of this study. The first-stage F-statistic is 2265.88, far exceeding the conventional threshold of 10 (the critical value of 16.38 at the 10% significance level), thereby ruling out concerns of weak instrument bias. Furthermore, the Kleibergen-Paap rk LM statistic of 457.222 is significant at the 1% level. It strongly rejects the null hypothesis of underidentification and suggests that the instrument is highly relevant.

To address potential selection bias, this study uses a Propensity Score Matching-Difference-in-Differences (PSM-DID) approach. Specifically, we use nearest-neighbor and caliper matching, using the control variables from the baseline model as covariates. After matching, the balance tests confirm that there are no systematic differences between the treatment and control groups. Based on the matched samples, we re-estimate the policy effect of the GFI on FEP, and the regression results are reported in Table 5, the coefficients for Gplant are all significant. These findings are consistent with the baseline results in Table 3, suggesting that the main conclusions of this study are highly robust to various matching specifications.

During the implementation period of the green factory policy, other concurrent policy measures may potentially influence the benchmark regression results. For instance, in China, the Low-Carbon City policy (LCT), as a major local-level initiative to promote green development, encompasses urban planning, energy structure adjustment and public transportation policies. Second, the revised Environmental Protection Law (EPL), which came into effect in 2015, represents a critical piece of environmental legislation in China. Third, the Carbon Emissions Trading Market (CET) has emerged as a key measure to address carbon emissions. To mitigate the potential confounding effects of these policy interventions and improve the robustness of the estimation, we include dummy variables for the LCT, the EPL and the CET in the benchmark regression model. Columns (1)–(6) of Table 6 present the regression results after controlling for these policy dummies individually, while columns (7)–(8) show the empirical results when all three policies are included simultaneously. The results suggest that, once these additional policy variables have been taken into account, the coefficient of Gplant remains significantly positive, with only slight changes in magnitude.

From the perspective of firm characteristics, we conduct a heterogeneity analysis based on ownership type by dividing the sample into domestic and foreign-invested enterprises. As shown in Table 7, the empirical results indicate that GFI significantly improves the export performance of domestic enterprises, whereas its impact on foreign-invested enterprises is not statistically significant. A possible explanation is that foreign-invested enterprises are generally more accustomed to meeting international environmental standards. Many multinational corporations already implement rigorous sustainability strategies across their global operations. For instance, developed countries in Europe and North America impose stringent environmental regulations, prompting foreign-invested enterprises to establish comprehensive green production systems to access these markets.

Given the significant variation in pollution intensity and environmental regulatory stringency across industries, the impact of GFI may also differ by industry type. Currently, green factories are widely distributed across 15 manufacturing sectors, including heavily polluting industries such as chemicals, building materials, steel and paper. Following the classification method proposed by Ren et al. (2024), this study categorizes firms into heavily polluting and nonheavily polluting industries.

The empirical results, as reported in Table 8, show that GFI significantly promotes the FEP in nonheavily polluting industries. This divergence may be attributed to two main factors. First, heavily polluting industries typically rely on energy-intensive and high-emission production models. Transitioning to green manufacturing in these sectors requires substantial investments in environmental equipment upgrades, clean energy substitution and production process transformation. Although GFI may provide some level of policy support and market recognition, the high short-term transition costs can offset the potential export-enhancing effects. Second, firms in heavily polluting industries are often subject to stricter environmental regulations. These firms may already be compelled to implement environmental measures before certification, which diminishes the marginal incentive effect of GFI.

To explore this heterogeneity, this section classifies sample firms into capital-intensive and labor-intensive industries. The regression results presented in Table 9 indicate that GFI has a stronger positive impact on FEP in capital-intensive industries, whereas its effect on labor-intensive firms is relatively limited. This pattern may be explained by the following factors: capital-intensive industries tend to rely on advanced equipment and technological innovation, making them more responsive to technological upgrades and environmental policies. The GFI emphasizes clean production, energy efficiency and effective management, goals that align closely with the pursuit of productivity and technological advancement in capital-intensive sectors, thereby enhancing their export capacity. In contrast, labor-intensive industries’ comparative advantage lies primarily in labor cost rather than technological upgrading, the incentives and capacity to respond to green factory initiatives remain relatively weak, resulting in a smaller improvement in export performance.

To explore the underlying mechanisms through which GFI affects FEP, this study builds on the baseline regression model (equation (1)) and constructs a mechanism testing model (equation (4)) to empirically test the proposed hypotheses:

In the model, Channel_it represents the mediating variables that capture the potential transmission channels, including technological innovation effects, resource allocation effects and corporate reputation effects.

A firm’s innovation performance can be measured in several ways. We use the natural logarithm of one plus the total number of invention patents, utility model patents and design patents independently filed by listed firms as a proxy for innovation (Patent). Given that invention patents are generally considered the most indicative of originality among the three patent types, this study further uses the natural logarithm of one plus the number of invention patent applications (Invention1) and the natural logarithm of one plus the number of granted invention patents (Invention2) as additional measures of firm technological innovation. As shown in Table 10, the results indicate that GFI has a significantly positive effect on all three measures at the 1% significance level. These findings suggest that GFI significantly enhances firms’ innovation capabilities, particularly in terms of invention patent applications and grants. This indicates that green manufacturing policies are crucial for promoting technological upgrading and innovation and that technological innovation serves as an important pathway through which GFI improves FEP.

Based on the preceding theoretical framework, government and financial institution incentives play a crucial role in promoting firms’ transition toward green manufacturing. Therefore, this paper selects three proxy variables, namely, government subsidy intensity (government subsidy amount/firms’ total assets, Gov), bank credit (firms’ long and short-term borrowing/firms’ total assets, Bnk) and firms’ tax burden (firms’ income tax expense/business revenues, Tax), to examine whether GFI contributes to the FEP through the resource allocation effect (Hu et al., 2019; Deng et al., 2020).

The regression results are shown in columns (1)–(3) of Table 11, and the main findings are as follows. First, the ongoing implementation of GFI significantly increases both government subsidy intensity and bank credit at the 5% significance level. Second, GFI significantly reduces firms’ tax burdens at the 1% level, suggesting that it effectively eases fiscal pressure. In summary, green factory construction not only alleviates firms’ financial constraints through multiple incentive channels but also reduces their tax burdens, thereby promoting continuous progress toward green and high-quality development.

To examine whether GFI enhances FEP through a reputation channel, this study uses corporate reputation (Fame) and media attention (News) as proxy variables for empirical analysis. Specifically, following Guan and Zhang (2019), the Fame score is derived using factor analysis based on stakeholder evaluations, incorporating 12 indicator variables. The resulting scores are scaled from 1 to 10, with higher values indicating stronger reputations. In addition, drawing on the methodology of Wang and Wan (2024), News is measured using data from the CNRDS database, defined as the natural logarithm of one plus the number of positive newspaper and online reports a firm receives annually. A higher News score reflects greater media visibility and positive public perception. As shown in columns (4)–(5) of Table 11, GFI has a statistically significant positive impact on both Fame and News at the 1% level. In summary, GFI not only improves firms’ environmental performance but also enhances their public image and market acceptance, thereby reinforcing their export competitiveness through reputation effects.

To empirically test the moderating effect, this study includes an interaction term between Gplant and Coz in the baseline regression model and conducts the analysis based on equation (2). Following the approaches of He and Huang (2017) and Chen et al. (2021), Coz is defined as the total proportion of shares held by all common institutional investors in a listed firm in a given year. The results presented in columns (1)–(6) of Table 12 show that the interaction term between Gplant and Coz is consistently and significantly positive, regardless of whether year-fixed effects, firm-fixed effects or control variables are included. This suggests that common institutional ownership plays a moderating role by providing financial support, professional management expertise and reputational endorsement, thereby strengthening the positive effect of GFI on FEP and enabling firms to achieve more stable and sustainable export outcomes amid international market fluctuations and competitive pressures.

To examine whether GFI generates spillover effects on the export performance of other firms within the same industry or region, this study follows the approach of Leary and Roberts (2014) by incorporating the average export performance of peer firms within the same industry and region into equation (1). Furthermore, given the potential lag in GFI’s impact, using the FEP in the following period helps mitigate concerns about reverse causality. To address possible sample selection bias arising from the binary nature of export propensity (Ene), this section focuses on the spillover effects on the continuous export scale (Osa).

Table 13 presents the empirical results on spillover effects at the industry and regional levels. The empirical results show that, at the industry level, the coefficient of Gplant is 0.2133 and statistically significant at the 1% level, indicating that GFI generates a significant industry-level spillover effect, thereby encouraging other firms within the industry to expand their export scale. Furthermore, when regressing using the one-period lagged industry spillover effects（F.Osa）, the results remain largely unchanged. However, at the regional level, the estimated coefficient of Gplant fails to reach statistical significance, suggesting that geographical proximity does not yield effective spillover channels. One plausible explanation is the “resource competition effect” (Liu et al., 2025), whereby certified firms secure greater policy and financial support, leaving noncertified firms at a disadvantage. This uneven allocation of resources reduces uncertified firms’ incentives for green upgrades and limits positive regional spillovers.

This paper treats the implementation of GFI, a voluntary environmental regulation policy, as a quasi-natural experiment and uses a multi-period DID approach to examine its effect on FEP. The results show that GFI significantly enhances FEP. Unlike previous studies that only consider the direct effect of GFI (Hu et al., 2025), we find that common institutional ownership plays a significant moderating role in amplifying its positive impact. Heterogeneity analyses further reveal that the effect is more pronounced among domestic enterprises, nonheavily polluting industries and capital-intensive industries. Consistent with existing research showing that GFI creates environmental incentives to stimulate innovation and ease financing constraints (Liu et al., 2025; Chen et al., 2025), this study corroborates those findings. It further extends the literature by demonstrating that technological innovation, resource allocation and corporate reputation are the key mechanisms through which GFI enhances FEP. Finally, we provide empirical evidence that the establishment of GFI generates spillover effects. These effects are significantly positive at the industry level but not statistically significant at the regional level.

First, efforts to expand the green manufacturing demonstration program and extend GFI coverage should be strengthened. Governments, particularly in developing countries, should accelerate standardized green factory certification, improve policy coordination and promote broader firm participation. Special attention should be given to supporting small and medium-sized enterprises in their green transformation through technological and managerial innovation. In addition, comprehensive financial, technological and human capital support should be reinforced throughout the transformation process to ensure successful upgrades and sustained export performance.

Second, policies should be tailored to firm-specific characteristics to maximize their effectiveness. Our heterogeneity analysis indicates that domestic firms, nonheavily polluting industries and capital-intensive industries benefit more from GFI. Hence, the government should adopt differentiated support measures. For domestic firms, stronger financing guarantees and technology subsidies can enhance their absorption of green innovation incentives. For nonheavily polluting industries, energy efficiency tax credits, digital-green integration pilots and other flexible tools can accelerate benefit diffusion. For capital-intensive industries, prioritizing long-term green credit and infrastructure investment can ease financing frictions and promote large-scale adoption of advanced technologies.

Finally, the spillover effects of green factory construction should be actively guided to promote coordinated development between upstream and downstream enterprises. Sector-wide green technology platforms can be established to reward firms with significant spillover potential and incentivize knowledge transfer through licensing, equipment leasing and collaborative innovation. In addition, cross-regional coordination mechanisms should be explored to facilitate the exchange of green technologies and managerial expertise, thereby raising the overall level of green manufacturing and international competitiveness across regions.

There are several directions for future research. First, regarding the research sample, green factories include listed firms, their subsidiaries and other enterprises, yet due to data limitations, this paper focuses only on listed firms. Future studies could investigate the green transformation and export performance of small and medium-sized enterprises once relevant data become available. Second, in terms of methodology, this paper relies primarily on a multi-period DID approach, while future research may incorporate machine learning techniques to capture more nuanced effects of GFI on FEP. Finally, future studies could adopt a broader perspective to assess the effectiveness of GFI’s mechanisms, for example, by examining its role in shaping international green trade rules and the restructuring of global value chains.