GTI is a critical pathway to achieving sustainable development (Liu M. et al., 2024). Existing studies have extensively explored its driving factors, focusing on policy mechanisms, green finance, and environmental regulation. In terms of policy mechanisms, Norberg-Bohm (1999) emphasized that well-designed policy incentives are essential for promoting environmentally friendly technological innovation, and that clear policy signals and strong incentives can encourage firms to adopt green technologies. Hong et al. (2021) further showed that green credit policies significantly enhance GTI by easing corporate debt financing constraints. Regarding green finance, Jiakui et al. (2023) provided empirical evidence that green finance plays a key role in improving green total-factor productivity and highlighted the need to establish a comprehensive legal framework for green finance. In the field of environmental regulation, Zhang et al. (2023) distinguished between real green technological innovation (RGTI) and strategic green technological innovation (SGTI), finding a U-shaped relationship between environmental regulation and both forms of GTI, with SGTI being more sensitive to regulatory intensity. These studies provide valuable theoretical and empirical insights into the multifaceted drivers of green innovation. However, little attention has been paid to the role of innovative industrial clusters as a meso-level policy instrument for promoting green innovation at the urban level. There is still a lack of systematic analysis on how such clusters facilitate the diffusion of green technologies, enable collaborative innovation, and enhance the allocation of innovation resources.

Industrial policy plays an increasingly prominent role in shaping environmental outcomes, particularly as green development becomes a national strategic priority. A growing body of literature has begun to examine the environmental impacts of industrial policy, including its potential to improve urban energy efficiency and promote sustainable industrial transformation. From the perspective of urban economic development, studies such as Deakin et al. (2018) propose smart city frameworks based on the Triple Helix model, highlighting the importance of IoT infrastructure and fair benefit-sharing in driving green transformation beyond conventional goals like energy savings and emissions reduction. In terms of economic system optimization, Yang et al. (2022) evaluated the National Innovative City Policy in China (2005–2017) and found that it improved urban energy efficiency through mechanisms such as innovation promotion and industrial upgrading, thereby enhancing the overall functioning of the urban economic system. Focusing on high-quality industrial development, Hong et al. (2022) employed a DID model to analyze the impact of carbon emissions trading in 276 Chinese cities from 2003 to 2016. Their results suggest that green innovation and improved resource allocation served as key transmission channels for increasing both single-factor and total-factor energy efficiency. Furthermore, they noted that market-oriented reforms and industrial agglomeration significantly amplified the policy effects. From the perspective of economic structural transformation, existing studies suggest that industrial policy can support the shift from resource-intensive to innovation-driven development models; however, the mechanisms linking this transformation to improvements in environmental efficiency remain insufficiently explored. Addressing this gap, Wang et al. (2025) examine how digitalization fosters urban sustainability by analyzing the impact of the digital economy on per capita CO₂ emissions across 286 Chinese cities from 2000 to 2023. Their findings show that digital economic growth significantly reduces emissions through increased urban population density and green innovation, with the effects further strengthened by industrial upgrading and partially offset by expanded transportation infrastructure.

Existing research has explored the environmental impacts of industrial policy from multiple perspectives, including urban economic development, system optimization, high-quality industrial growth, and economic structural transformation. While the broader effects of industrial policy are well-documented, the literature has largely overlooked the specific role of innovative industrial cluster policies as a meso-level intervention in this process. In particular, there remains a substantial gap in understanding how such policies causally promote green technological innovation, and how green innovation, in turn, serves as a key transmission mechanism for improving urban energy efficiency.

Improving urban energy efficiency is a critical objective for achieving sustainable development and addressing environmental challenges in rapidly urbanizing regions. Enhancing energy efficiency not only reduces resource consumption and carbon emissions but also promotes economic vitality and urban livability. Existing research identifies three main drivers influencing urban energy efficiency improvements: technological progress, urban transformation, and policy frameworks.

From the perspective of technological progress, Deakin et al. (2018) proposed a smart city development theory based on the Triple Helix model, emphasizing the role of regional IoT innovation and digital infrastructure. They highlight that beyond energy savings and emission reductions, equitable benefit-sharing mechanisms are essential for sustainable urban energy transitions. Regarding urban transformation, Yang et al. (2022) analyzed China’s National Innovative City Policy using panel data from 2005 to 2017, finding that the policy improves urban energy efficiency through both direct energy assessment and indirect pathways such as industrial upgrading and innovation. In terms of policy and institutional factors, Hong et al. (2022) applied a difference-in-differences model to evaluate the impact of carbon emissions trading across 276 Chinese cities from 2003 to 2016. Their findings indicate that the carbon trading system significantly increases single-factor and total-factor energy efficiency through stimulating green innovation and optimizing resource allocation, with the effects further amplified by marketization and industrial agglomeration.

However, despite these insights, the role of innovative industrial cluster pilot policies in promoting urban energy efficiency remains underexplored, particularly the key mediating role of green technological innovation. This study aims to fill this gap by deconstructing the mechanism linking industrial cluster policies, green technological innovation, and improvements in energy use efficiency.

In summary, although existing literature provides a solid theoretical foundation for understanding GTI, industrial policy, and urban energy efficiency, substantial research gaps remain. Table 1 summarizes the relevant literature and corresponding research gaps. First, the urban-level impact of industrial policies—particularly those embedded in innovative industrial clusters—has not been systematically explored in the context of GTI. Second, most industrial policy studies focus on economic growth and structural upgrading, with limited attention to their environmental and innovation-oriented outcomes. Third, research on energy efficiency tends to emphasize technological and spatial factors, often overlooking the role of meso-level industrial policy instruments. Notably, few studies have integrated innovative industrial cluster pilot policies, GTI, and energy efficiency within a unified analytical framework. This article aims to bridge these gaps by examining how cluster-based policy interventions influence GTI and, in turn, increase urban energy efficiency. By highlighting the mediating role of GTI and emphasizing the importance of targeted policy support, the study contributes to a more comprehensive understanding of how industrial policy can promote sustainable urban development across economic, social, and environmental dimensions.

In June 2013, the Ministry of Science and Technology officially released 10 pilot innovative industrial clusters, including the Beijing Zhongguancun mobile Internet industrial cluster, marking that China’s pilot of innovative industrial clusters has entered a new stage. The pilot cities for innovative industrial clusters are listed in Table 2. These innovative industrial clusters were able to obtain financial subsidies and preferential policies from governments that helped to provide technical expertise as well as tax reductions for enterprises in the clusters (Xu and Hu, 2024).

In view of the accessibility and completeness of data, the final sample selects 281 prefecture-level cities from 2006 to 2021, including 55 innovative industrial cluster pilot cities. The data sources are as follows: first, pilot data are from the Torch Center of the Ministry of Science and Technology of China. The second is the green patent data from the State Intellectual Property Office of China. Third, city-level data of other characteristic indicators are taken from the National Bureau of Statistics of China and statistical yearbooks over the years. The missing data are imputed using the mean value method. Cities with a large amount of missing data, such as Laiwu, Chaohu, Tongren, and Lhasa, were deleted.

Table 3 presents the descriptive statistics of the main variables in this study. The values of urban green technology innovation exhibit a wide range between the minimum and maximum, indicating an uneven development of green technology innovation activities across Chinese cities. The average value of the innovative industrial cluster policy is 0.0827, meaning that 8.27% of the samples in this study belong to the treatment group, which suggests that China's industrial cluster policy is still in an emerging stage of development.

Table 4 presents the baseline regression outcomes, where the explanatory variable’s coefficient, t r e a t i t × t i m e i t , is significantly positive at the 1% level. This indicates that implementing pilot policies for innovative industrial clusters positively affects urban GTI, significantly improving its quality as well as quantity. Therefore, Hypothesis 1 is validated.

This study uses the event analysis method to test whether the hypothesis of parallel trend is met before implementing the policy. That the experimental group and the control group meet the parallel trend hypothesis is an important prerequisite for applying the DID model; otherwise, the difference after implementing the policy may be caused by other differences. In this study, there should be no substantial difference in green technology innovation between the two groups of cities before the policy implementation, indicating that the experimental group and control group meet the parallel trend hypothesis. To further examine dynamic effects, this study selects the year before policy implementation as the base period. The study uses the interaction term between dummy variables of 3 years before and after policy implementation and the treatment variable t r e a t i t × t i m e i t as the core explanatory variable for dynamic effect analysis. Figure 2 reveals the estimated coefficients and 95% confidence intervals of the year-by-year interaction terms. Before the policy implementation, there was no remarkable distinction between the experimental group and the control group, establishing the parallel trend hypothesis. When implementing the innovative industrial clusters pilot policy, from the third period onward, the estimated coefficient became significantly positive and showed an upward trend. This indicates that the pilot policy significantly boosts green technology innovation in pilot cities over the sample period, with the effect strengthening over time, creating a virtuous cycle.

Drawing on the research of Tang et al. (2023), this article uses an undirected super efficiency SBM model that includes unexpected outputs and uses Dearun Tools software to calculate the green total-factor energy efficiency under the assumption of constant returns to scale to represent the energy utilization efficiency (eff) of each city. In the selection of variables, labor, capital, and energy are used as input factors. Each city’s GDP is regarded as the expected output, while industrial sulfur dioxide, soot, and wastewater emissions are considered as unexpected output. Control variables include the level of financial development, high-level urban industrialization level, and industrial structure (ais) with the added value of the third industry and measured by the ratio of the added value of the second industry (Zhang et al., 2019), population density (lndensity), with urban population and urban administrative area of the measured values of the ratio of the land area (Baur et al., 2014).

Table 11 presents the empirical test results of how the pilot policy for innovative industrial clusters impacts urban energy efficiency. To address potential endogeneity issues, we used a mixed OLS model, a panel two-way fixed effect model, and a multi-dimensional panel fixed effect model for regression analysis. Results are shown in columns (1), (2), and (3) of Table 11, with an estimated coefficient of 0.0366, significant at 1%. This indicates that the pilot policy positively promotes urban energy efficiency, with robust results confirmed by tests such as the parallel trend test and the placebo test. The pilot policy for innovative industrial clusters likely boosts urban energy efficiency by guiding the development of high-tech and strategic emerging industries, which tend to have higher energy efficiency and lower energy consumption. This contributes to optimizing the city’s industrial structure. Additionally, innovative industrial cluster pilot policy encourages enterprises to increase R&D investment, fostering technological innovation that increases energy use efficiency and drives continuous advancements in energy technology, thereby strongly supporting overall improvements in urban energy efficiency.

The article selects empirical GTI and strategic GTI as mechanism variables to analyze more deeply how innovative industrial cluster pilot policies impact urban energy utilization efficiency. The mechanism analysis results are shown in Table 12. The coefficients of t r e a t i t × t i m e i t are 0.4659 and 0.1937, respectively, which are significant at the 1% level. Meanwhile, scholars have found that GTI is crucial for achieving sustainable development for humanity. Using data covering 17 OECD countries, Wurlod and Noailly (2018) found that green innovation can reduce the energy intensity of the industrial sector. In addition, Du and Li (2019) found that green technology innovation can reduce the cost of clean energy and increase the proportion of clean energy in the energy consumption structure. In light of the above analysis, the findings explain that the quality and quantity of urban green technology innovation are important mechanisms for innovative industrial clusters pilot policies to increase urban energy utilization efficiency. Quality in green technology innovation is reflected in both the depth and breadth of technological progress. High-quality advancements significantly improve energy efficiency, reduce consumption and emissions, and achieve economies of scale through widespread application. Quantity, on the other hand, is measured by the number of innovative achievements, which provide more energy-saving technologies and products for selection and application, thereby increasing energy efficiency. Active innovation activities further drive continuous improvements in energy utilization.

Therefore, the quality and quantity of urban green technology innovation are crucial for improving urban energy utilization efficiency under the pilot policies for innovative industrial clusters. By strengthening research and application, fostering industrial agglomeration and synergistic effects, and under policy guidance and support measures, cities can continuously improve energy efficiency and promote sustainable development.

The empirical findings of this study provide strong evidence for the pivotal role of China’s innovative industrial cluster pilot policy in promoting urban GTI and increasing energy efficiency. Robustness checks confirm that targeted industrial policies can function as effective meso-level instruments. They not only stimulate innovation but also translate technological advances into measurable environmental outcomes. This extends prior research on the importance of policy design and institutional support in fostering green innovation ecosystems (Norberg-Bohm, 1999; Jiakui et al., 2023). The results underscore the multiple pathways through which innovation can mediate environmental performance.

Placing these findings in an international context highlights their broader relevance. China’s cluster policy shows parallels with South Korea’s heavy and chemical industry strategy, demonstrating how coordinated institutional frameworks and innovation-driven upgrading can generate competitive advantages and structural economic transformation (Lane, 2025). Similarly, evidence from the European Union and OECD countries indicates that integrated policies—combining market mechanisms, institutional robustness, and strong intellectual property regimes—are critical for advancing energy efficiency and green growth (Deakin et al., 2018; Hong et al., 2022). Together, these comparisons reinforce the view that supportive institutional environments are essential to maximize the impact of industrial innovation policies.

Experiences from other developing countries provide important comparative insights into the role of institutional design and policy frameworks in shaping green industrial transformation. In Ethiopia, weak institutional capacity and limited implementation often undermine the effectiveness of green industrial policies, contrasting with China’s stronger institutional coordination (Shen et al., 2023). India illustrates a different pathway, where state-organized financing with minimal fiscal expenditure has enabled a rapid scale-up of solar deployment. This model differs from China’s more fiscally expansive, state-driven investment approach, which has more effectively fostered technological upgrading and industrial clustering (Larsen, 2025). Brazil, India, and South Africa, as second movers in electromobility, benefit from lower technological risks but face fiercer global competition, unlike China’s early and state-driven experimentation that shaped EV development under high uncertainty (Lema et al., 2024). Brazil’s policy orientation toward biofuels and emerging green hydrogen pathways also reflects a distinct strategy, but fragmented governance and limited emissions reduction stand in sharp contrast to China’s large-scale industrial cluster strategy, which more successfully integrates innovation with environmental outcomes (Dos Santos Martins et al., 2024). Taken together, these comparisons highlight that while local institutional quality, financial capacity, and industrial legacies condition policy effectiveness, the fundamental mechanisms linking industrial policy, innovation, and environmental sustainability remain broadly applicable. Accordingly, China’s experience not only deepens theoretical understanding of meso-level policy design but also offers transferable lessons for both developed and developing countries seeking to advance sustainable industrial transformation.

Several policy recommendations can be drawn based on the empirical findings to maximize the effectiveness of China’s innovative industrial cluster pilot policy in promoting GTI and energy efficiency. First, targeted fiscal and institutional support is essential to promote GTI, especially in ecologically fragile or industrially transforming regions. Measures such as dedicated R&D funds, tax incentives, preferential green financing, and institutional innovations like R&D centers and technology incubators can accelerate green technology commercialization. This approach offers a transferable lesson for other developing and emerging economies facing similar resource constraints.

Second, robust intellectual property (IP) protection enhances the effectiveness of innovation policies. Strengthening legal frameworks, establishing IP courts or mediation centers, expanding legal aid, and supporting subject matter experts (SMEs) through training and integrated IP service platforms can help ensure that technological advances translate into market outcomes. Globally, regions with weak innovation systems can apply these insights to improve green innovation performance.

Finally, aligning industrial cluster strategies with local comparative advantages and fostering collaborative innovation ecosystems can maximize policy impact. Strategic planning, supply chain coordination, and technology matchmaking, together with public–private partnerships and university–enterprise–research collaboration, create synergies that enhance both GTI and sustainability. These principles offer actionable guidance for policymakers in developed and developing countries seeking to leverage industrial clusters for sustainable urban development.

This study provides both theoretical and empirical insights into the role of China’s innovative industrial cluster pilot policy in advancing urban GTI and energy efficiency. The results indicate, first, that the policy exerts a sustained positive effect on urban GTI, which remains consistent after a range of robustness checks, including parallel trend assessments and endogeneity treatments. Second, the heterogeneity analysis suggests that the policy effect is more evident in the eastern and western regions, in cities with stronger environmental protection orientations, and in those with more developed intellectual property protection. Third, the mechanism analysis identifies three key channels through which the policy promotes GTI: enhancing entrepreneurial activity, increasing the concentration of science and technology talent, and expanding financial investment in innovation. Finally, further evidence shows that the policy contributes to improvements in urban energy efficiency, with GTI serving as a mediating mechanism.

Overall, these findings contribute to a better understanding of how meso-level industrial policies may link innovation and environmental outcomes. While the conclusions are derived from the Chinese context, the evidence may hold broader relevance for other emerging economies seeking to design industrial policies that integrate technological upgrading with sustainable development.

Several limitations of this study merit careful consideration. First, data availability constrains the analysis, limiting the examination of micro-level firm behaviors, sectoral heterogeneity, and dynamic innovation processes. This may obscure important within-industry variations in how industrial policies affect GTI and energy efficiency. Second, although China’s innovative industrial cluster pilot policy provides a valuable quasi-natural experiment, the transferability of the findings to other institutional contexts remains uncertain. Differences in governance capacity, industrial structures, and policy implementation mechanisms may influence both the magnitude and direction of policy effects. Third, the study relies primarily on prefecture-level data, which may overlook intra-urban heterogeneity, such as variations across districts, industrial parks, or firm networks. These limitations highlight the need for future research that integrates multi-scalar data, cross-national comparisons, and firm-level evidence to better understand how industrial policies foster sustainable innovation.

Future research could extend this study in several concrete ways. First, cross-national comparative analyses could evaluate the effectiveness of innovative industrial cluster policies across countries with varying institutional capacities, governance structures, and levels of economic development. Such studies could employ comparable city- or region-level panel data to ensure consistency. Second, firm- or establishment-level research could leverage microdata to examine how policy interventions affect innovation behaviors, R&D investment, and technology adoption. This approach would provide a more detailed understanding of the causal mechanisms linking policy to GTI and energy performance. Third, sector-specific studies could focus on high-impact industries, such as clean energy, advanced manufacturing, or green transportation. These studies could investigate how policy design interacts with industry characteristics, resource endowments, and technological trajectories, offering practical insights for designing targeted interventions that enhance both innovation and environmental outcomes.