Exploring how green location-oriented industrial policies promote urban energy efficiency: evidence from National eco-industrial demonstration parks in China

Abstract

Enhancing energy efficiency is a critical pathway toward achieving energy conservation, emissions reduction, and green development. However, traditional location-oriented industrial policies, while promoting economic growth, are often accompanied by issues such as a high proportion of energy-intensive industries and an unreasonable energy consumption structure. In contrast, green location-oriented industrial policies, exemplified by the National Eco-industrial Demonstration Parks (NEDPs) play a significant role in enhancing energy efficiency. Against this backdrop, grounded in theories such as market failure, Porter hypothesis and the pollution haven hypothesis, this study employs panel data from 196 Chinese cities between 2005 and 2020 and applies a time-varying difference-in-differences (TV-DID) approach to analyze the impact of NEDPs on urban energy efficiency. The results reveal that NEDPs serve as an effective policy instrument for boosting urban energy efficiency. This enhancement operates primarily through three mechanisms: the strengthening of environmental regulation, the promotion of green technology innovation, and the facilitation of industrial structure upgrading. A further investigation into the heterogeneity of the data indicates that the positive impact is more pronounced in cities that are not old industrial bases, non-resource-based, or those where the parks have already been established. Conversely, the impact is less pronounced in conventional industrial cities, resource-dependent cities, and those where parks are still under development. This paper puts forward several policy implications, including the acceleration of the green transformation of traditional development zones, the design of differentiated green industrial policies tailored to urban types, and the promotion of regional collaborative governance.

Introduction

Low-carbon energy consumption contributes to environmental sustainability (Caglar et al., 2024a)¹. The Tracking SDG7: Energy Progress Report 2023 indicates that the average annual improvement in global energy intensity over the period 2010–2020 was only 1.8 per cent, a figure that falls well below the target. In order to compensate for this shortfall, the report emphasizes that the improvement in global energy intensity must exceed 3.4% from 2020 to 2030. This context underscores the critical relevance of this topic to global sustainable development goals. Furthermore, to achieve net-zero emissions by 2050, as proposed by the International Energy Agency, it is necessary that the average annual improvement in energy intensity exceed 4% for the remainder of the period until 2030. The scaling up of energy efficiency policies and investments to improve energy efficiency has become the most important agenda for global sustainable development (Van Doren et al., 2016; Flachenecker and Rentschler, 2019; Caglar et al., 2024b)^2,3,4. China is the second largest consumer of energy worldwide, with its coal consumption still in the growth stage. Furthermore, it should be noted that total carbon dioxide emissions have yet to reach a peak, and the resource consumption and environmental problems caused by energy extraction and utilisation are becoming increasingly serious (Lin and Li, 2015)⁵, thus rendering it a critical and relevant case study.

Improving energy efficiency is a necessary means of overcoming the constraints posed by limited resources and environmental concerns (Buonomano et al., 2022)⁶. However, China’s traditional development zone policies, while promoting economic growth, are often accompanied by issues such as high pollution emissions and high energy consumption, thereby hindering the improvement of energy utilization efficiency. In response to growing environmental pressures and the imperative to achieve net-zero emissions, the Chinese government has leveraged the framework of traditional development zones to implement a green location-oriented policy, establishing National Eco-industrial Demonstration Parks (NEDPs) guided by clean production standards, circular economy principles, and industrial ecology (Hua and Tan, 2022)⁷. By 2022, 95 NEDPs had been approved, representing 18% of all national development zones. Given that energy efficiency serves as a nexus between economic growth and environmental governance, and considering the ongoing dominance of state-led industrial policy in shaping China’s energy landscape (Yuan and Xie, 2014)⁸, a evaluation of this policy’s effectiveness and underlying mechanisms is essential.

Against this backdrop, the present study investigates the influence of NEDPs on urban energy efficiency, with the aim of providing empirical evidence to support policy refinement and sustainable urban development. To provide robust empirical evidence, this study employs a panel dataset encompassing 196 Chinese prefecture-level cities over the period 2005–2020. The choice of this sample provides a comprehensive spatial and temporal coverage to capture the policy’s effects. Empirical strategy is grounded in a time-varying difference-in-differences (TV-DID) model, a methodology particularly appropriate for evaluating the gradual rollout of policies like the NEDPs, as it accurately identifies causal effects by accounting for heterogeneous treatment timing. Furthermore, given the differences in regional economic structures and resource endowments, NEDPs may have significantly heterogeneous effects on energy efficiency. Therefore, this paper will further analyze their mechanisms and heterogeneity to examine their impact pathways and applicable scenarios.

Despite the fact that extant literature has evaluated the implementation outcomes of NEDPs, the majority of studies focus on indicators such as carbon emissions and pollution reduction, thereby failing to fully incorporate energy efficiency considerations within the “economic-environmental” dimension. Indeed, as a comprehensive environmental policy, the impact mechanisms and pathways of NEDPs on energy utilisation efficiency warrant further exploration. This gap highlights the originality of the study focus. In comparison with extant literature, the potential contributions of this study are primarily reflected in the following three aspects:

Firstly, it extends the research scope by focusing on urban energy efficiency, offering new insights into NEDPs’ policy dividends. The present paper is grounded in the comprehensive characteristics of urban sustainable development. An in-depth analysis is conducted of the impact of NEDPs on urban energy efficiency, with explicit reference to China’s energy structure features and carbon emission constraints. This provides new theoretical perspectives and practical methods for promoting urban energy efficiency improvements. Secondly, with regard to the research content, the present paper systematically analyses the mediating pathways of the effects of environmental regulatory pressure, the effects of green innovation and the effects of industrial structure optimisation at two levels - namely, endogenous motivation and implementation mechanisms - by integrating the combination attributes of NEDPs policies and the behavioural motivations of urban actors. Concurrently, it differentiates the heterogeneity of city attributes, resource endowments, and policy implementation stages to identify the differentiated impacts of NEDPs on energy efficiency, thereby providing comprehensive and effective empirical support for refining NEDPs policies. Thirdly, this study offers important and clear policy implications for the green transformation of traditional development zones and the design of differentiated industrial policies, which are crucial for policymakers in China and other industrializing countries.

The remainder of this paper is structured as follows: Sect. 2 reviews the relevant literature and outlines research gap. Section 3 outlines the policy context, theoretical framework, and research hypotheses. Section 4 describes the empirical methodology, variable construction, and data sources. Section 5 presents the main results, robustness checks, and discusses the mechanisms and heterogeneity effects. Section 6 discusses the empirical findings and outlines research limitations and future directions. Section 7 concludes with the key findings and their policy implications.

Literature review

Environmental and economic effects of NEDPs

As a distinctive location-oriented environmental policy in China, the conceptual foundations of China’s NEDPs can be traced to earlier theoretical in industrial ecology and circular economy. The pioneering work of Lowe (1997)⁹ and Chertow (2000)¹⁰on industrial symbiosis and system-level resource cycling provided a framework for transforming traditional industrial zones into closed-loop systems that mimic natural ecosystems. Simultaneously, the circular economy paradigm, notably advanced by Boulding (1966)¹¹ and Pearce and Turner (1989)¹², emphasized “closing material loops” through reduce-reuse-recycle principles, shifting policy focus from end-of-pipe pollution control to proactive, systemic resource management. These ideas collectively underscored the potential of location-oriented industrial policies to integrate environmental and economic objectives, directly influencing China’s design of NEDPs, which now serve as a policy instrument for implementing SDG-oriented strategies (Bai et al., 2014)¹³.

By leveraging mechanisms such as industrial symbiosis, systemic optimization, and shared infrastructure, NEDPs offer a practical approach to reconciling economic growth with environmental sustainability (Yu et al., 2015)¹⁴. Empirical studies indicate that NEDPs contribute positively to urban economic development (Pu et al., 2021)¹⁵ and exhibit notable environmental performance (Sacirovic et al., 2019)¹⁶. Specifically, they have been shown to mitigate air pollution (Li et al., 2023)¹⁷, reduce carbon emissions (Hua and Ye, 2023)¹⁸, enhance carbon emission efficiency (Liu et al., 2023)¹⁹, and lower industrial sulfur dioxide emissions (Chen et al., 2023)²⁰. Moreover, NEDPs play a significant role in promoting urban green innovation (Wu and Gao, 2022; Wu et al., 2023)^21,22 and facilitating the green transition and sustainable development of urban industrial sectors (He and Li, 2023)²³. However, some studies suggest that NEDPs have not fully met their intended environmental goals. Although they enhance economic performance along the industrial chain (Kim et al., 2018)²⁴, they may also lead to the displacement of foreign direct investment due to stringent environmental regulations (Hua and Tan, 2022)⁷. Furthermore, research by Cao et al. (2023)²⁵ and Chen et al. (2023)²⁰ indicates that while NEDPs effectively reduce carbon dioxide and sulfur dioxide emissions within their boundaries, they may simultaneously increase pollution levels in adjacent cities, highlighting the need for regional-scale policy assessment and a broader ESG (Environmental, Social, and Governance) perspective in evaluation.

The debate on environmental regulations and energy efficiency

The relationship between governmental environmental regulation policies and energy efficiency has been the subject of extensive debate, particularly within the context of global sustainability agendas (Işık et al., 2024a)²⁶. Conventional wisdom in the field suggests that restrictive environmental policies may temporarily reduce energy efficiency by imposing costs associated with pollution control (Zhang et al., 2021)²⁷. Conversely, the “Porter hypothesis” proposes that environmental regulation policies can enhance energy efficiency by catalysing technological innovation (Dzwigol et al., 2023; Chen and Chen, 2024)^28,29. In addition to the implementation of restrictive policies, governments also implement incentive-based environmental policies with the aim of enhancing corporate energy efficiency. Such policies include emissions trading (Shi and Li, 2020)³⁰, energy innovation R&D policy (Caglar et al., 2024c)³¹, and pilot policies for energy rights trading (Xue and Zhou, 2022)³². Notably, the integration of SDG-oriented benchmarks into policy frameworks is increasingly recognized as essential for achieving systemic energy efficiency improvements (Caglar et al., 2024d; Işık et al., 2024b)^33,34. Recent research emphasises that transitioning to renewable resources can reduce energy dependence in the energy market, playing a crucial role in enhancing energy efficiency and optimising the allocation of natural resources, thereby advancing the achievement of multiple Sustainable Development Goals (Işık et al., 2024c)³⁵.

The extant literature suggests that environmental regulation policies can significantly improve energy efficiency; however, the majority of research in this field focuses exclusively on incentive-based or restrictive environmental policies, with a paucity of empirical studies of location-oriented environmental policies that combine both incentive-based and restrictive attributes. Moreover, while extant literature has evaluated the environmental and economic outcomes of NEDPs, a comprehensive analysis focusing on urban energy efficiency, with particular regard to its mechanisms and heterogeneity, remains under-explored. The present study aims to address this lacuna in the existing literature.

Policy background and theoretical analysis

Policy background

While enhancing enterprise performance and promoting regional economic growth (Kong; 2021; Wang et al., 2022)^36,37, traditional location-oriented policies, mainly in development zones and high-tech zones, have resulted in environmental issues such as high energy consumption, deterioration of water quality, and exceeding of pollutant emissions (Gao et al., 2023)³⁸. China’s ecological and environmental sectors have increasingly acknowledged the necessity of integrating green development objectives into traditional location-oriented policies and implementing green location-oriented policies. This is of great significance in resolving the contradictory problem of economic development and environmental protection. The NEDPs, which serves as the focus of this paper, represents a typical example of a green location-oriented policy oriented towards sustainable development and ecological environmental protection. The NEDPs is established on the basis of two types of zones: national economic and technological development zones and high-tech industrial development zones. These NEDPs are created in accordance with a process that may be described as follows: “self-declaration-approval of creation-acceptance and naming-periodic review.” Furthermore, a dynamic management mechanism is employed, whereby NEDPs that fail to meet the requisite qualifications are promptly notified, criticized, rectified, and, if necessary, withdrawn within a limited period of time.

In order to facilitate the establishment of NEDPs, the Chinese government has issued a series of policy documents. The policy documents delineate the requisite number of new eco-industrial chain projects in the parks, the ratio of energy use, the utilization rate of industrial wastewater, as well as the necessity that key pollutant emissions must meet standards and that key enterprises must implement cleaner production. Since the implementation of the NEDP policy in 2000, the number of NEDPs created has increased annually, with 95 NEDPs created in China as of 2022. The construction of NEDPs in the eastern region of China is influenced by a number of factors, including resource endowment, location conditions, economic development level, industrial structure, and so forth. Consequently, the construction of NEDPs in the eastern region commences at an earlier stage and encompasses a greater number of NEDPs than in the central and western regions. This has resulted in a phenomenon of imbalance in the development of the region. A total of 71 out of 95 NEDPs are located in the China eastern region, representing 74.7% of the total number. The central region accounts for 19.0% of the total number with 18 NEDPs, while the western region accounts for 6.3% with 6 NEDPs. The distribution of NEDPs demonstrates a gradual expansion from the eastern to the central and western regions of China (Chen et al., 2021)³⁹.

Theoretical analysis and research hypothesis

The establishment of NEDPs represents a pivotal integration of environmental objectives into location-oriented industrial policy. To systematically analyze its impact on urban energy efficiency, this study constructs a theoretical framework that synthesizes the theory of environmental externality internalization, the Porter Hypothesis, and theories of structural change. This paper posit that the NEDPs policy, through its stringent standards and guiding principles, primarily influences energy efficiency via three interconnected channels: first, by strengthening environmental regulation to correct market failures and force efficiency gains (the environmental regulation effect); second, by inducing green technological innovation that creates compensation effects and alters factor substitution relationships (the technological innovation effect); and third, by catalyzing industrial structure upgrading that reallocates energy factors from low-productivity to high-productivity sectors, thereby generating a structural dividend (the industrial structure effect). These pathways, as illustrated in Fig. 1, constitute the core analytical framework for deriving research hypotheses. This paper will analyze the mechanism of NEDPs on urban energy efficiency from these three dimensions.

Fig. 1

Mechanism of action analysis.

Environmental regulation effects of the NEDPs

Grounded in the theory of internalizing environmental externalities, environmental regulation transforms the negative externalities of environmental pollution into internal costs for enterprises by setting stringent pollutant emission and energy consumption standards. This incentivizes companies to adopt more efficient energy utilization practices to achieve compliance. The NEDPs are devised in accordance with the stipulations of cleaner production, the tenets of the circular economy, and the principles of industrial ecology. This approach enhances the effectiveness of urban environmental regulations. Conceptually, the policy exerts both a “selection effect” and a “technique effect”. On the one hand, the Pollution Sanctuary Hypothesis posits that the stringent environmental constraints of NEDPs may prompt highly polluting firms to relocate to regions with less stringent environmental policies, thereby reducing pollution emissions within the park (Hamaguch, 2024)⁴⁰. On the other hand, the environmental regulatory policies of NEDPs also strictly control the entry of new projects with high energy consumption and pollution, thereby preventing the entry of high-polluting enterprises through the “environmental barriers” generated by the policies (Hua and Ye, 2023)¹⁸. Furthermore, the NEDPs policy introduces mandatory regulations on the proportion of clean energy use, thereby promoting the replacement of coal and other fossil energy sources with clean energy sources (e.g., wind and solar energy) through environmental regulation. This substitution mechanism serves to reduce reliance on coal, improve the energy consumption structure, and drive energy efficiency (Han et al., 2007)⁴¹.

The theory of industrial ecology emphasizes the recycling of materials and energy within systems in order to maximize resource efficiency. NEDPs are designed based on the concept of a circular economy, which requires the establishment of a circular economy system within the park and sets forth clear guidelines regarding the utilization rate of industrial wastewater. Consequently, the by-products or wastes generated in the production process of upstream enterprises in the park can become raw materials or intermediate inputs for downstream enterprises, thereby facilitating the recycling of wastes generated by enterprises within a diversified industrial structure. This, in turn, contributes to the promotion of energy efficiency (Wan et al., 2019)⁴². In summary, the NEDPs policy enhances energy efficiency by strengthening environmental regulations, which is known as the “environmental regulation effect.” The following research hypothesis is hereby proposed:

H1

NEDPs policy facilitate urban energy efficiency by reinforcing environmental regulations.

Technological innovation effects of the NEDPs

According to endogenous growth theory and green innovation theory, technological innovation is the core factor driving the improvement of energy efficiency (Berrone et al., 2013)⁴³. The impact of NEDPs on green technological innovation is mainly manifested in the following aspects: Firstly, NEDPs policy effectively mitigates the risk of green technological innovation by providing a rational framework for the allocation of financial resources. The high difficulty, high risk, and public goods attributes of green technology innovation lead to insufficient motivation for independent innovation when enterprises weigh the costs and benefits. NEDPs’ tilted policies on science and technology innovation, green development funds, and incentive policies such as R&D subsidies and Tax exemptions can reduce the cost of green R&D for enterprises and offset the risk of externalities associated with green technological innovation activities (Dao et al. 2025)⁴⁴, thereby enhancing the enterprises’ enthusiasm for green technology innovation. Secondly, the NEDPs policy can “force” enterprises to engage in green technological innovation. According to the Porter hypothesis, NEDPs binding environmental policies will result in the “innovation compensation effect,” compelling some polluting enterprises with superior production efficiency and technological foundation to increase their investment in green technology research and development (Demirel and Kesidou, 2011; Berrone et al., 2013)^43,45. Thirdly, the NEDPs policy facilitates the aggregation of innovation factors through demonstration and leadership effects. Grounded in the signaling theory (Alós-Ferrer and Prat, 2012)⁴⁶, NEDPs policy orientation has a signaling effect, which not only guides more social capital to invest in green technological innovation projects, but also promotes the influx of innovative talents and advanced knowledge into the park. This enriches the resource elements for enterprises in the park to carry out green technological innovation. Consequently, this promotes green innovation of polluting enterprises.

Green technological innovation, as a distinct form of environmentally directed technological progress, serves as a critical pathway for improving energy efficiency and realizing the “double carbon” goals (Kazemzadeh et al., 2025)⁴⁷. On the one hand, it can reshape the technological trajectory of energy use, thereby altering the output efficiency of various factors of production, enhancing enterprise productivity, and promoting the efficient utilization of energy along this new path (Chang and Hu, 2010)⁴⁸. On the other hand, grounded in the factor substitution theory, green technological innovation influences energy efficiency by optimizing the reallocation of energy within production processes (Dao et al. 2025)⁴⁴. Specifically, in the production process, such innovation alters the marginal rate of technical substitution among factors, leading firms to substitute energy inputs with relatively lower-cost factors such as capital and labor. As a result, energy consumption per unit of output is reduced. In summary, NEDPs policies enhance energy efficiency by promoting green technological innovation, which is known as the “technological innovation effect.” The following research hypothesis is hereby proposed:

H2

NEDPs policy facilitate urban energy efficiency by fostering the green technology innovation.

Industrial structure effects of the NEDPs

The “Matchday-Clark Theorem” elucidates the evolution of industrial structure, and the optimization and upgrading of industrial structure can directly alter the relationship between input and output of energy factors, which in turn affects the total factor energy efficiency. Specifically, NEDPs optimize the industrial structure through the following three channels: Firstly, NEDPs eliminates traditional polluting enterprises. The implementation of binding policies will increase the cost of sewage treatment, which will “squeeze out” pollution-intensive enterprises and inhibit the entry of high-polluting and high-energy-consuming enterprises, thus achieving the optimization of the industrial structure. Secondly, NEDPs can facilitate the transformation of traditional industries. The introduction of green technology in the park has the potential to transform the traditional production technology equipment and production process, thereby increasing the utilization rate of high-energy consumption, high-pollution and high-emission industries (Hua and Tan, 2022)⁷. This will facilitate the gradual development of an industrial structure that is oriented towards high productivity, technology, and capital intensification, thus enabling the realization of an upgraded industrial structure. Thirdly, NEDPs can facilitate the growth of the cleanliness industry. The green investment promotion policy of NEDPs and the mandatory requirement of the number of new eco-industrial chain projects can attract cleanliness industry deployment. Furthermore, it can facilitate the growth of upstream and downstream supporting industries, such as environmental protection equipment, environmental protection engineering construction, and environmental protection services within the park. This, in turn, can initiate a green transformation of the entire industrial chain, leading to the optimization and upgrading of the industrial structure.

The upgrading of industrial structures can promote energy efficiency through two distinct mechanisms: the resource allocation effect and the structural dividend effect. On the one hand, industrial structure upgrading will result in alterations to the scale ratio among industries. The energy demand, energy intensity, and energy-saving potential of different industries vary considerably. The secondary industry consumes the most energy, followed by the primary industry and finally the tertiary industry. Consequently, the optimization of industrial structure, namely the transition from the secondary industry, which is characterized by high energy consumption and low value-added, to the tertiary industry, which is distinguished by low energy consumption and high value-added, can directly reduce the demand and consumption of energy. This, in turn, can enhance the level of total factor energy efficiency (Kazemzadeh et al., 2023)⁴⁹. On the other hand, the “structural dividend hypothesis” argues that industrial structure upgrading is a process of factor reallocation and efficiency improvement among different industries. When energy factors are transferred from low-productivity industries to high-productivity industries, an equal amount of energy factors will result in a greater economic output. This indicates that the unimpeded flow of energy factors will enhance the overall energy efficiency (Maddison, 1987)⁵⁰. In summary, NEDPs policy enhances energy efficiency through the optimization of industrial structure, which is referred to as the “industrial structure effect.” The following research hypothesis is hereby proposed:

H3

NEDPs policy facilitate urban energy efficiency by promoting industrial structure upgrading.

H4

NEDPs policy facilitate the promotion of urban energy efficiency.

Data and methods

Data sources and description

In order to provide a more comprehensive estimation of the policy effect of NEDPs, this paper selects the data of 196 Chinese cities with national-level development zones (including economic development zones and high-tech zones) from 2005 to 2020 as the research sample. A total of 48 cities with NEDPs were selected as the treatment group, while 148 cities with national-level development zones but without approved NEDPs were designated as the control group. The List of NEDPs, released by the Ministry of Environment and Ecology of China in 2017, indicates that a total of 94 NEDPs have been approved for creation nationwide. This paper examines 92 NEDPs, which are primarily located in 48 cities.

It is pertinent to highlight that the inaugural year of the NEDPs was 2001, during which the Guigang’s NEDPs was established. However, the park did not pass the national acceptance until 2022, while the construction cycle of NEDPs is generally 3–7 years. Therefore, it can be reasonably concluded that this park is an outlier and therefore excluded from the study. In 2003, Shandong Lubei Enterprises Group became the second one to be approved to build the NEDPs. However, due to the nature of the park as a corporation, it is no longer within the scope of this paper. Furthermore, in 2021, the Ministry of Environment and Ecology of China approved the construction of NEDPs in the Jinggangshan Development Zone and the Jiaxing Development Zone. However, as this construction commenced outside the sample period of this paper’s study, it is also not considered. The sample period encompasses multiple construction phases, commencing with the formal approval of Suzhou Industrial Park and other parks as NEDPs in 2008 and concluding with the approval of Changsha Hi-Tech Industrial Development Zone to build NEDPs in 2015.

In the case of cities with multiple NEDPs, this paper employs the approved creation time of the initial NEDPs as the point at which the treatment is initiated. In order to narrow the initial discrepancy between the control and treatment groups and to reduce the potential for selection bias, this paper employs cities with national development zones that have not yet been approved for NEDPs as the control group. In the subsequent robustness test, this paper further removes provinces without NEDPs in order to narrow the gap between the control and treatment groups. With the exception of the data on NEDPs, which were collected manually by the authors, the remaining data required for this article were obtained from a variety of sources, including the China Urban Statistical Yearbook, the China Regional Economic Statistical Yearbook, the China Energy Statistics, the China Economy Statistical Database, and the statistical yearbooks of the corresponding cities, as well as the National Economic Development Statistical Bulletin. Any individual missing data were supplemented by interpolation. It should be noted that industrial smoke and dust emissions were only counted until 2010, and were changed to industrial smoke and dust emissions since 2011. The data on patents are sourced from the State Intellectual Property Office (SIPO) and are classified and screened according to the Green List of International Patent Classification, which was issued by the World Intellectual Property Organization (WIPO) in 2010. Furthermore, the relevant nominal economic variables are deflated by the GDP deflator.

Variables and descriptive statistics

Explained variable

The explained variable is energy efficiency, which is measured by total factor energy efficiency (etfp) in this paper. This paper employs the super-efficiency SBM-DEA with the Global Malmquist-Luenberger (GML) index method, as outlined in the study by Shi and Li (2020)³⁰, to analyze the data of 196 Chinese cities from 2004 to 2020. The total factor energy efficiency for the period from 2005 to 2020 was obtained by cumulative multiplication. The input factors selected were labor, capital, and energy. The desired output was the real GDP of the city, measured in ten thousand CNY at constant 2004 prices. The non-desired outputs were sulfur dioxide emissions (measured in tons), industrial wastewater emissions (measured in 10,000 tons), and industrial smoke and dust emissions (measured in tons) of the city’s industrial sector.

Among the aforementioned factors, the labor input is quantified by the number of employees at the end of the year (in 10,000 persons). Capital inputs are estimated using the perpetual inventory method, calculated as Kt = It + (1 - ∂)Kt − 1, where Kt and Kt-1 are the capital stock in period t and t-1, respectively (unit: ten thousand CNY, constant 2004 prices), It is the amount of capital inputs in period t, ∂ is the depreciation rate, and the initial capital stock is in 2004. Energy inputs are selected to measure urban energy consumption. This paper builds upon the work of Hu and Yu (2023)⁵¹ to select three energy sources: total natural gas supply (10,000 cubic meters), total LPG supply (tons), and electricity consumption of the whole society (10,000 kwh). These energy sources are then converted to standard coal consumption in order to measure urban energy consumption (measured in 10,000 tons). The standard coal conversion coefficients are based on those issued by the Ministry of Industry and Information Technology of China. These are 1.33 kg of standard coal/m3 for natural gas, 1.7143 kg of standard coal/kg for liquefied petroleum gas, and 0.1229 kg of standard coal/kWh for electricity.

Explanatory variables

The explanatory variables are NEDPs. This paper employs a dummy variable to indicate whether a city has been approved as a NEDPs ($\:{NEDP}_{it}$). The variable is expressed as the product of the approval time of NEDPs and the approved city. If a city is approved as an NEDPs in a given year, the value is 1; otherwise, it is 0.

Mechanism variables

The mechanism variables encompass environmental regulation (regu), green technology innovation (giov), and industrial structure upgrading (indust).

Environmental regulation (regu) is typically quantified by scholars as the cost per unit of pollutants controlled (Song et al., 2022)⁵². However, this approach is unable to assess the regulatory impact of administrative instruments such as monitoring. This paper draws on the study of Wang et al. (2021)⁵³ and employs an environmental regulation effect-based indicator, namely GDP (unit: ten thousand CNY, constant 2005 prices)/carbon dioxide (CO2) emissions (10,000 tons), as a proxy variable. To date, government departments have not yet compiled data on CO2 emissions at the city level, and there is a problem of missing historical data. This paper adopts the approach of Zhang et al. (2021)²⁷ to measure the CO2 emissions of cities based on the inverse simulation of nighttime total light luminance value data. First, provincial energy consumption data are employed to calculate the CO2 emissions of each province via the methodology prescribed by the Intergovernmental Panel on Climate Change (IPCC). Subsequently, a correlation was established between provincial CO2 emissions and nighttime lighting data (DN values). Finally, the urban CO2 emissions were calculated from the urban nighttime lighting data, based on the concept of the top-to-down method and the weighted average result of the DN values. The basic logic underlying this approach is that the higher the brightness of nighttime lighting, the more nighttime economic activities are occurring, which in turn implies a higher level of economic development and the corresponding energy consumption and CO2 emissions.

Green technology innovation (giov) is measured by the number of green invention patents granted per 10,000 population., as measured by Wu et al. (2023)²², directly reflects the results of R&D and innovation performance. Industrial structure upgrading (indust) is measured by the ratio of the value-added of the tertiary industry to that of the secondary industry, as proposed by Yuan and Xie (2014)⁸.

Control variables

To investigate the net effect of NEDPs on urban energy efficiency and mitigate potential omitted variable bias, a set of control variables was selected based on economic theories and empirical literature concerning the determinants of energy consumption and efficiency, such as Xue and Zhou (2022)³² and Hua and Ye (2023)¹⁸. Level of Economic Development (pgdp): Measured by the natural logarithm of real GDP per capita, measured at constant 2005prices (Unit: CNY). This controls for the scale effect and the different stages of economic development, as postulated by the Environmental Kuznets Curve (EKC). Higher income levels are associated with both increased energy demand and greater capacity to invest in efficient technologies. Consumption Level (spe): Quantified by the ratio of total retail sales of consumer goods (ten thousand CNY) to regional GDP (ten thousand CNY). This variable accounts for the structure of economic activity, as a shift towards a consumption-driven economy may influence the energy intensity of GDP. Population Density (scale): Defined as the population (10,00 persons) divided by the administrative area (square kilometers). This captures agglomeration economies. Denser urban areas may benefit from shared infrastructure and more efficient public transportation, potentially leading to higher energy efficiency, but may also face increased congestion and energy demand. Level of Financial Development (fina): Measured by the ratio of loan balances of financial institutions (ten thousand CNY) to GDP (ten thousand CNY). Well-developed financial markets can alleviate financing constraints for firms, facilitating investments in energy-saving technologies and green innovation. Level of Informatization (nfor): Quantified by the logarithm of the number of international Internet users (in individuals). This controls for the role of the digital economy. Greater informatization can promote smart energy management, optimize production processes, and enable a shift towards less energy-intensive service sectors, thereby enhancing overall energy efficiency.

Descriptive statistics

The descriptive statistics of the variables are presented in Table 1. An examination of the descriptive statistics in Table 1 reveals distributions that are consistent with the real-world economic phenomena. The urban energy efficiency variable (etfp) shows a median of 1.50 and a maximum of 7.57, indicating the presence of cities with exceptionally high energy efficiency. Similarly, the green technology innovation variable (giov) has a mean (0.81) significantly larger than its median (0.27), with a maximum value of 7.27. This right-skewed distribution is expected, as green patenting activity is highly concentrated in a few innovative urban centers. The same pattern is observed for environmental regulation (regu) and economic development (pgdp), where the maximum values are substantially higher than the medians, reflecting the significant disparity in regulatory intensity and economic development levels across Chinese cities. These ‘outliers’ are not data errors but represent genuine heterogeneity within the sample. To ensure the results are not driven by these extreme yet valid observations, this paper performed a robustness check in which all continuous variables were winsorized at the 1 st and 99th percentiles.

Table 1 Descriptive statistical analysis.

Methods

Benchmark regression model

Given that the NEDPs policy was implemented in a staggered manner across different cities and years, it constitutes a typical quasi-natural experiment. The TV-DID is particularly suited for such a setting, as it can effectively address the potential biases arising from heterogeneous treatment timing. This approach is widely employed in similar studies assessing the impact of environmental policies (Hua and Ye, 2023)¹⁸. Therefore, the TV-DID was applied to assess the impact of the NEDPs policy on the city’s energy efficiency. The model was set up as follows:

$$\:{Energ}_{it}=c+{\beta\:}_{1}{NEDP}_{it}+\theta\:{X}_{it}+{\delta\:}_{i}+{\gamma\:}_{t}+{\epsilon\:}_{it}$$

(1)

In Eq. (1), $\:{Energ}_{it}$ represents the energy efficiency of city i in year t. The policy dummy variable, denoted by $\:{NEDP}_{it}$, is the interaction term of the area dummy variable $\:{trea\text{t}}_{i}$ and the time dummy variable $\:{post}_{t}$. During the sample period, if city i is approved as a demonstration park, then $\:{trea\text{t}}_{i}$=1; otherwise, it is 0. In the year of the approval of the demonstration park and thereafter, $\:{post}_{t}$= 1, otherwise it is 0. $\:{X}_{it}$ denotes other variable control variables affecting energy efficiency. $\:{\delta\:}_{i}\:$is the city fixed effect, $\:{\gamma\:}_{t}\:$is the time fixed effect, and $\:{\epsilon\:}_{it}$ is the random perturbation term. The estimated coefficient $\:{\beta\:}_{1}$ represents the difference in energy efficiency between cities that have been approved as NEDPs (the experimental group) and those that have not (the control group).

Parallel trend test method

The fundamental premise of employing the TV-DID method is to satisfy the parallel trend hypothesis test. It is assumed that, in the absence of NEDPs, the trend of energy efficiency in pilot and non-pilot cities should be similar. To rigorously test this assumption, this paper draws on Jacobson et al. (1993)⁵⁴ and employs the event analysis method for parallel trend hypothesis testing, with the model set up as follows:

$$\:{Energ}_{it}=c+{\beta\:}_{k}\sum\:_{\begin{array}{c}k\ge\:-10,\\\:K\ne\:-1\end{array}}^{k=10}{D}_{it}^{k}+\theta\:{X}_{it}+{\delta\:}_{i}+{\gamma\:}_{t}+{\epsilon\:}_{it}$$

(2)

The dummy variable $\:{D}_{it}^{k}$ in Eq. (2) indicates whether year t is the kth year before or after the construction of the NEDPs in city i. If the answer is affirmative, the variable is assigned a value of 1; conversely, it is assigned a value of 0. $\:{D}_{it}^{0}$ indicates whether year t is the year in which NEDPs were constructed in city i. The value of k is used to denote the kth year before the construction of the park, and the value of k is used to denote the kth year after the construction of the park. Given the limited data available in the five years preceding and following the policy implementation, k takes on values between − 5 and 5. To avoid multicollinearity, the year prior to the policy implementation is designated as the base period, with k=−1 excluded from the regression equation. The remaining components of the model are aligned with the baseline model. $\:{\beta\:}_{k}$ reflects the difference in energy efficiency between the treatment and control groups before and after the construction of NEDPs. When the coefficient $\:{\beta\:}_{k}$ is not significant when k < 0, it indicates that the estimation satisfies the parallel trend test.

Mechanism testing model

In order to explore the mechanism of the impact of approved NEDPs on energy efficiency, the following model is constructed in this paper with reference to the study of Jiang (2022)⁵⁵:

$$\:{\:\:\:\:\:\:\:\:Mach}_{it}={\alpha\:}_{0}+{\alpha\:}_{1}{NEDP}_{it}+\eta\:{X}_{it}+{\delta\:}_{i}^{{\prime\:}}+{\gamma\:}_{t}^{{\prime\:}}+{\epsilon\:}_{it}^{{\prime\:}}$$

(3)

In Eq. (3), $\:{Mach}_{it}$ represents the mechanism variables, which encompass environmental regulation, green technology innovation, and industrial structure upgrading. $\:{\alpha\:}_{1}\:$denotes the coefficient of core variables, while $\:{\alpha\:}_{0}$ signifies the constant term. The remaining variables are identical to those in Eq. (1). All econometric analyses were conducted using Stata version 17.0.

Empirical results

Benchmark model regression results

The present study utilised the variance inflation factor (VIF) to examine multicollinearity. The findings suggest that the mean VIF is 1.56, which is well below 10, indicating that multicollinearity issues are not significant. Table 2 presents the regression results for the impact of NEDPs on energy efficiency. Column (1) presents the estimation results without consideration of control variables and fixed effects, while column (2) presents the estimation results controlling for city and year fixed effects. Columns (3) and (4) present the estimation results with control variables added to columns (1) and (2), respectively. The Hausman test rejects the original hypothesis at the 5% significance level, indicating that the estimation results of the two-way fixed effects model are superior to those of the random effects model. Consequently, this paper develops the analysis based on the estimation results of the two-way fixed effects model. The regression coefficients for NEDPs are consistently positive across all scenarios, and they all pass the 1% significance test, indicating that NEDPs significantly contribute to energy efficiency improvement. Therefore, hypothesis 4 is confirmed. As previously stated, NEDPs are not only conducive to the provision of beneficial industrial support policies for the park, but also assist in the reduction of undesired outputs, such as pollutants, the improvement of the energy consumption structure, and the promotion of the transformation of urban industrial ecology and energy efficiency.

Table 2 Benchmark regression analysis.

Robustness test

Parallel trend test

In this paper, the dynamic impact of NEDPs on energy efficiency is tested using model (2), based on the parallel trend test. The results are shown in Fig. 2. Prior to the construction of NEDPs, there was no discernible difference in energy efficiency (the regression coefficient β was not significantly different from 0), which satisfied the parallel trend hypothesis. However, following the construction of NEDPs, the energy efficiencies of the treatment groups exhibited a notable increase in comparison to the control group (regression coefficient β was greater than 0 at the 1% significance level). This indicates that NEDPs have a long-term impact on energy efficiency.

Fig. 2

Parallel trend test.

Placebo test

In order to ascertain whether the impact of NEDPs policies on energy efficiency is not the result of extraneous factors, this paper employs a placebo test to identify the contingency of NEDPs policy effects. As outlined by Shi et al. (2018)⁵⁶, a “pseudo-policy dummy variable” is constructed through random sampling 1,000 times, with regression estimation performed using model (1). The results are presented in Fig. 3. The results indicate that the mean value of the regression coefficients of the “pseudo-policy dummy variables” is close to 0 and considerably smaller than the benchmark regression coefficients. The distribution of the estimated coefficients is approximately normal and is not statistically significant at the 10% level. This indicates that the impact of NEDPs on energy efficiency is not attributable to other random factors, thereby reinforcing the reliability of the aforementioned conclusions.

Fig. 3

Placebo test.

PSM-DID method

The DID method is susceptible to “selectivity bias,” which refers to the inability to ensure that the experimental group and the control group possess identical individual characteristics prior to the implementation of the policy. This phenomenon is particularly prevalent in large sample sizes. Consequently, this paper employs the propensity score matching (PSM) method to re-match the control group with the cities in the treatment group in order to mitigate the sample selection bias. Specifically, the near-neighbor matching method is employed to match control variables such as GDP per capita and population density as covariates, thereby ensuring that the treatment group is not systematically different from the control group. After excluding the few samples that are not matched, the TV-DID model is re-estimated using the TV-DID model and the regression results are presented in column (1) of Table 3. The regression coefficient for NEDPs is 0.4437, which is significant at the 1% level. This indicates that NEDPs can promote energy efficiency. In other words, the original model is not subject to significant sample selection bias issues, and the conclusions drawn are therefore reliable.

Tests for heterogeneous treatment effects

De Chaisemartin and D’Haultfoeuille (2020)⁵⁷ demonstrate that the estimates of the TV-DID model are a weighted average of the treatment effects across all individuals. Negative weights may occur when there are heterogeneous treatment effects, and the model’s estimates may not be robust when negative weights account for a large proportion of the total. To ascertain the robustness of the benchmark regression results under heterogeneous treatment effects, this paper refers to the study of Hu and Yu (2023)⁵¹, which employed the twowayfeweights command of Stata software. The results indicate that among the 409 weights, 403 are positive and 6 are negative. The sum of the positive weights is 1.0049, while the sum of the negative weights is −0.0049, representing a small proportion of negative weights. This suggests that the heterogeneity treatment effect has a minimal impact on the benchmark regression results, which are robust.

The elimination of interference from other policies

China’s practice demonstrates that it is promoting energy conservation, emission reduction, and green low-carbon policies in parallel. In the process of promoting NEDPs as an environmental policy, a number of other environmental policies aimed at enhancing the ecological environment have also been implemented. The National Low Carbon City Pilot Policy is designed to achieve low-carbon development in high-pollution and high-energy-consumption areas by optimizing the energy structure and reducing the proportion of high-carbon industries. Consequently, the national low-carbon city pilot policy is inextricably linked to the present study and must be excluded from the influence of the policies. Since October 2010, the Chinese government has initiated the low-carbon city pilot policy, continuously expanding the scope of the pilot program. To date, three batches of pilot cities have been announced. This paper introduce a dummy variable indicating whether the city was covered by the low-carbon city pilot policy in that year into the model. The results of the regression are presented in column (2) of Table 3. The sign and statistical significance of the coefficients of the NEDPs remain consistent with the baseline regression results after excluding the effect of low-carbon city pilots. This further substantiates the assertion that NEDPs facilitate urban energy efficiency.

Sample data screening process

The first is the “trimming” of the sample data. Given the potential influence of extreme values on the outcomes of benchmark regression, this paper employs a data reduction technique, whereby the sample is reduced by 1%. The resulting re-estimation results are presented in column (3) of Table 3. The results indicate that the positive impact of NEDPs on urban energy efficiency remains significant. Second, the sample interval is modified. The emergence of the Corona Virus Disease (COVID-19)at the end of 2019 had a profound impact on the global economy and a significant negative impact on the urban economy. In order to avoid the potential effects of the COVID-19 on energy efficiency, this paper excludes the sample data for 2020. The re-estimation results are presented in column (4) of Table 3. The regression coefficient for NEDPs is 0.5045, which remains significant at the 1% level. Third, outlier provinces and cities are removed. To enhance the comparability between the treatment and control groups, this paper deletes provinces and cities (including Ningxia, Gansu, Qinghai, Heilongjiang, Hainan, Guangxi, and Chongqing Municipality) that have not been approved for NEDPs during the sample period. Meanwhile, this paper eliminates two additional outliers: Shanghai with 10 parks and Suzhou with 8 parks. Furthermore, municipalities directly under the central government exhibit distinct political and economic characteristics compared to other prefecture-level cities. Consequently, Beijing and Tianjin are excluded from the analysis. The re-estimation results after the exclusion of the outlier samples are presented in column (5) of Table 3. The findings indicate that the positive impact of NEDPs on urban energy efficiency remains statistically significant. Following a series of robustness tests, it can be demonstrated that the estimation results presented in this paper are robust.

Table 3 Robustness test analysis.

Mechanism test

This paper examines the role of NEDPs in affecting energy efficiency from the perspective of environmental regulation, green technology innovation, and industrial structure optimization. The estimation is based on Eq. (3), and the results are presented in Table 4. The estimated coefficients of NEDPs in columns (1) to (3) are all significantly positive at the 1% significance level, indicating that NEDPs significantly enhance the strength of environmental regulation, promote green technological innovation, and facilitate industrial structure optimization and upgrading. And some studies have shown that environmental regulation, green technology innovation and industrial structure optimization and upgrading can effectively promote energy efficiency (Maddison, 1987; Han et al., 2007; Chang and Hu, 2010; Wan et al., 2019; Kong et al., 2021)^{36,41,42,48,50}. It can be concluded that NEDPs can promote energy efficiency through three mechanisms: strengthening environmental regulation, promoting green technology innovation, and industrial structure optimization and upgrading. This verifies research hypotheses H1, H2, and H3.

Table 4 Mechanism test analysis.

Heterogeneity analysis

Heterogeneity analysis of urban industrial characteristics

This paper examines the disparate impact of NEDPs on old industrial base cities and non-old industrial base cities, as defined by the 120 old industrial base cities identified in the National Old Industrial Base Adjustment and Reform Plan (2013–2022). Old industrial bases are typically responsible for supplying major technological equipment or products related to the national economy and people’s livelihood. Their industrial structures are primarily dominated by the secondary industry, which, in general, exhibits significant characteristics of high energy consumption and high pollution. Whether NEDPs can play a significant role in promoting energy efficiency in old industrial base cities is related to whether old industrial bases can realize high-quality development.

This paper employs a quasi-natural experiment to investigate the impact of NEDPs on the energy efficiency of old industrial base cities. The experimental group comprises old industrial base cities with NEDPs, while the control group comprises the remaining old industrial base cities. In order to explore the effect of NEDPs on energy efficiency of non-old industrial base cities, non-old industrial base cities located in NEDPs are also used as the experimental group, and the remaining non-old industrial base cities are used as the control group. As evidenced by the results presented in columns (1) and (2) of Table 5, the impact of NEDPs on energy efficiency in cities with old industrial bases is positive but not statistically significant. The impact of NEDPs on non-old industrial bases is statistically significant at the 1% level. This suggests that the impact of NEDPs on energy efficiency improvement in non-old industrial base cities is more pronounced.

Heterogeneity analysis of urban resource endowment

The resource endowment on which urban development depends is the basis for influencing urban energy utilization. This paper examines the heterogeneous impact of NEDPs on energy efficiency from the perspective of differences in urban resource endowments. This paper is based on the Circular of the State Council on the Issuance of the National Sustainable Development Plan for Resource-based Cities (2013–2020), which divides the 196 sample cities into 66 resource-based cities and 130 non-resource-based cities. The estimation results are shown in columns (3) to (4) of Table 5. The results indicate that the effect of NEDPs on energy efficiency in non-resource cities is significantly positive, whereas the coefficient on energy efficiency in resource cities is positive but insignificant. This suggests that NEDPs contribute to energy efficiency in non-resource cities, but not in resource cities.

Heterogeneity analysis of the construction stage of NEDPs

The List of NEDPs , issued by China’s Ministry of Environment and Ecology in 2017, categorizes NEDPs into two types: those that have been approved as NEDPs and those that are under construction. Those under construction are required to complete the indicators of economy, environment, and energy use according to the plan. Upon reaching the criteria for the construction of NEDPs and receiving regular assessment and evaluation by the national government, they can be approved as official NEDPs. The two types of parks may differ in their initial energy transition, pollution reduction, and carbon reduction. Additionally, the length of the park construction time may influence the effect. To this end, the study test for heterogeneity in the construction stage of NEDPs and the results are presented in columns (5) and (6) of Table 5. The results are presented in columns (5) and (6) of Table 5. The results indicate that the NEDPs coefficient for cities that have been approved for the park is 0.0199, which is statistically significant at the 1% level. Cities that are in the process of constructing a park have a positive but insignificant impact coefficient. This suggests that the approved parks can significantly enhance urban energy efficiency, whereas the parks in the construction stage have not yet demonstrated an effective impact.

Table 5 Heterogeneity analysis.

Discussions

Discussion on influential relationships

The empirical analysis provides robust evidence that the implementation of NEDPs significantly enhances urban energy efficiency. This finding not only reinforces the growing body of literature on the positive environmental outcomes of green location-oriented policies but also extends it by offering a nuanced understanding of the energy efficiency mechanism. While existing research has demonstrated that NEDPs contribute to urban green innovation (Wu and Gao, 2022)²¹, reduce carbon emissions (Hua and Ye, 2023)¹⁸, lower industrial sulfur dioxide emissions (Chen et al., 2023)²⁰, and promote sustainable urban development (He and Li, 2023)²³, this study shifts the focus to energy efficiency as a central channel through which NEDPs achieve broader environmental-economic synergies. The present findings thus bridge a critical gap between studies focusing on specific pollutant reductions and those examining broader economic outcomes, positioning energy efficiency as a linchpin that connects these two dimensions.

Importantly, the results highlight a distinct policy-driven pathway that is complementary to conventional determinants of energy efficiency, such as, industrial structure (Jenne and Cattell, 1983; Wang et al., 2021)^53,58, renewable energy consumption (Caglar et al., 2024e)⁵⁹, and technological progress (Chen et al., 2021)³⁹. Unlike these traditional factors, which often evolve gradually and remain subject to market path dependency, NEDPs represent a targeted policy intervention capable of accelerating efficiency improvements—even in contexts where typical drivers are weak. This perspective challenges the conventional understanding that energy efficiency improvements stem solely from market forces or endogenous technological change, thereby enriching research on the political economy of energy transition.

This research is particularly salient in the context of China’s ongoing efforts to reconcile economic growth with energy conservation and emission reduction. Previous development strategies, though successful in stimulating rapid economic expansion, have also led to substantial environmental degradation and excessive energy consumption, with estimated economic costs of pollution ranging from 8% to 15% of GDP (Han and Hu, 2015)⁶⁰. Compared to developed economies, China continues to face the dual challenge of fostering high-quality economic development while overcoming structural dependence on energy-intensive industries. In this regard, improving energy efficiency is widely recognized as a crucial lever for harmonizing economic, energy, and environmental objectives (Kazemzadeh et al., 2024)⁶¹. The success of NEDPs in this context offers a potentially valuable model for other large, industrializing economies, such as India and Vietnam, which face similar challenges of locked-in energy infrastructure and rapid urban-industrial growth. The key for generalization lies in adapting the core principles of integrated planning, strict environmental standards, and innovation incentives to different institutional settings, rather than replicating the policy verbatim.

By rigorously establishing the causal link between NEDPs and elevated urban energy efficiency, this study provides timely and policy-relevant insights into how traditional development zones can serve as experimental arenas for sustainability transitions. It also contributes to the theoretical discourse on the antecedents of energy efficiency by highlighting the role of integrated green and location-oriented industrial policies. Rather than viewing economic and environmental goals as incompatible, the study findings suggest that well-designed policy instruments such as NEDPs can facilitate a transition toward both greener and more efficient production systems. Thus, this study not only advances academic understanding of the mechanisms behind location-oriented environmental policies but also offers empirical support for scaling and optimizing such policies in China and other industrializing economies facing similar developmental trade-offs.

Discussion of the mechanism

This paper builds upon and extends the existing literature on green location-oriented policies by elucidating the mechanistic pathways through which NEDPs enhance urban energy efficiency. While prior studies have established a positive correlation between NEDPs and environmental performance (He and Li, 2023)²³, the analysis delves deeper, identifying and validating three synergistic internal mechanisms: enhanced environmental regulation, spurred green technology innovation, and guided industrial structure optimization.

Firstly, the environmental regulation intensification mechanism aligns with the theory of internalizing environmental externalities, which posits that the negative externalities of environmental pollution can be transformed into internal costs for enterprises by setting stringent pollutant emission and energy consumption standards. This incentivizes companies to adopt more efficient energy utilization practices to achieve compliance. NEDPs operationalize this by setting stringent unit energy consumption carbon emission targets and implementing differentiated environmental policies. This creates a binding constraint, forcing firms within parks to abandon extensive energy use patterns and prioritize energy conservation. This finding is consistent with the results of other studies on environmental regulations (Dzwigol et al., 2023; Shi and Li, 2020)^28,30, but it emphasises the unique advantage of green location-oriented industrial policies—namely, the implementation of environmental policies that combine constraints with incentives. Such policies can simultaneously promote economic development and prevent environmental pollution.

Secondly, the present study provides empirical validation of the pivotal mediating function of green technological innovation in the process of NEDPs enhancing energy efficiency. In comparison with preceding studies, the present paper further reveals that NEDPs, as a comprehensive environmental governance instrument, possess unique advantages in promoting green technological innovation that distinguish them from traditional command-and-control or market-incentive policies. Specifically, NEDPs have been shown to significantly reduce the uncertainty and costs associated with green technological innovation by building industrial symbiosis networks, facilitating knowledge and technology spillovers among enterprises, and providing targeted R&D collaboration platforms. From a mechanistic perspective, this finding emphasizes the pivotal role of NEDPs as a comprehensive environmental governance policy in fostering green technological innovation. Furthermore, it provides essential policy insights for China and other developing countries that are pursuing a green transformation through the implementation of pilot park models.

Thirdly, the industrial structure optimization mechanism functions through dual channels: upgrading within traditional sectors and cultivating new low-carbon industries. NEDPs enforce strict market access based on energy efficiency benchmarks, phasing out backward production capacity while attracting high-value, low-energy-consumption industries like green manufacturing and modern services. This transforms the park’s economic fabric, significantly lowering the aggregate energy intensity. This structural shift embodies the concept of “green productivity” and demonstrates how industrial policy, when aligned with environmental goals, can fundamentally redefine a region’s economic trajectory towards less energy-dependent activities.

Importantly, these three mechanisms do not operate in isolation but are deeply interconnected and mutually reinforcing within the NEDPs framework, forming a virtuous cycle. Stringent environmental regulations create the demand pull for green technological innovations, which in turn enable and reduce the costs of industrial upgrading. Conversely, a more advanced industrial structure possesses greater capacity and resources to comply with regulations and invest in innovation. This systems perspective, often underexplored in discrete policy analyses, is a crucial contribution of the study. In practice, numerous provinces and cities in China have not yet been approved for NEDPs, and there is considerable scope for expanding the policy’s coverage.

Discussion on heterogeneity

The heterogeneity test section indicates that NEDPs facilitate energy efficiency in non-old industrial base cities, but do not play a significant role in old industrial base cities. One possible explanation for this discrepancy is that old industrial bases are dominated by high-polluting and high-energy-consuming industries, and there is limited space for the reduction of pollutant emissions and energy intensity in the short term. The NEDPs provide incentives to enterprises producing major technological equipment to invest funds in energy-saving technology research and development. However, the path-dependence of the production technology makes the effect of the energy efficiency improvement limited. The majority of non-old industrial bases are cities with more developed economies and more rational industries. These cities have greater demands for environmental quality and economic development, and therefore have higher incentives to participate in NEDPs. This ultimately facilitates the realization of urban energy efficiency. This heterogeneity underscores that the “one-size-fits-all” policy implementation is unlikely to be effective. It calls for a more tailored approach, a lesson that is highly relevant for other large countries with significant regional disparities, such as Brazil or Indonesia, when designing their own regional development policies.

Secondly, NEDPs have a positive effect on energy efficiency in non-resource cities, but no effect on resource cities. The theoretical rationale for this assertion is that the industrial structure of resource cities is predominantly based on the secondary industry, which is highly polluting and energy-consuming. This sector is facing significant challenges in terms of pollutant emission and energy consumption. Concurrently, resource cities are endowed with substantial coal reserves and other resources, and the cost of utilizing energy is relatively low. The lack of motivation on the part of enterprises to engage in green technological innovation and energy saving and emission reduction is a significant obstacle to regional energy efficiency. Enterprises in non-resource cities have relatively high energy costs, and as a result of the constraints of profit maximization, most of them pay more attention to saving energy consumption or reinforcing investment in research and development of energy utilization technologies. Given the objective technical differences between enterprises, the constraints of energy costs provide a strong incentive for most enterprises to improve energy utilization technology. This can reduce corporate energy consumption and pollution emissions. Furthermore, there is a spillover effect of technology. Small-scale enterprises with limited research and development capabilities can also achieve green production and reduce energy consumption due to technology spillover. Consequently, the implementation of NEDPs policy is more beneficial for non-resource cities in terms of enhancing energy efficiency. This finding is consistent with the findings of Zhang et al. (2021)²⁷, who examined the impact of the pilot low-carbon city policy on energy efficiency. Their findings indicated that the policy had a significant impact on non-resource cities, but not on resource cities. This consistent pattern across different environmental policies in China highlights the profound challenge of overcoming resource curses and structural inertia. The policies for resource-based cities may need to be coupled with stronger regional transition assistance and diversification strategies to be effective.

Additionally, the study findings indicate that NEDPs in the construction phase did not result in a notable improvement in energy efficiency. One potential explanation for this is that the policy advantages of institutional innovation and stress testing have not yet been fully developed and utilized in the parks that are under construction. The relative weakness of the relevant eco-construction investment, technology market maturity, and green technology innovation capacity at this stage makes the parks less obvious in promoting energy efficiency. Nevertheless, at least in terms of the direction of influence, NEDPs in the construction stage can also positively influence energy efficiency. The effects of such policy interventions will gradually become apparent over time, requiring policymakers to maintain a long-term commitment.

Limitations and outlook

Limitations of the study

Enhancing energy efficiency is a key way of aligning economic, energy and environmental objectives. Although NEDPs have had a positive impact on improving energy efficiency, how to further incentivize traditional development zones to transition into NEDPs remains an urgent issue that needs to be addressed. Another concern is its potential to create a “green bubble” within designated zones, which could inadvertently lead to the displacement of pollution-intensive activities to neighboring non-NEDPs cities, thereby exacerbating regional inequalities in environmental quality and economic development. This spillover effect, whether positive or negative, warrants thorough investigation to assess the net environmental impact of the policy at a regional scale. Furthermore, the current policy framework may place disproportionate emphasis on top-down regulatory measures, underutilizing market-based incentives and failing to fully engage the enthusiasm of key stakeholders, including private enterprises and local communities.

Beyond the limitations of the aforementioned policies, this study still has some shortcomings. For example, the analysis focuses primarily on the medium-term effects of the policies; the long-term sustainability of these efficiency gains and their persistence beyond the policy’s initial implementation phase remain open questions. Additionally, the generalizability of the conclusions is also limited to the Chinese context, and their applicability to other countries with different institutional structures and economic conditions requires further verification.

Potential directions for future research

Future studies should prioritize several avenues. First, examining the long-term dynamics of NEDPs’ effects on energy efficiency and carbon reduction is crucial for evaluating their sustained efficacy. Second, research should investigate the policy’s spillover effects on surrounding non-pilot cities to understand its broader regional implications. Third, comparative international analyses are needed to explore the transferability of the NEDPs model to other developing economies. Finally, scholarly and practical attention should focus on how to enhance the policy’s design. This includes: establishing differentiated energy-saving and emission-reduction standards tailored to local resource endowments, better integrating market-based instruments to stimulate participant engagement, and effectively harnessing scientific and technological innovation to foster strategic emerging industries (e.g., new energy, advanced manufacturing) within the parks. These are significant theoretical and practical matters essential for optimizing the NEDPs policy framework.

Conclusions and recommendations

Conclusions

This study contributes to the sustainability policy literature by grounding theoretical frameworks like the Porter Hypothesis in the Chinese context, providing robust empirical evidence from its location-oriented industrial policies. The present study utilised a panel data set encompassing 196 Chinese cities from 2005 to 2020, employing a TV-DID approach for analysis. The analysis establishes NEDPs as a potent instrument for enhancing urban energy efficiency, confirming a statistically robust positive impact with an average effect size of 0.5644. Robustness tests (including parallel trend, placebo test, and substitution of core explanatory variables) further confirm that this causal relationship is free from endogeneity interference such as sample selection bias, ensuring the reliability of the quantitative results. The core contribution, therefore, extends beyond quantifying this average treatment effect to uncovering the mechanisms and contingencies underlying it.

Firstly, the present study provides detailed insights into the causal mechanisms underpinning the policy’s effectiveness. The present study moves past the “black box” of policy impact by theoretically articulating and empirically validating a tripartite transmission channel: the stringency of environmental regulation, the inducement of green technological innovation, and the facilitation of industrial structure upgrading. The mediation analysis confirms that environmental regulation (β = 0.2164, p < 0.01), green technology innovation (β = 1.0531, p < 0.01), and industrial structure upgrading (β = 0.0894, p < 0.01) all serve as statistically significant channels. This mechanistic evidence serves to reinforce the theoretical underpinnings of the Porter Hypothesis within the specific context of place-based environmental policy, thereby demonstrating that the implementation of well-designed regulations can indeed catalyse innovation and efficiency gains that offset compliance costs. Concurrently, the present findings offer a refined interpretation of the Pollution Haven Hypothesis by demonstrating that within a substantial, internally heterogeneous nation such as China, the implementation of rigorous green policies can effectively attract clean industries and innovation factors to specific regions, as opposed to the indiscriminate dispersal of these elements.

Secondly, the heterogeneity analysis reveals the critical role of local initial conditions in policy outcomes, with statistical differences across city types explicitly supporting contextual adaptability of the policy. The finding that the positive effects are more pronounced in cities that are not old industrial bases, non-resource-based, or with established parks, underscores that the success of green industrial policies is not automatic. Specifically, the regression coefficient of NEDPs on energy efficiency in old industrial bases is 0.1386 (insignificant at the 10% level), while in non-old industrial bases it reaches 0.3715 (significant at the 5% level); in resource-based cities, the policy effect is 0.0714 (insignificant at the 10% level), whereas in non-resource-based cities it is 0.5988 (significant at the 1% level). Additionally, the policy effect in mature NEDPs (coefficient: 0.9598, significant at the 1% level) is significantly stronger than that in under-construction parks (coefficient: 0.1394, insignificant at the 10% level), reflecting the cumulative effect of policy implementation. These statistical results confirm that policy effectiveness is contingent upon the pre-existing industrial structure, resource endowment, and institutional capacity. This underscores a salient policy quandary: the regions exhibiting the most urgent need for energy efficiency enhancement (i.e., lagging industrial and resource-dependent cities) may concurrently be the least equipped to actualize these gains within the confines of a standardized policy framework. This necessitates a fundamental shift from a one-size-fits-all approach to a more nuanced, context-sensitive implementation strategy.

In summary, the broader theoretical implication of this study is that the efficacy of location-oriented industrial policies hinges on their embeddedness within local socio-economic contexts and their architectural design that actively stimulates specific micro-level pathways. The NEDPs model, as analyzed, demonstrates that marrying stringent environmental standards with mechanisms that foster innovation and structural change can create a virtuous cycle. This conclusion is firmly supported by the empirical relationships established in our discussion: the identified mechanisms and the systematic heterogeneity in policy effects. This offers a replicable, yet adaptable, framework for other industrialising nations grappling with the dual challenges of economic development and energy sustainability. Future research should focus on longitudinal studies to track the long-term evolution of these effects and on refining the policy instruments to better support cities with less favorable initial conditions, thereby enhancing the inclusivity and overall effectiveness of the green transition.

Recommendations

This study finding clarifies the mechanistic pathways through which green location-oriented industrial policies influence energy efficiency, and offer empirical evidence for governments to accelerate NEDPs construction deliberately. Policymakers can focus not just on policy designation but on actively fostering these three reinforcing pathways, to maximize the energy efficiency gains of such location-oriented industrial policies. The following specific policy recommendations are hereby proposed.

Firstly, incentives and constraints should be used to increase the motivation of traditional development zones to build NEDPs. In practice, a significant number of Chinese provinces and cities have yet to be designated as NEDPs, indicating that there is considerable scope for extending the scope of the policy. The declaration and creation of NEDPs in traditional development zones should be included in the assessment system of local governments. Specifically, the central government can initiate a “Green Transformation of Development Zones” campaign, setting a target to double the number of NEDPs within five years. Provincial governments should then be required to submit annual plans outlining which specific development zones will apply for NEDPs status and the timeline for doing so. The performance evaluation system for local officials should include specific weights for the successful establishment and energy efficiency performance of NEDPs. Moreover, the objective of enhancing energy efficiency during the construction of NEDPs should be underscored. it is necessary to establish more rigorous criteria for entry into parks, and regulate the number of high-polluting and high-energy-consuming enterprises. Meanwhile, the park should proactively cultivate and introduce new energy, new materials, advanced manufacturing, electronic information, and other strategic emerging industries and future industries.

Secondly, given the heterogeneous effects of NEDPs, government should design differentiated and adaptive NEDPs policies tailored to the resource endowments and industrial foundations of different cities. For non-resource-based cities and non-old industrial bases, support should emphasize accelerating the approval and development of new NEDPs, incorporating energy efficiency targets into local government performance evaluations, and establishing clear energy-intensity benchmarks. In contrast, for resource-based cities and old industrial bases, policies should focus on overcoming structural lock-ins through tailored transitional assistance. This includes establishing special green transition funds, and offering preferential interest loans for energy-saving technological transformation. Under external shocks—such as energy market fluctuations—these regions should receive enhanced fiscal buffer support to avoid rebound pollution or economic decline during transition periods. Furthermore, it is imperative to vigorously develop low-energy consumption, low-pollution high-tech industries and modern service industries in order to facilitate the transformation of resource-based cities and old industrial base cities.

The specific measures to be implemented should include. It is recommended that the Ministry of Ecology and Environment, in conjunction with the National Development and Reform Commission, publish a Guideline for Differentiated NEDPs Development. The proposed guideline would categorise cities according to their industrial structure and resource dependence, employing a system of classification that includes the designations “Advanced,” “Catching-up,” and “Transitional.” It is proposed that each category be accompanied by a bespoke set of support policies, application criteria, and key performance indicators (KPIs). In the context of “transitional” cities (for example, those with a strong focus on natural resource-based economic activities), the guideline stipulates the establishment of a designated “Transition Office”. The primary mandate of this office would be the management of a special fund designated for the transition to a green economy, in addition to the coordination of access to preferential loans from state-owned banking institutions.

Thirdly, policy tools should strengthen market-based incentives and innovation support under uncertainty. On one hand, introduce differentiated green credit support and carbon emission reduction-linked tax incentives to encourage enterprises within NEDPs to invest in green technology R&D and application. Specifically, the “NEDP Green Innovation Voucher” program is to be implemented in selected industrial parks. In accordance with the provisions of the programme, small and medium-sized enterprises (SMEs) within NEDPs are permitted to submit applications for vouchers that are intended to provide financial assistance to a proportion of the cost of contracting R&D services from universities or accredited research institutes for energy efficiency projects. Concurrently, the China Banking and Insurance Regulatory Commission (CBIRC) could issue a directive encouraging banks to offer a 0.5–1.0.5.0% point interest rate discount on loans for projects aligned with the NEDP’s green technology catalog. On the other hand, the park should enhance the effectiveness of regulatory services, improve the business environment and innovation environment, attract high-quality innovative talents, technologies, capital and other factors to form a cluster in the park, and improve the efficiency of green innovation. Furthermore, establish a dynamic monitoring and early warning mechanism for emerging industries, enabling industrial support strategies to respond swiftly to external shocks such as supply chain disruptions.

Finally, government should promote regional synergy and spillover management. It is recommended that the establishment of joint mitigation mechanisms between NEDPs and non-NEDPs cities be encouraged. Specifically, launch a formal “Inter-City Green Partnership” program, managed by the National Development and Reform Commission. The program would facilitate the signing of partnership agreements between an NEDPs city and a neighboring non-NEDPs city. The agreement would include specific clauses on technology transfer, joint R&D projects focused on energy efficiency, and a framework for sharing best practices in environmental regulation. The central government could provide matching grants for collaborative projects under these partnerships. On the other hand, the development of “green partnership” initiatives between resource-based and non-resource-based cities is recommended to foster industrial complementarity. Such mechanisms may take the form of cross-regional carbon emission trading and energy efficiency compensation programmed.