Impact of Resource-Saving and Environment-Friendly Society Construction on Sustainability

Abstract

Promoting coordinated and environmentally sustainable development of the Chinese economy is one of the primary tasks at the moment, as well as one of the sustainable development goals of the United Nations. The Resource-saving and Environment-friendly Society (RES) has created a pilot promotion policy aimed at China, with the goal of supporting the sustainable development of economic production activities while preserving natural ecosystems. In this investigation, we used the global Malmquist–Luenberger index with a slack-based measure of the directional distance function to calculate the industrial green total factor productivity (IGTFP) of 105 prefecture-level cities along the Yangtze River from 2004 to 2019; IGTFP was used as a proxy for sustainable development. Then, by considering the RES construction as a quasi-natural experiment, we used propensity score matching difference-in-differences (PSM-DID) to determine the effect of RES construction on sustainable development of the Yangtze River economic belt. We also employed the mediating effect model and triple difference (DDD) model to further analyze the mechanisms underlying the heterogeneous impacts of different urban characteristics. The results revealed three key findings. (1) According to the IGTFP calculation results, RES construction can effectively promote green technological progress while inhibiting green technological efficiency. (2) After carrying out a series of robustness tests, we found that RES construction increased the IGTFP of pilot cities by 4%, indicating that RES construction can significantly promote the sustainable development of pilot cities along the Yangtze River. (3) The results of the mechanism analysis showed that RES construction had a significantly positive impact on sustainable development through technological innovation, human capital, energy conservation, and emission control. In terms of different urban characteristics, the RES construction promoted sustainable development in first-tier cities, second-tier cities, and resource-based cities. We summarized the practical experience of the RES construction as a typical pilot promotion policy. It provides an empirical basis for planning the construction of eco-friendly urban areas in the 21st century and responding to the international community’s sustainable development goals from a Chinese perspective.

1. Introduction

Faced with limited resources and laws to protect the environment from economic overdevelopment, both the international community and the Chinese government are attempting to create sustainable development policies to overcome these obstacles [1]. In 2015, the UN2030 agenda for sustainable development presented seventeen sustainable development goals (SDGs) [2]. The SDGs addressed a range of economic, environmental, and social factors known to be involved in sustainable growth planning [3]. The SDGs were designed to be universally applicable to both developing and developed countries. While an industry-centered country may focus on SDGs related to ecological changes and how they affect the human living environment, a developing country may prioritize different goals, such as reducing pollutant emissions and improving the health system and quality of life. At the end of 2007, China proposed the construction of a resource-saving and environment-friendly society (RES), which sought to create a harmonious and sustainable relationship between human activities and natural ecosystems. Furthermore, RES construction corresponded to the eleventh sustainable development goal (SDG11) and the fifteenth sustainable development goal (SDG15), which called for the development of a sustainable human community and the promotion of a sustainable terrestrial ecosystem. Thus, RES construction is a beneficial and innovative Chinese practice for achieving sustainable development. The study of RES policy not only provides data on China’s SDGs, but can also serve as a model for other regions and countries.

The Chinese government’s pilot promotion policy follows the logic of ‘from point to surface’. The Wuhan City circle and Chang-Zhu-Tan city group were identified as reformed pilot areas and were granted policy innovation privileges during the RES construction process. To coordinate the economic and social development with people, resources, and the environment, the RES policy demanded that reformed pilot areas accelerate the transformation of their economic development based on their own experience. It was incumbent upon them to create a new mode of industrialization and urbanization development different from the traditional one. After that was done, the best pilot work would be incorporated into the central government’s policies, which were then promoted nationally. At the Symposium on the Development of the YREB in 2014, Chinese President Xi Jinping maintained that it was essential to improve the ecological environment in the YREB and make it an eco-friendly showpiece with clean water, green land, and unpolluted air. With the government focusing more on the construction of an ecological civilization, formulating policies through pilot projects and promotions not only helped to evaluate the effects of RES construction, but also provided beneficial policy enlightenment for promoting the sustainable development of the Yangtze River economic belt.

The YREB includes the Yangtze River’s eastern, central, and western areas, including eleven provinces and cities (Figure 1). It contains >40% of China’s population and gross domestic product and is crucial to the socioeconomic development of the country [4]. In recent years, the YREB’s rapid economic development has been primarily dependent on the expansion of the energy sector, the heavy chemical industry, and other industries [5]. According to the National Bureau of Statistics’ 2017 annual report, the industrial value added of the YREB above the designated size was approximately CNY 14 trillion, representing 50% of the country’s GDP. However, the rapid economic development of the YREB came at the expense of the over-exploitation of resources and the accumulation of environmental debts [6]. Facing the dual pressures of slowing economic growth and aggravating environmental pollution under China’s policies with respect to the ‘new normal’, it became increasingly clear that the YREB needed to transform into a greener economy with workable environmental regulations for sustainable development.

Sustainable development was first proposed by the United Nations Commission on Environment and Development in 1987, for the purpose of coping with the challenges of increasing energy scarcity and pollution. According to the neoclassical growth theory, if no technological progress and efficiency improvement is assumed, extensive economic growth dominated by factor expansion will face the problem of diminishing returns of scale. Only intensive economic growth with continuous improvement of productivity is sustainable [7,8,9]. Total factor productivity (TFP) is an additional output index with fixed input that also serves as a source of output growth. It is not only an indicator of growth efficiency or input growth, but also represents the point at which the government should formulate development policies and change the mode of economic growth. Due to limited energy supply and environmental carrying capacity, some studies introduced energy consumption and ecological influences as new input factors in traditional TFP analysis [10,11], which makes green total factor productivity (GTFP) an important variable in research on sustainable development.

How to improve a country’s overall prosperity under the restriction of natural capital has emerged as a critical component of achieving sustainable development. Environmental regulation is an important tool for governments to use in implementing sustainable development. Many studies have shown that environmental regulation affects not only the ecological environment of cities [12], but also their economic growth and social development [13,14]. RES construction has more than ten years of working experience as an important practice in promoting sustainable development in China. Our research was aimed at answering the following questions: (1) How can the effectiveness of the RES construction policies be assessed? (2) Can RES construction facilitate the sustainable development of the YREB? (3) What are the underlying mechanisms of influence? (4) How heterogeneous are the effects of RES construction on sustainable development?

In this paper we treated the RES construction in China as a quasi-natural experiment, and the propensity score matching difference-in-differences (PSM-DID) method was used to measure the impact of RES construction on sustainable development. First, this study applied the global Malmquist–Luenberger index using the slack-based measure of directional distance function (SBM-GML) to calculate the IGTFP of 105 prefecture-level cities along the Yangtze River from 2004 to 2019 and reorganize it into a green technology progress (GTP) index and a green technology efficiency (GTE) index. Second, we used the PSM-DID method to assess the impact of RES construction on sustainable development, based on the novel perspective of measuring sustainable development with IGTFP. Third, the mediating effect model was used to identify the factors that influenced RES construction and sustainable development. Lastly, this paper applied the triple difference (DDD) method to assess the heterogeneous influence of different types of cities on sustainable development during the RES construction process.

The remainder of this study is structured as follows: Section 2 provides a literature review. Section 3 introduces the hypotheses and framework for analyzing the mechanism. Section 4 presents the methods, research scope, and data sources. Section 5 reports empirical results. Section 6 focuses on the intermediary mechanism analysis and heterogeneous analysis. Section 7 contains the discussions, conclusions, and policy suggestions.

2. Literature Review

Since the idea of GTFP was advanced, it has been widely used to assess the coordinated development of the economy and the environment [15,16]. The core idea of GTFP centers on low input, low pollution, and high income, and the existing research primarily focused on three main perspectives. The first aspect was the method of calculation. Since the traditional DEA model could not overcome the problem of variable relaxation in radial selection, Tone (2001) proposed a non-radial, non-angular, slack-based measure (SBM) method [17]. Following that, Fukuyama and Weber (2009) combined the SBM method with the directional distance function (DDF), which gave a more accurate reading of technical efficiency [18]. To evaluate the dynamic information from the movement of the production frontier, the Malmquist–Luenberger index method was required to determine the GTFP index. Oh (2010) proposed a global Malmquist–Luenberger (GML) index which could avoid the non-transitivity defects of the traditional ML index [19]. Because of the accuracy of the SBM model and the GML index, it has been widely used in recent years for the measurement of GTFP [20,21]. The second aspect is research dimension. Early research on GTFP mainly focused on micro-fields such as products and enterprises [22], but failed to address the externalities. When the research progressed to the macro-level, we were able to examine the temporal and spatial evolution characteristics and the effective aspects of GTFP [23,24]. The third aspect is the purpose of the research. Scholars typically use GTFP to assess green development and high-quality economic development [25,26].

Sustainable development is measured by constructing an evaluation index system [27,28] and efficiency measurement method [29,30]. Economic development, social harmony, and environmental quality improvement were the most important elements in constructing the evaluation index system [31]. In addition, studies also found that economic level, industrial structure, government intervention, openness to global markets, and other factors had a substantial effect on promoting regional sustainable development and improving resource utilization efficiency [32,33]. In this study, we used the industrial green total factor productivity (IGTFP) index to measure sustainable development. There were three reasons for doing this. First, given its clear association with economics, the purpose of strategic planning is to achieve sustainable growth of economic output with limited resources and environmental carrying capacity. In other words, the organization can increase production efficiency by the application of advanced production technology and the improvement of management methods and concepts, including the related concepts of TFP and GTFP. To an economist, sustainable development involves the improvement of TFP. Given the limited environmental capacity and energy reserves, the problem should also involve the improvement of GTFP. Second, IGTFP refers to the total efficiency of resource utilization of industrial enterprises after accounting for various factors such as capital, labor, land, energy, etc., under environmental constraints. IGTFP considers industrial development, environmental quality, and economic development, with economic development quality preceding economic growth speed [34,35]. It can be viewed as a kind of compromise between further economic growth and environmental conservation, which is consistent with the goals of strategic sustainability. Third, relying on the golden waterway, the industry in the YREB has developed rapidly in recent years, and a relatively complete industrial system has been established. However, the extensive industrial development has resulted in a series of issues, including tight resource constraints and environmental pollution, which have seriously hampered YREB’s sustainable development. As a result, we chose IGTFP as the index for assessing sustainable development, which is critical for evaluating the process of sustainable development in China from a fresh perspective during this new period.

As a concrete environmental policy, RES construction aims to help people realize the importance of sustainable development of the environment, the economy, and the society. The enactment of this policy involved the mutual influences of ecology, economy, society, and politics, which showed the complex interactions, variability, and uncertainty in practice. RES construction exhibits the characteristics of decision making in the post-normal scientific era because of the multiplicity of factors among individuals, groups, society, and nature [36]. Ravetz and Funtowicz (1999) believed that “quality” was a relevant and stable guiding principle for decision making [37]. We should pay more attention to “quality” as the core evaluation criterion when making environmental decisions and evaluating environmental policies. Following this principle also emphasizes the appropriateness of choosing industrial green total factor productivity as a proxy for evaluating sustainable development. Furthermore, the ecological economics involved in RES construction, as a post-normal science, has further enhanced the complexity and uncertainty of the system. When solving environmental problems, we need to rely on managers, policy makers, the public, and others involved in decision-making issues, as well as the knowledge and information brought by these people [38,39]. Evaluating the influence of RES construction on sustainable development actually involves the intersection of ecology, economics, sociology, and other disciplines. Policy formulation is oriented towards the whole society, and the promotion of public awareness of environmental protection is an integral part. Therefore, using mathematical models to evaluate RES construction is actually a multi-dimensional evaluation method [40]. Applying this kind of policy evaluation for resolving the conflict between environment and economy has become very popular and an essential element of the research field.

In recent years, the Chinese government has actively promoted sustainable development through its pilot promotion policies. Evaluation of the effectiveness of the pilot promotion policies, such as the low-carbon city pilot policy [41,42], smart city construction [43], and free-trade zone pilot reform policy has become a hot topic [44]. However, little research has been conducted on the policy effect of RES construction and its mechanism. To examine the effects of RES policy implementation, early studies simply compared the economic or environmental data before and after policy implementation [45,46]. Later researchers constructed an index system to evaluate the policy’s effectiveness [47,48]. However, because these methods clearly could not accurately identify the causal relationship between RES construction and economic or environmental indicators, the evaluation results should be reconsidered. Few scholars treat the RES policy as a quasi-natural experiment, employing a DID model to explore the economic effect and green ecological effects of the RES activities [49]. Our investigation determined the effects of RES construction on sustainable development using a PSM-DID method, because the PSM method can resolve the heterogeneity of urban samples and the endogenous problems of models. This method has been extensively used for policy assessment [50,51].

Few studies were found that used quantitative methods to investigate the mechanism underlying the relationship between RES construction and sustainable development. As a result, there was a limited group of reports addressing the mechanism of influence of other pilot promotion policies. Qiu et al. (2021) claimed that the low-carbon city pilot policy significantly promoted urban green development by a combination of innovative technology, industrial restructuring, resource reallocation, energy saving, and sequestering carbon [52]. Jiang et al. (2021) argued that smart city construction could improve GTFP by boosting urban technological innovation [43]. Zhou et al. (2022) found that the free-trade zone had an important part to play in emphasizing high-quality green development by attracting foreign investors, enhancing the growth of green production [53]. Furthermore, other studies have suggested that government intervention and human capital could be potential influencing mechanisms for such policies. Due to the complexity of these mechanisms and the fact that they are not directly observable, the conclusions from previous papers have often been inconsistent. We examined the available literature to find the mechanisms with the greatest potential for having a significant effect on sustainable development based on economic theory and empirical analysis.

We summarized the effects of China’s major national development projects on sustainable development, in order to analyze the relationship between these projects and the selected impact variables. For example, Long et al. (2020) found that economic development, government support, and industrial structure upgrades were the driving forces in the Yangtze River economic belt strategy for directly enhancing green technology innovation [54]. Kang et al. (2018) discovered that China’s foreign direct investment could be significantly boosted by the “one belt one road initiative” [55]. Fang et al. (2018) found that the equalization of per capita fiscal expenditure, the optimization of investment structures, and the establishment of an ecological compensation mechanism could promote collaborative and sustainable development in the Beijing-Tianjin-Hebei region [56]. It was found that economic development, government support, industrial structure, foreign investment, and fiscal expenditure were important variables that affected the sustainable development of important national development projects, and which also provided a basis for the selection of variables in the following empirical analysis.

For multi-scale policies, the environmental management of settlements is also an important concern. China’s RES pilot policy requires that urban construction should achieve a reasonable urban spatial layout through scientific urban planning, and improvements to public infrastructure and urban public service capacity, so as to better meet the basic living needs of urban residents and promote sustainable development. Foreign scholars have also conducted extensive research on this subject. Macfarlane et al. (2015) provided some examples of the types of initiatives undertaken by the Coalitions Linking Action and Science for Prevention to support healthy public policies in the built environment [57]. They suggested that the “Healthy Cities” approach was a useful framework to promote policy change in the built urban environment. Tortajada and Castelan (2003) concluded that there was an urgent need to improve the current water supply and wastewater management practices of the Mexico City Metropolitan Area, to become sustainable [58]. Collado and Wang (2020) scrutinized three recent slum upgrading programs in Latin America and discovered that slum upgrading was both a policy mechanism to address socio-economic issues and an instrument by which built-environment interventions could enhance adaptation and mitigation in informal settlements [59].

Substantial gaps remain in the literature and future researchers need to address the missing issues. Studies on the evaluation of RES construction are particularly scarce, and there have been even fewer investigations into the impact of RES policy on sustainable development. However, as a typical pilot promotion policy, promoting RES construction and rebuilding harmony between humans and nature is a key issue that policy makers and researchers need to focus on. Furthermore, exploring the policy effect of constructing a RES society is critical for promoting China’s sustainable development and achieving the UN’s objectives on sustainability. Second, in terms of methodology, the early literature primarily focused on qualitative analysis, with few quantitative investigations on the construction of an evaluation index system. More recently, the DID method was used more extensively in quantitative research on RES construction, but its assumption was to satisfy the equilibrium trend test. Furthermore, since it was limited by the availability of high-quality data, it was difficult to capture the policy effect over a longer time span. Therefore, based on the data from 2004 to 2019 and applying the PSM-DID method to evaluate the effects of RES construction as accurately as possible, we were able to make the experimental results more credible. Third, the internal mechanism was mostly concealed in the theoretical elaboration, which lacked a clear theoretical explanation and standardized empirical test. Therefore, in this paper we performed an in-depth analysis of the mechanism and heterogeneity influence of RES construction on sustainable development.

The potential contributions of this paper to the body of literature include: (1) Discussion of the impact of RES construction on sustainable development, which not only fills the gaps in research, but also summarizes China’s practical experience in promoting global sustainable development. RES construction is a typical Chinese pilot promotion policy, as well as an important element in China’s sustainable development plans; few reports have been published on the effectiveness of RES construction. (2) Sustainable development requires the coordinated growth of the economy, the environment, and society, and it emphasizes the quality of economic development under the constraints of the environment and resources. For this to happen, the assessment of the plans for sustainable development should not only reflect the economic benefits, but also capture the cost of the resources, and reduce the environmental pollution as far as possible. IGTFP affords a way to monitor the increase in growth efficiency while adhering to the environmental resource constraints and industrial development procedures. In this study, IGTFP as calculated by the SBM-GML index was used to evaluate sustainable development, reflecting the degree of coordination between ecological environment protection and industrial economic development. (3) We employed the PSM-DID method to explore the effects of RES construction on sustainable development. This method can help by correctly identifying the net effect of the RES policy, overcoming the endogenous and sample selection bias problems, and preventing other factors from interfering. In addition, we used the mediating effect model to discuss the specific mechanism of the effects of RES construction on sustainable development, which enriches the in-depth study of the micro-impact mechanism. We also employed the DDD method to analyze the heterogeneous influence of RES construction on sustainable development, by evaluating the policy effect of RES construction in multiple dimensions.

3. Theoretical Analysis and Hypotheses

3.1. Possible Impact Mechanism of RES Construction

By analyzing the RES construction process, we found that a series of policies implemented by local governments could effectively change the micro-internal efficiency and macro-allocation efficiency of the experimental areas. These policies mainly include government intervention, technological innovation, and human capital; also, opening up to the outside world, saving energy, and reducing emissions. The effectiveness of policy experiments will have a cyclic cumulative causal effect on IGTFP over time and ultimately affect regional sustainable development.

(1)

Effects of government intervention

Government performance appraisal is an important component of the RES construction system reform. Local governments in non-pilot cities only use GDP growth to evaluate government performance, and their proposals are always focused on rapid economic growth with little concern for the environment. In contrast, local governments in pilot cities use green GDP as the primary performance indicator and increase the weight of resource and ecology indicators in the evaluation system. This encourages the pilot cities to prioritize economic quality improvement over economic growth. However, local governments in underdeveloped cities may have to prioritize the development of heavy industry through central financial support. Heavy industry development has the potential to not only pollute the environment, but also to inhibit IGTFP growth. As a result, empirical testing is required to determine whether RES construction can affect sustainable development through government intervention.

(2)

Effects of technological innovation

Hubei and Hunan provinces have formulated a series of related measures and plans in their RES construction process. Accordingly, local governments are encouraged to: (1) guide researchers in developing key technologies for RES construction; (2) improve the incentive system for enterprise technological innovation and increase enterprise research investment; (3) encourage enterprises to absorb, digest, and reconfigure advanced technologies for current needs. Many studies found that technological innovation played an important role in environmental improvement and GTFP growth [60,61]. Therefore, technological innovation is considered one of the main engines for local governments to promote RES construction. It can stimulate various market participants to improve their R&D levels and promote further sustainable development.

(3)

Effects of human capital

The Wuhan city circle relies on innovation and entrepreneurship and attracts a large number of talented people to work in their high-tech development zone. Human capital is required for converting scientific research findings into procedures for improving production efficiency. By learning efficient energy-saving and emission-reduction technologies, advanced human capital not only improves the probability of successful technological innovation, but also stimulates scientific innovation to promote GTFP and high-quality economic development [62,63]. Therefore, it is reasonable to expect that RES construction will have an indirect impact on sustainable development through human capital.

(4)

Effects of opening up to the outside world

Many cities selected for RES construction have been identified as key areas for attracting foreign investment and developed industries. Some scholars believe that FDI can intensify competition by crowding out domestic market share [64]. As a result, the survival-of-the-fittest principle can compel domestic enterprises to implement technological reform and improve the resource allocation efficiency of IGTFP. However, according to the ‘pollution refuge’ hypothesis, polluting industries will relocate to countries or regions with looser environmental regulations, leading to increased pollution in those countries. Some researchers believe that FDI may inhibit IGTFP growth [65]. Thus, the impact of FDI on sustainable development in RES construction must be empirically tested.

(5)

Effects of energy saving and emissions reduction

By establishing a circular economy demonstration area, the Hubei provincial government was able to explore the cross-regional circular economy development model. This business model has the potential to increase resource utilization and waste recycling rates. To reduce pollution, the Hunan provincial government also explored a benign development model of resource recycling including enterprise, industrial chain, and regional circulation. In conclusion, RES construction can effectively reduce energy waste by integrating energy industries and establishing an energy price mechanism. Meanwhile, by adhering to the circular economy model and low-carbon processes, the RES construction can improve the environmental quality while achieving the goals of saving energy and reducing emissions. Therefore, the circular economy policy of the RES construction can promote regional sustainable development by saving energy and reducing pollution.

3.2. Impact of Heterogeneity on Effects of RES Policies

As illustrated in Figure 1, the Yangtze River economic belt is an economic circle along the Yangtze River, spanning the eastern, central, and western regions, covering eleven provinces and cities from Shanghai in the east to Yunnan in the west. The differences in economic and environmental development among cities are caused by the different geography and resource endowments of cities in the upper, middle, and lower reaches of the YREB. Therefore, we discuss the heterogeneous influence of RES construction on sustainable development from two aspects: innate resource endowment differences and acquired urban development differences.

In terms of resource endowment differences, 36% of cities in the YREB are identified as being resource-centered, and these cities are distributed across nine provincial regions in the YREB’s middle and lower reaches. As a result, the policy effects of RES construction may differ across cities due to resource endowment differences. Not only do the resource-based cities act as long-term sources of ‘natural nourishment’ for large industrial regions, they also provide considerable natural resources for sustainable development in the YREB. However, endless exploitation and reliance on resource-based cities will eventually lead to unsustainable development and resource exhaustion. The study showed that the key to rejuvenating resource-exhausted cities lies in transforming the economic development pattern from resource-dependent to innovation-driven [66]. However, because of the decline in the dominant industries, the resource-exhausted cities cannot develop a support system based on emerging alternative industries in the short term. In attempting to coordinate the development between the economy and the environment, the government may not be able to make ends meet. Then, it becomes more difficult for resource-exhausted cities in the YREB to achieve green transformation. With a foundation based on diversified industry and high-tech innovation, resource-based cities will be better able to maintain sustainable development with enhanced environmental regulation.

From the perspective of urban development differences, cities in the YREB are distributed in the eastern, central, and western regions of China, and the differences in resource endowments, population size, infrastructure, and other aspects result in cities having different degrees of development. In general, large cities have more solid economic foundations, more high-level talents to create green innovation, and can more easily attract capital for green environmental protection. Thus, large cities are able to achieve both economic development and environmental protection, which is conducive to sustainable development. However, due to a long-term dependence on natural resources, the smaller, underdeveloped cities in the upper reaches of the Yangtze River form an extensive economic development model. These areas may fall into the trap of the ‘resource curse’ [67], while small cities in the middle reaches of the Yangtze River may become pollution shelters for the relocation of high-polluting and high energy-consuming industries, which makes the environmental conditions of small cities worse. These industries are usually from large cities in the lower reaches of the Yangtze River.

Based on the above discussion, we propose the following hypotheses and test them in the fifth section.

Hypothesis 1 (H1). 

Compared with non-experimental cities, the RES pilot cities can indirectly affect sustainable development along the Yangtze River through government intervention, technological innovation, human capital, opening to the outside world, energy saving, and emissions reduction.

Hypothesis 2 (H2). 

The RES construction has a heterogeneous impact on the sustainable development of cities along the Yangtze River with different urban characteristics.

The mechanism of RES construction’s specific impact on sustainable development is summarized in Figure 2 below.

4. Research Design

4.1. Model Design

4.1.1. SBM-GML Index

(1)

Global directional SBM

According to the research of Fukuyama and Weber (2009) [18], we defined the global directional SBM, covering undesirable outputs, as follows:

$$\begin{gathered} S_v^G\left(x_0^t,y_0^t,b_0^t;g^x,g^y,g^b\right)=\underset{S^x,S^y,S^b}{\max }\frac{\frac{1}{M}{\sum }_{m=1}^M\frac{S_m^x}{g_m^x}+\frac{1}{N+1}\left({\sum }_{m=1}^M\frac{S_n^y}{g_n^y}+{\sum }_{i=1}^I\frac{S_i^x}{g_i^x}\right)}{2} \\ s.t. \sum _{t=1}^T\sum _{q=1}^0{\lambda }_q^t{x}_{qm}^t+S_m^x=x_{m0}^t, m=1,\cdots ,M \\ \sum _{t=1}^T\sum _{q=1}^0{\lambda }_q^t{y}_{qn}^t-S_n^y=y_{n0}^t, n=1,\cdots ,N \\ \sum _{t=1}^T\sum _{q=1}^0{\lambda }_q^t{b}_{qi}^t+S_i^b=b_{i0}^t, i=1,\cdots ,I \\ \sum _{q=1}^0{\lambda }_q^t=1, {\lambda }_q^t\ge 0,q=1,\cdots ,Q \\ {S}_m^x\ge 0,S_n^y\ge 0,S_i^b\ge 0 \end{gathered}$$

(1)

where $\left(x_0^t,y_0^t,b_0^t\right)$ represents input, desirable output, and undesirable output vector, respectively. $\left(g^x,g^y,g^b\right)$, $\left(S_m^x,S_n^y,S_i^b\right)$ represents the direction vector and slack vector, respectively. $\left(S_m^x,S_n^y,S_i^b\right)$ is the slack vector that indicates the redundant input, insufficient desirable output, and excessive undesirable output, respectively.

(2)

GML index

Oh (2010) proposed the GML index model by combining global production technology and the ML index [19]. Based on the global directional SBM model, we constructed the GML index as follows:

$$\begin{array}{c}GM{L}_t^{t+1}=\frac{1+{\overrightarrow{S}}_v^G\left(x^t, y^t,b^t;g^t\right)}{1+{\overrightarrow{S}}_v^G\left(x^{t+1}, y^{t+1},b^{t+1};g^{t+1}\right)}\end{array}$$

(2)

where ${\overrightarrow{S}}_v^G\left(x^t, y^t,b^t;g^t\right)$ is the global directional SBM based on the global production possibility set. $GM{L}_t^{t+1}$ indicates the changes in IGTFP of decision-making units in two adjacent periods during the study period. $GM{L}_t^{t+1}>1$ ($<1$) indicates the increase (decline) in the GML, and $GM{L}_t^{t+1}=1$ which indicates that the GML index is stable over the period from $t$ to $t+1$.

4.1.2. Difference-in-Differences (DID) Method

In this paper, twelve cities covered by the Wuhan city circle and Chang-Zhu-Tan city group were taken as the treatment group, and the remaining 93 cities in the YREB were taken as the control group. Because of the RES construction in December 2007, the year 2008 is considered the time node for policy implementation. We constructed the bidirectional fixed-effect DID model as follows:

$$IGTF{P}_{it}={\alpha }_0+{\alpha }_{1}HR{S}_{it}+{\alpha }_2\sum Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}$$

(3)

where $IGTF{P}_{it}$ stands for the $IGTFP$ of city $i$ in year $t$, $HR{S}_{it}$ represents the interactive term of the product of $D{W}_{it}$ and $D{T}_{it}$, and the coefficient ${\alpha }_1$ represents the impact of RES construction on sustainable development. $D{W}_{it}$ represents the dummy variable of the city, and $D{T}_{it}$ represents the dummy variable of time. ${\mu }_i$ represents the city fixed effect, and ${\gamma }_t$ represents the year fixed effect. $\sum Control{s}_{it}$ is a set of control variables, and ${\epsilon }_{it}$ is the random error term.

4.1.3. Propensity Score Matching (PSM) Method

In order to use the DID method, it must follow the premise that the treatment group and the control group share a common trend. Thus, if there is no RES policy, then there is no systematic difference in the overall factor change trend of cities in the experimental areas and other areas over time. However, in reality, there are enormous differences between the different regions, and this assumption is unsatisfactory. To control the unobservable, but not the time-varying differences between groups, we used the more accurate PSM-DID method. We first applied the PSM method to match all city samples and determine those with the highest degree of matching. Then, we repeated the DID estimation of the two groups of cities screened by PSM and obtained the net effects of the RES construction policy.

4.1.4. Mediating Effect Model

The RES construction not only has a direct impact on the sustainable development of cities along the YREB, but also has an indirect impact on sustainable development through government intervention (FAN), technological innovation (TI), human capital (HC), and opening up to the outside world (FDI). To verify these indirect effects, we constructed the models as follows:

$$T{I}_{it}=\beta 1_0+\beta 1_{1}HR{S}_{it}+\beta 1\sum Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}$$

(4)

$$FA{N}_{it}=\beta 2_0+\beta 2_{1}HR{S}_{it}+\beta 2\sum Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}$$

(5)

$$H{C}_{it}=\beta 3_0+\beta 3_{1}HR{S}_{it}+\beta 3\sum Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}$$

(6)

$$FD{I}_{it}=\beta 4_0+\beta 4_{1}HR{S}_{it}+\beta 4\sum Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}$$

(7)

The value of IGTFP indicates the influence of energy saving and emissions reduction. If energy consumption and pollutant emissions are reduced, IGTFP will rise. If RES construction can result in saving energy and reducing emissions in pilot cities, then IGTFP will be improved. Therefore, energy consumption (ES) and pollutant emission (PM) can be treated as potential factors that affect IGTFP and further used for the mediating effect test. Then, we constructed the models as follows:

$$E{S}_{it}=\beta 5_0+\beta 5_{1}HR{S}_{it}+\beta 5\sum Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}$$

(8)

$$P{M}_{it}=\beta 6_0+\beta 6_{1}HR{S}_{it}+\beta 6\sum Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}$$

(9)

4.1.5. Triple Difference (DDD) Method

The hypothesis related to the common trend of the DID model requires that any change in the treatment group be the same as that in the control group before the experiment. Only then, can the DDD model be introduced to control trend differences and to achieve an unbiased estimation of the treatment effect by constructing a new control group. Based on the DID estimation results, to further analyze the influence mechanism of resource endowment differences and urban development differences on the policy effect of the RES construction, the resource-exhausted cities ($R{E}_{it}$) dummy variable and city level ($C{L}_{it}$) dummy variable were introduced as triple difference items. The triple difference model was constructed as follows:

$$\begin{array}{c}IGTF{P}_{it}={\rho }_0+{\rho }_{1}D{W}_{it}\times D{T}_{it}\times R{E}_{it}+{\rho }_{2}D{W}_{it}\times D{T}_{it}+{\rho }_{3}D{W}_{it}\times R{E}_{it}\\ +{\rho }_{4}D{T}_{it}\times R{E}_{it}+{\rho }_5\sum Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}\end{array}$$

(10)

$$\begin{array}{c}IGTF{P}_{it}={\gamma }_0+{\gamma }_{1}D{W}_{it}\times D{T}_{it}\times C{L}_{it}+{\gamma }_{2}D{W}_{it}\times D{T}_{it}+{\gamma }_{3}D{W}_{it}\times C{L}_{it}\\ +{\gamma }_{4}D{T}_{it}\times C{L}_{it}+{\gamma }_5\sum Control{s}_{it}+{\mu }_i+{\gamma }_t+{\epsilon }_{it}\end{array}$$

(11)

The National Development and Reform Commission identified 69 resource-exhausted cities in 2008, 2009, and 2012, and we determined the resource endowment differences by distinguishing resource-exhausted cities from the others. Based on Equation (10), the value of the dummy variable ($R{E}_{it}$) was set to 1 if the city was resource-exhausted, otherwise to 0.

To explore the urban development differences, we divided the research samples into four types: first-tier, second-tier, third-tier, and lower-tier cities. In agreement with Song et al. (2022), we merged the new first-tier cities and first-tier cities into new first-tier cities and conducted the empirical analysis accordingly [68]. Based on Equation (11), the dummy variable ($C{L}_{it}$) was set to 1 if the city belonged to a certain type of city, otherwise to 0. The triple difference items, $D{W}_{it}\times D{T}_{it}\times R{E}_{it}$ and $D{W}_{it}\times D{T}_{it}\times C{L}_{it}$, were the core variables, and the coefficients ${\rho }_1$ and ${\gamma }_1$ indicated the heterogeneous impacts of different urban characteristics.

4.2. Variables and Measurements

4.2.1. Dependent Variable (Sustainable Development)

In order to better reflect the sustainable development of pilot cities along the Yangtze River, this study considered the dual constraints of energy consumption and environmental pollution, in addition to the traditional production factors of labor and capital. We defined the environmental pollutants, including industrial sulfur dioxide, industrial effluent, and discharge of general solid waste, as the unexpected outputs after the comprehensive assessment by using the principal component analysis method. In addition, the total industrial output of industrial enterprises as included served as the expected output, while labor, capital stock, and energy consumption were the input factors. Specifically, labor was denoted by the urban industrial employment population, capital stock was calculated by employing fixed capital in 2004 as the initial value with the perpetual inventory method, and energy referred to total industrial energy consumption.

4.2.2. Independent Variables

With regard to the independent variables, this paper adopted two dummy indexes including the city dummy variable ($DW$) and the time dummy variable ($DT$). In particular, the value of the city dummy variable ($DW$) was set to 1 if the city belonged to the experimental city group, otherwise the value was 0; the time dummy variable ($DT$) was assigned to 1 in 2008 and the subsequent years, otherwise to 0. Thus, the $HRS$ showed whether or not the city was influenced by RES construction in a certain year.

4.2.3. Control Variables

To accommodate the characteristics of each city, this paper employed several control variables gathered from prior studies: (1) gross domestic product (GDP), which used 2004 as the baseline year to diminish the impact of inflation; (2) industrial structure (IS), which was calculated by the proportion of the output value of secondary industry to local GDP; (3) technological innovation (TI), which was measured by the amount of internal R&D expenditure of urban industrial enterprises; (4) human capital (HC), which was measured by the number of students in urban colleges and universities; (5) opening up to the outside world (FDI), which was calculated by the proportion of foreign investment actually included in the local GDP; (6) population density (POP), which was measured as the ratio of urban resident population to the built-up area of a city; (7) urbanization level (UB), which was measured as the number of employed individuals in the urban population; (8) government intervention (FAN), which was calculated as the proportion of the general public budget allocated to local GDP.

The statistical description of the aforementioned variables is shown in

Table 1. Statistical description of the variables.

Variables	Observations	Mean	Std. Dev.	Min	Max
IGTFP	1680	1.01	0.22	0.24	4.24
GDP	1680	3236.08	4592.97	130.42	38,155.32
TI	1680	69.44	109.38	0.01	1238.60
IS	1680	0.46	0.21	0.03	4.76
HC	1680	160,589.64	223,365.13	4481.00	907,426.00
FDI	1680	50.91	86.91	0.22	1090.41
POP	1680	753.24	2953.87	117.84	71,023.00
UB	1680	115.02	157.18	4.17	1066.54
FAN	1680	0.18	0.17	0.06	2.52

.

4.3. Data Sources

After excluding those cities with administrative changes and missing information, the panel data from 105 cities at the prefecture-level or above in the YREB from 2004 to 2019 were included as the research dataset. The data sourced in this paper included the China Statistical Yearbook, China City Statistical Yearbook, China Industrial Statistical Yearbook, China Environmental Statistical Yearbook, China Energy Statistical Yearbook, statistical bulletins of prefecture-level cities, the Wind database, and the official website of the National Bureau of Statistics.

5. Empirical Results

5.1. Evolution Characteristics of IGTFP

The values of IGTFP and its decomposition items, green technical progress (GTP), and green technical efficiency (GTE), are shown in

Table 2. Values of IGTFP and its decomposition items in different YREB regions.

Year	Total Sample			East			Central			West
	IGTFP (1)	GTP (2)	GTE (3)	IGTFP (4)	GTP (5)	GTE (6)	IGTFP (7)	GTP (8)	GTE (9)	IGTFP (10)	GTP (11)	GTE (12)
2005	1.018	1.010	1.007	0.760	0.634	1.200	0.747	0.963	0.701	1.366	1.273	1.073
2006	1.033	0.983	1.054	0.972	1.077	0.902	0.765	0.940	0.814	0.752	0.831	0.905
2007	0.977	0.940	1.038	1.122	1.060	1.059	1.102	1.000	1.050	1.458	1.308	1.114
2008	0.980	0.967	1.014	1.005	1.023	0.982	0.802	0.780	1.028	0.923	0.908	1.017
2009	0.944	0.812	1.154	0.870	0.898	0.970	1.590	1.142	1.392	0.822	0.893	0.921
2010	1.024	1.000	1.006	0.957	0.936	1.022	0.837	0.747	1.120	1.193	1.182	1.010
2011	0.984	1.048	0.935	0.991	0.786	1.260	0.969	1.028	0.942	0.851	0.880	0.967
2012	0.996	1.056	0.940	0.474	1.806	0.263	1.058	1.062	0.997	1.159	1.179	0.983
2013	0.941	0.990	0.954	1.006	1.015	0.991	1.041	1.114	0.935	0.987	0.963	1.025
2014	0.998	0.982	1.015	0.990	1.020	0.971	0.935	0.901	1.039	0.805	0.797	1.009
2015	0.938	1.051	0.887	0.967	0.992	0.975	1.298	1.229	1.056	1.558	1.548	1.007
2016	1.002	1.034	0.979	0.943	0.921	1.024	1.009	1.010	0.999	0.674	0.959	0.702
2017	0.937	0.987	0.946	1.101	1.105	0.996	0.960	0.961	1.000	0.807	1.023	0.789
2018	1.014	1.101	1.003	1.037	1.012	1.025	1.135	1.133	1.002	0.915	1.000	0.916
2019	1.015	1.045	0.980	1.051	0.998	1.053	1.067	1.248	0.856	0.834	0.945	0.883
Mean	0.987	1.000	0.994	0.933	0.995	0.938	1.001	1.009	0.993	0.975	1.028	0.949

. The mean value of IGTFP in cities along the Yangtze River in column one was <1.000, and it decreased by 1.3 units from 2004 to 2019, indicating that industrial pollution not only destroys the environment but also affects IGTFP growth. It is worth noting, however, that IGTFP increased significantly from 2009 to 2010, reaching 1.024 in 2010 and implying that environmental protection measures implemented as part of RES construction improved environmental quality and promoted sustainable development in the YREB. Furthermore, IGTFP increased significantly after one or two years of RES policy implementation rather than changing immediately, suggesting that IGTFP improvement lagged behind policy changes.

After decomposing IGTFP into GTP and GTE, we found that GTE was >1.000 in many years while the average value in the sample period was 0.994, implying that there was still room for improvement in the overall green technical efficiency of cities along the Yangtze River. In comparison, the average value of GTP was 1.000, indicating that it was on the rise. In summary, technological progress played a leading role in promoting IGTFP in cities along the Yangtze River, while technical efficiency showed great potential for improvement. This conclusion was supported by many existing studies [69,70]. Under the rigid constraints of resources and environment, GTP became the engine driving sustainable development in the YREB. Thus, when investing in pollution control equipment, the government should strengthen equipment utilization and focus on stimulating the potential of GTE. This can inject a new impetus into the sustainable development of cities along the Yangtze River.

From the division of the YREB into eastern, central, and western regions, we can observe that the growth trend of IGTFP in each region is similar to that of the whole sample. When the mean values in columns five and six, eight and nine, and eleven and twelve are compared, the GTP of each region is greater than the GTE, demonstrating that GTP dominates IGTFP growth. The mean value of IGTFP in column seven is 1.001, indicating that IGTFP is increasing in the central region. The IGTFP in columns four and ten is <1.000, showing varying degrees of inhibition of IGTFP in the eastern and western regions, which is more significant in GTE.

5.2. Propensity Score Matching and Balance Test

To validate the matching method, we determined whether there was a significant difference in the means of each variable between the treatment group and the control group. The matching and balance test results are reported in

Table 3. Result of the matching and balance tests.

Variables	Logit Estimation		Type	Mean		Std. Dev. (%)	p Value
	Coefficient	p Value		Experimental	Control
TI	0.005 *** (0.001)	0.000	Unmatched Matched	114.392 115.081	62.485 115.13	44.2 −6.4	0.000 0.525
FDI	−0.033 *** (0.098)	0.053	Unmatched Matched	3.085 3.140	3.330 3.200	−16.7 −7.5	0.001 0.276
IS	1.257 ** (0.535)	0.019	Unmatched Matched	0.490 0.492	0.457 0.500	19.5 −2.7	0.145 0.478
HC	0.265 ** (0.139)	0.057	Unmatched Matched	10.895 10.881	11.159 10.951	−20.9 8.5	0.064 0.346
FAN	4.634 *** (0.729)	0.000	Unmatched Matched	0.304 0.263	0.165 0.258	49.9 1.6	0.000 0.895
POP	0.209 (0.224)	0.352	Unmatched Matched	6.292 6.291	6.209 6.332	15.8 −6.8	0.027 0.198
UB	0.341 *** (0.098)	0.056	Unmatched Matched	3.887 3.871	4.185 3.967	−30.0 −9.5	0.011 0.208
GDP	0.115 * (0.202)	0.045	Unmatched Matched	7.309 7.289	7.452 7.291	−12.9 0.9	0.044 0.245
Constant	0.082 *** (1.536)	0.009
R2	0.157

Notes: robust t statistics are shown in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

. After PSM, the standard deviation of each control variable decreased and all were <10%. Their P values were >0.1, verifying that no significant difference existed between the treatment group and the control group. The distributions of the chosen variables were balanced, and the comparability requirement was satisfied.

5.3. Baseline Regressive Results

5.3.1. Impact of RES Construction on Sustainable Development

Based on Equation (3), we conducted empirical analyses (

Table 4. Estimated results of baseline regression.

Variables	DID (1)	PSM-DID (2)	PSM-DID (Total) (3)	PSM-DID (Central) (4)
$HR{S}_{it}$	0.027 * (0.025)	0.029 * (0.038)	0.040 ** (0.047)	0.044 * (0.052)
TI	0.013 (0.000)		0.039 ** (0.001)	0.083 * (0.000)
IS	−0.012 (0.041)		−0.048 (0.060)	−0.031 (0.106)
FDI	0.108 ** (0.007)		0.161 ** (0.015)	0.203 ** (0.015)
HC	−0.021 (0.009)		0.051 * (0.023)	0.074 * (0.021)
POP	−0.088 (0.016)		0.091 (0.038)	0.270 * (0.038)
FAN	−0.014 (0.051)		−0.025 ** (0.073)	−0.051 * (0.075)
UB	0.096 (0.012)		0.015 (0.035)	0.027 (0.034)
GDP	0.116 (0.014)		0.117 (0.039)	0.158 (0.025)
Constant	1.278 *** (0.107)	1.032 *** (0.027)	1.297 *** (0.240)	1.707 *** (0.255)
City and Year FE	Yes	Yes	Yes	Yes
R2	0.018	0.024	0.031	0.053
N	1680	1360	1360	608

Notes: standard error in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

), and compared with using the DID method alone (col. one), the coefficients of the interactive term $HR{S}_{it}$ (col. three) were not only significantly positive at 5% significance level but also statistically increased after combining with the PSM method, implying that the endogenous problem was improved and the empirical results were optimized. When no control variables were added, the coefficients of $HR{S}_{it}$ (col. two) were positive and significant at the 10% significance level. After adding all control variables, the interaction coefficients (col. three) were positive and significant at the 5% significance level, implying that RES construction fostered IGTFP growth in the pilot cities. Compared to the non-pilot cities, the IGTFP in pilot cities was increased by 4.0%, confirming that RES construction in China not only achieved remarkable performance but also significantly promoted the sustainable development of YREB. In addition, when the sample area was narrowed down to the central region of YREB, the interaction coefficient in column four shows that RES construction promoted IGTFP by 4.4%, confirming the reliability of the benchmark regression results.

5.3.2. Impact of RES Construction on GTP and GTE

Since IGTFP can be decomposed into GTP and GTE, we further explored the impact of RES construction on cities along the Yangtze River. As can be seen in

Table 5. Results of industrial green technology progress and technical efficiency in the central region of the Yangtze River economic belt.

Variables	Total Sample		Central Region
	(1) GTE	(2) GTP	(3) GTE	(4) GTP
$HR{S}_{it}$	−0.008 * (0.005)	0.023 * (0.015)	−0.005 * (0.012)	0.010 * (0.012)
Control variables	Yes	Yes	Yes	Yes
Constant	1.261 *** (0.264)	1.052 *** (0.080)	0.817 *** (0.093)	1.900 *** (0.281)
City and Year FE	Yes	Yes	Yes	Yes
R2	0.033	0.044	0.083	0.064
N	1360	1360	608	608

Notes: standard error in parentheses; *** p < 0.01, * p < 0.1.

, the coefficients of $HR{S}_{it}$ (col. two) were positive and significant at the 10% significance level, while the coefficients in col. one were significantly negative at the 10% significance level, indicating that RES construction effectively increased GTP in pilot cities but inhibited GTE. The reason for the improvement in GTP may be that the implementation of environmental regulations and the introduction of energy-saving technologies in pilot cities promoted green technology innovations in industrial enterprises in the YREB. One possible cause of the GTE inhibition may have been because the local government used market-oriented means to shut down some ‘three-high’ enterprises with high energy consumption, high emissions, and high pollution levels. These policies would stimulate resource flow to enterprises and industries with green development and improve factor allocation efficiency. However, it is a lengthy and difficult process to change an irrational industrial structure, as is found in the heavy chemical industry in the YREB. This hinders the improvement of GTE even at economic scale efficiency in cities over the same period.

When the sample area was narrowed down to the central region of the YREB, the results in columns three and four were similar to the complete sample, indicating that RES construction increased GTP but inhibited GTE. When combined with columns eight and nine in Table 2, we concluded that GTP was an essential path for fostering IGTFP growth in the central region during the RES construction process.

5.4. Robustness Tests

5.4.1. Counterfactual Test

An important prerequisite of the DID method is that the treatment group and the control group must be comparable. This means that in the absence of RES construction, the IGTFP values in the treatment group relative to the control group do not change significantly with time. To verify the reliability of the baseline regressive results, a counterfactual test was also performed. The starting time of the RES construction was advanced to 2006 and postponed to 2009 and 2010. It can be seen (

Table 6. Results of the counterfactual test.

Variable	Treatment Year
	(1) 2009	(2) 2010	(3) 2006
$HR{S}_{it}$	−0.075 (0.046)	−0.106 (0.051)	0.062 (0.037)
Constant	1.191 *** (0.086)	1.237 *** (0.097)	1.152 *** (0.069)
Control variables	Yes	Yes	Yes
City and Year FE	Yes	Yes	Yes
R2	0.026	0.037	0.021
N	1680	1680	1680

Notes: standard error in parentheses; *** p < 0.01.

) that the $HR{S}_{it}$ coefficients in columns one, two, and three are not significant, revealing that there was no real difference in the effects of RES construction on IGTFP between the treatment and control groups at the 2008 baseline year. Thus, the choice of time node is reasonable.

5.4.2. Effects of Environmental Policy in Other Regions

Along with the RES policy, other environmental policies were enacted during the same period with a high likelihood of influencing the benchmark regression results. We collected and compiled the city-scale environmental policies enacted during the sampling period, which included the Two Control Zone policy, the low-carbon city pilot policy, and the emission-trading system. We added the relevant policy dummy variables and the cross-terms for the linear time-trend into the regression equation to control for the effects of related regional environmental policies on the results (

Table 7. Results of controlling for the influence of other policies.

Variables	Two Control Zones Policy (1)	Low-Carbon City Pilot Policy (2)	Emission-Trading System (3)
$HR{S}_{it}$	0.041 ** (0.023)	0.040 ** (0.028)	0.039 ** (0.022)
Constant	1.110 *** (0.067)	1.333 *** (0.188)	1.252 *** (0.154)
Control Variables	Yes	Yes	Yes
City and Year FE	Yes	Yes	Yes
R2	0.060	0.074	0.068
N	1680	1680	1680

Notes: standard error in parentheses; *** p < 0.01, ** p < 0.05.

). It can be seen that the coefficients are similar to the results of the benchmark regression in Table 4—the coefficients of $HR{S}_{it}$ are all positive and significant at the 5% significance level. Thus, the results in this paper remained reliable in spite of the influence of other regional environmental policies.

5.4.3. Changing the Sample Time Window

The randomness principle of the DID model is not only reflected in policy time nodes, but also in the sample time windows. To eliminate the estimation errors caused by possible exogenous variables during the sample period, we tested different sample time windows: 2004–2012, 2004–2014, and 2004–2016. As shown in

Table 8. Results of changing the sample time window.

Variables	2004 to 2012	2004 to 2014	2004 to 2016
$HR{S}_{it}$	0.132 * (0.030)	0.090 ** (0.071)	0.086 ** (0.050)
Constant	1.128 *** (0.052)	1.291 *** (0.123)	1.231 *** (0.093)
Control Variables	Yes	Yes	Yes
City and Year FE	Yes	Yes	Yes
R2	0.051	0.043	0.036
N	945	1155	1365

Notes: standard error term in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

, the coefficients of $HR{S}_{it}$ were significantly positive at the 1% and 5% significance levels, indicating that RES construction significantly promoted sustainable development in cities along the Yangtze River. The results after adjusting the sample time window were consistent with those of the benchmark regression.

5.4.4. Synthetic Control Method

To verify the robustness of the experiment, we referred to Abadie et al. (2015) and randomly selected Wuhan and Changsha for the experiment by using the synthetic control method (Figure 3) [71]. Before the implementation of the RES policy (2004–2008), the trends of Wuhan and ‘synthetic Wuhan’, and Changsha and ‘synthetic Changsha’ were similar, and the values all fluctuated around zero. After the implementation of the RES policy, the differences between Wuhan and ‘synthetic Wuhan’ were negative in 2010, 2012, and 2013, and positive in the other years, and the average fluctuation was 0.038. Similarly, the difference between Changsha and ‘synthetic Changsha’ was negative in 2013, but it was positive in the other years and the average fluctuation was 0.043. It can be seen that the Ln (IGTFP) in Wuhan was higher than that of ‘synthetic Wuhan’ after the RES construction, indicating that the latter did indeed promote sustainable development in Wuhan. The results for Changsha were similar, confirming the robustness of the benchmark regression experiment.

6. Further Analysis

6.1. Mediating Effect Test

After the application of innovative technology, the coefficient of $HR{S}_{it}$ (col. one) and the coefficient of $TI$ (col. seven) were both positive and significant at the 1% significance level. This means that RES construction had a significantly positive effect on IGTFP through technological innovation (

Table 9. Mediating effect test.

Variables	TI (1)	FAN (2)	FDI (3)	HC (4)	ES (5)	PM (6)	IGTFP (7)
$HR{S}_{it}$	0.175 *** (10.016)	0.267 ** (0.019)	0.020 * (0.128)	0.061 ** (0.104)	−0.110 *** (0.072)	−0.098 *** (0.011)	0.031 * (0.026)
TI							0.013 *** (4.945)
FAN							−0.002 ** (0.028)
FDI							0.082 (0.192)
HC							0.004 * (0.156)
ES							0.010 *** (0.107)
PM							0.101 *** (0.017)
Constant	4.619 *** (4.664)	0.126 *** (0.077)	2.556 *** (0.521)	3.168 *** (0.424)	0.363 *** (0.291)	0.967 *** (0.045)	1.193 *** (0.092)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
City and Year FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes
R2	0.320	0.086	0.189	0.475	0.590	0.544	0.683
N	1680	1680	1680	1680	1680	1680	1680

Notes: standard error in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1.

). Innovative RES programs on the application of advanced technology can promote environmental protection as well as confer competitive advantages to green enterprises able to utilize high-tech processes. It can also have a demonstratively competitive effect on polluting businesses, encouraging local enterprises to invest more in scientific research and technical innovation. Independent innovation and industry–university research collaboration are both beneficial to sustainable development in the YREB.

In the government intervention category, the coefficient of $HR{S}_{it}$ (col. two) was significantly positive at the 5% significance level, while the coefficient of $FAN$ (col. seven) was significantly negative, implying that government intervention was not a mediating factor (Table 9). Based on their factor endowments, local governments in RES pilot areas implemented a variety of environmental and industrial policies that were part of a model of sustainable development with local characteristics. However, government intervention may not have much impact on the proportion of heavy industry along the Yangtze River. It is difficult to quickly replace an irrational industrial structure and improve economic scale efficiency. Excessive government intervention, however, can result in overinvestment and overcapacity, both of which can impede sustainable development.

For opening up to the outside world, it can be seen that the coefficient of $HR{S}_{it}$ (col. three) was significantly positive at the 10% significance level, and the coefficient of $FDI$ (col. seven) was non-significant, showing that opening up was not a mediating factor. By introducing advanced technology and providing experience, the RES construction could stimulate technology spillover and demonstration effects by opening up to the world. This accessibility has the potential to increase productivity and environmental quality of host-country enterprises. However, while it contributes to the industrial restructuring and green transformation of the YREB by increasing the scale of foreign investment, the positive effect of FDI on sustainable development cannot be fully realized due to the low quality of foreign investment.

For human capital, the coefficient of $HR{S}_{it}$ (col. four) and the coefficient of $HC$ (col. seven) were significantly positive at the 5% and 10% significance levels, respectively, implying that RES construction improved IGTFP by enhancing human capital. Cutting-edge clean technology research, the application of advanced environmental facilities, and the development of high-tech, high-value green products all require advanced human capital. RES construction significantly increases the demand for advanced human capital by promoting enterprise technology innovation. It has the potential to encourage businesses to increase their human capital investment and attract a higher-quality labor force able to participate in advanced technological innovation activities. Through knowledge accumulation and spillover, human capital can improve resource utilization, which can reduce the reliance on traditional energy, resulting in energy savings and emissions reductions, as well as promoting sustainable development.

For saving energy and reducing emissions, the coefficients of $HR{S}_{it}$ (col. five; six) were both significantly negative at the 1% significance level, whereas the coefficients of $ES$ and $PM$ in column seven were both significantly positive at the 1% significance level. This means that RES construction can improve sustainable development by saving energy and reducing emissions.

6.2. Heterogeneous Analysis

6.2.1. Degree of Heterogeneity in Urban Resource Dependency

Based on Equations (10) and (11), we conducted an empirical analysis, and the results are shown in

Table 10. Results of heterogeneous analysis.

Variables	Resource-Exhausted Cities (1)	First-Tier Cities (2)	Second-Tier Cities (3)	Third-Tier Cities (4)	Below Third-Tier Cities (5)
$D{W}_{it}\times D{T}_{it}\times R{E}_{it}$	−0.029 ** (0.079)
$D{W}_{it}\times D{T}_{it}\times C{L}_1$		0.053 ** (0.034)
$D{W}_{it}\times D{T}_{it}\times C{L}_2$			0.044 ** (0.024)
$D{W}_{it}\times D{T}_{it}\times C{L}_3$				−0.036 (0.020)
$D{W}_{it}\times D{T}_{it}\times C{L}_4$					−0.018 (0.023)
RD	0.065 (0.001)	0.041 (0.001)	0.061 (0.001)	0.035 (0.001)	0.041 (0.001)
FAN	−0.190 (0.785)	−0.031 (0.052)	−0.002 (0.049)	−0.008 (0.049)	−0.003 (0.049)
FDI	0.260 (0.023)	0.092 ** (0.008)	0.107 ** (0.007)	0.115 *** (0.008)	0.102 ** (0.008)
HC	−0.075 (0.026)	0.030 (0.009)	0.023 (0.009)	0.027 (0.009)	0.026 (0.009)
IS	−0.415 (0.159)	−0.013 (0.052)	−0.011 (0.041)	−0.013 (0.041)	−0.008 (0.041)
EB	0.581 (0.046)	0.090 * (0.011)	0.080 (0.012)	0.087 (0.012)	0.088 (0.012)
POP	0.319 (0.029)	0.086 (0.016)	0.097 (0.016)	0.090 * (0.016)	0.086 (0.016)
GDP	0.248 (0.040)	0.125 * (0.014)	0.138 * (0.014)	0.120 (0.014)	0.122 (0.014)
Constant	0.641 *** (0.257)	1.319 *** (0.109)	1.325 *** (0.110)	1.309 *** (0.110)	1.302 *** (0.118)
City and Year FE	Yes	Yes	Yes	Yes	Yes
R2	0.267	0.021	0.021	0.018	0.017
N	1680	1680	1680	1680	1680

Notes: (1) standard error in parentheses; *** p < 0.01, ** p < 0.05, * p < 0.1; (2) in resource-exhausted cities, only administrative cities above prefecture-level are considered; (3) the classification of city hierarchies refers to the 2019 urban classification of the First Financial New First-tier City Research Institute.

. The coefficient of $D{W}_{it}\times D{T}_{it}\times R{E}_{it}$ (col. one) was negative and significant at the 5% significance level and showed that RES construction reduced IGTFP in resource-exhausted cities along the Yangtze River by 2.9 units and hindered their sustainable development. The most likely explanations for this observation include the following. First, industrial structure has a negative impact on IGTFP growth in resource-exhausted cities. The RES policy, with its characteristic weak incentives and weak constraints, had no effect on the YREB’s industrial layout, which was dominated by heavy industry. The imbalance in industrial structure resulted in an over-reliance on resource industries, which can stymie the development of tertiary or alternative industries, further limiting sustainable development in the YREB. Second, government intervention has a harmful influence on the growth of IGTFP. Some ‘zombie’ companies continue to receive long-term government support to alleviate employment pressures. It is not only difficult to realize market clearing, but also hinders sustainable development. Third, human capital has a negative effect on IGTFP. Foreign capital and private enterprises can be squeezed out of resource-based industries, resulting in a single employment structure and brain drain. Because of the high cost, substantial risk, and long cycle of technological innovation, it is difficult for enterprises to adopt green technology, further impeding sustainable development.

6.2.2. Degree of Heterogeneity of Urban Development

The coefficients of the core explanatory variables were both significantly positive at the 5% significance level (Table 10, col. two and three), indicating that RES construction in first-tier and second-tier cities significantly promoted sustainable development. However, the coefficients in columns four and five were negative and non-significant. This meant that the effect of third-tier cities and cities below the third tier on sustainable development was not significant. Governments can achieve both economic development and environmental protection because first-tier and second-tier cities have strong economic foundations. To meet the new production requirements and environmental protection standards in RES construction, enterprises and industries in first-tier cities and second-tier cities with high pollution, energy usage, and emissions may be transferred to third-tier cities or cities below the third tier, which can hinder sustainable development in those cities.

7. Discussion and Conclusions

7.1. Discussion

China is the largest developing country dedicated to actively promoting sustainable development. The implementation of pilot promotion policies is an important factor for realizing sustainable development from concept to practice, as well as a key support to allow China to achieve the United Nations’ SDGs as soon as possible. RES construction in China is an important part of the process of sustainable development. The construction of a society in which people live in harmony with nature is inextricably linked to the concept of sustainable development. This paper evaluated the impact of RES construction on sustainable development, which is of great significance for summarizing practical experience with Chinese characteristics and promoting sustainable development.

The majority of the literature on the evaluation of sustainable development in China is based on the construction of an evaluation index system, and few researchers measured sustainable development with GTFP. The essence of promoting development sustainably is to stimulate economic development while making energy use more efficient and reducing pollution, which results in an improvement in GTFP. As a result, using GTFP to evaluate sustainable development has reference value and significance. Furthermore, with China proposing a ‘double-carbon’ goal as well as a ‘made in China 2025’ initiative, the decoupling problem between industrial economic development and environmental pollution must be addressed as soon as possible. As a result, we selected IGTFP as a proxy for assessing sustainable development.

The goal of RES construction is to promote sustainable development but also to foster coexistence between humans and nature. This paper treats the RES construction in China as a quasi-natural experiment and evaluates the policy effect of the RES construction on sustainable development. It not only contributes to strengthening China’s ecological civilization construction in the new period and promoting the new pattern of harmonious development between humans and nature, but also provides meaningful empirical evidence for further deepening the implementation of sustainable development strategy and promoting the realization of global sustainable development goals.

7.2. Conclusions

Based on a panel of data from 105 cities at prefecture-level and above along the Yangtze River from 2004 to 2019, this paper treated RES construction in China as a quasi-natural experiment and adopted the PSM-DID method to evaluate the effect of RES construction on sustainable development in the YREB. We analyzed the influence mechanism and heterogeneity characteristics of the policy and came up with the following conclusions. (1) The IGTFP calculation results showed that GTP was most important in improving IGTFP in pilot cities, while GTE still had room for improvement. In terms of regional heterogeneity, RES construction primarily promoted IGTFP growth in the YREB’s central region. It had varying degrees of inhibitory effects on IGTFP in the eastern and western regions, which was more pronounced for GTE. (2) RES construction increased the IGTFP of pilot cities by 4.0 percent, which confirmed that RES construction did indeed promote sustainable development. When we considered the decomposition terms of IGTFP, they showed that RES construction promoted GTP but inhibited GTE in the pilot area. (3) The mediating effect model revealed that RES construction had a significantly positive effect on sustainable development through technological innovation, human capital, energy conservation, and emission reduction. However, government intervention offset some of the positive stimulus effects, and the effect of opening up to the outside world on sustainable development was not significant. (4) Heterogeneity analysis revealed that, compared to resource-based cities, the effects of sustainable development on resource-exhausted cities were limited. While RES construction could promote sustainable development in the first-tier and second-tier cities, it hindered sustainable development in third-tier cities and cities below the third tier.

Based on the above conclusions, we propose the following recommendations: (1) The conclusions show that technological innovation and human capital are positive and effective intermediary factors for promoting sustainability. Therefore, local governments should reconfigure the Yangtze River economic belt into a green innovative industrial hub by cultivating the application of new technological innovations and further developing human capital. Urban agglomerations in the middle reaches of the Yangtze River, represented by the Wuhan city circle and Chang-Zhu-Tan city group, should learn from the experiences of industry in the Yangtze River delta and combine their resource endowments to deploy advanced technology and environmental protections. Governments should invest more in developing human capital and strengthening talent collaborations between provincial capitals and coastal areas with abundant resources for innovation. The efficiency loss caused by resource mismatch and the poor use of human capital could be reduced by enhancing the knowledge spillover effect through more inter-regional collaboration and joint innovation projects. (2) The conclusions show that energy saving and emission reduction have a positive effect on promoting sustainable development. Therefore, local governments along the Yangtze River should optimize the energy consumption structure, increase their usage of clean energy, and reduce the proportion of coal consumption in industrial production, as well as promote the use of new energy-saving processes and technology to improve energy efficiency. The high industrial energy consumption of resource-exhausted cities is a key point in the YREB’s plans to reduce energy usage and emissions. As a result, local governments can establish more pilot projects in clean energy and green processes to support the green transformation and sustainable development of resource-exhausted cities. (3) The conclusions state that opening up investment to the outside world would not necessarily produce a significant improvement in sustainable development; however, this may be due to the type and quality of foreign capital available for investment. Therefore, when cities along the Yangtze River undertake industrial transfer projects from other regions or from abroad, governments should improve the foreign capital quality in accordance with the RES requirements. They must not only actively attract domestic and foreign capital with resource-saving and environmentally friendly characteristics, but must also prevent highly polluting industries from transferring operations from first- and second-tier cities to lower-tier cities. (4) The conclusions indicate that local government intervention could actually have an adverse effect by offsetting some positive effects in other areas. Therefore, policy makers should insist that the market be allowed to take a more prominent role in resource allocation, so as to reduce the direct intervention of the government in micro-economic activities. In addition, the government should tailor environmental regulations to cultivate and develop the application of new resource-saving processes and offer more market incentives to environmentally friendly industries.