Carbon constraints, industrial structure upgrading, and green total factor productivity: an empirical study based on the Yangtze River Economic Belt

As industrialization and urbanization continue to increase, carbon constraints have gained a prominent place in the discourse on global climate change. Carbon constraints are essentially environmental rules. Because of the close relationship between industrial energy saving and carbon emission reduction, the concept of carbon constraints includes energy saving and emission reduction. Industries are the lifeblood of economic development, but they are also the largest producers of carbon emissions. The traditional production mode characterized by high inputs, high energy consumption, and high emissions cannot support sustainable and high-quality economic development. This production mode has also set back the low-carbon green development of China's economy. In September 2020, China's central government pledged to ensure that CO₂ emissions peak by 2030 and achieve carbon neutrality by 2060, at the 75th United Nations General Assembly, which means that China's economy will have to carry out rapid and comprehensive decarbonization reforms in the next 40 years. The Yangtze River Economic Belt is one of the densest and critical economic regions and one of China's most important industrial corridors. Many of the modern industries, such as steel, automobiles, electronics, and petrochemicals, are concentrated here, which are high-energy-consuming, high-emissions, and large-scale enterprises. Industrial activities are closely related to water, and water is an essential resource in industrial development. However, the utilization of water resources in industries in the past was rough, and water resources were greatly wasted due to unreasonable structure and low-efficiency recycling, which hindered the green development of industry. Therefore, it is of great significance for China to study the green development of industry in the Yangtze River Economic Belt and put forward corresponding suggestions.

In this study, green total factor productivity (GTFP) is used to measure the level of industrial green development. Taking environmental factors into account, green total factor productivity is the ratio of total input to total output in a country or a region. It considers input factors including capital, labor, and energy consumption, as well as output factors including expected output and unexpected output, which is an important indicator to measure the coordinated development of resources, environment, and economy in a country or region.

How we effectively control industrial carbon emissions and improve green total factor productivity is an important concern in promoting the green development of industry. Wang & Xiang (2014) proposed that industrial structure upgrading could reduce carbon emissions. Industrial carbon constraints play a vital role in the upgrading of industrial structure. Although many scholars have studied the relationship between carbon constraints, industrial structure upgrading, and green total factor productivity using different methods, few have combined the three concepts into the same research framework. In order to broaden the research in this field, this study explored how carbon constraints and industrial structure upgrading affect green total factor productivity in the Yangtze River Economic Belt.

In order to promote the green development of industry, many scholars have discussed carbon emissions from different perspectives, mainly focusing on how carbon constraints affect industrial development. Xu et al. (2014) adopted the Logarithmic Mean Divisia Index model to analyze the influence factors of the emission intensity of industrial energy consumption. Ouyang & Lin (2015) studied the determinants of industrial emissions in China by using the co-integration method. Dong et al. (2017) analyzed carbon emissions and implied carbon transfer in the copper industry. Dong et al. (2016) and Guo et al. (2021) studied the relationship between carbon emissions and economic development from a provincial perspective. Yu et al. (2015) and Liu et al. (2022) studied the impact of carbon emissions on industrial low-carbon transformation at the urban level. Wang & Jiang (2019), Wang & Zhang (2020), and Apeaning (2021) used the decoupling principle to explore the relationship between industrial carbon emissions and economic growth. Wu & Zhang (2021) found that the Yangtze River Economic Belt has strategic importance in China's regional development and discussed the internal relationship between carbon emissions and regional economic development. Sanguinet et al. (2022) applied the cross-regional input–output model to determine the interaction between CO₂ emissions embedded in the commodities of resource-based industries and other industries. Based on the above analysis, this study proposes the following Hypothesis 1:

Hypothesis 1 Carbon constraints have a positive effect on green total factor productivity.

Industrial structure upgrading can directly promote industrial green development and be an essential driver for changing the mode of regional industrial development. New technologies emerge with the upgrading of industrial structure, for example, the integration of industrial technology and intelligent technology and Industry 4.0. Industry 4.0 uses information technology to promote industrial transformation – the era of intelligence, which will promote the green development of industry (Ghobakhloo 2018, 2020; Wichmann et al. 2019). Industrial structure upgrading mainly affects industrial green development through the replacement of production factors and change in industrial structure, more efficiency in industrial factor replacement, exploring the characteristics of industrial structure adjustment, and discussing the upgrading path to be taken (Zhu et al. 2019; Zeng et al. 2020). Industrial structure upgrading also affects industrial green development by coordinating the operations of various industrial sectors, promoting a more effective resource allocation in various industrial sectors, improving green total factor productivity, and achieving high-quality development of the industrial economy (Yang & Wang 2020). Based on the above analysis, this study proposes the following Hypothesis 2:

Hypothesis 2 The upgrading of industrial structure is conducive to improving green total factor productivity.

Carbon constraints have profoundly changed the development mode of China's economy and are significant in optimizing and upgrading China's industry. Carbon constraints are important for strategic industrial entrepreneurship and forming the basis of economic development (Boudreaux 2020). With carbon constraints, industrial structure can be optimized and upgraded through technological innovations and improvements in energy efficiency (Wu et al. 2019). Technological innovation not only promotes the adjustment of industrial structure, but also improves sharing of resources and benefits, promotes the flow of innovation resources, and improves energy and environmental efficiency, thus effectively promoting further upgrading of the industrial structure (Li & Zeng 2020). However, as industrial development is tilted toward the traditional production mode, economic and industrial development structure show three highs – high pollution, high emissions, and high energy consumption. Under the new carbon emissions constraint policy, upgrading industrial structure and achieving sustainable economic development are important (Dong et al. 2020). Transforming industrial low-carbon energy is an effective way to achieve energy conservation and emission reduction (He et al. 2022). Han & Zhao (2022) explained that the upgrading of industrial intelligence can narrow the gap in the regional economy. Based on the above analyses, this study proposes the following hypotheses:

Hypothesis 3 Carbon constraints are conducive to promoting the upgrading of industrial structure.

Hypothesis 4 The upgrading of industrial structure has played an intermediary role in the process of carbon constraints promoting industrial green development.

The GTFP of 11 provinces in the Yangtze River Economic Belt is selected as the explained variable. According to the definition of GTFP, the input–output system refers to Yan et al. (2020) and Huang & Wu (2021), and the input index variables are industrial capital inputs, labor inputs, and energy consumption. The output is divided into expected and unexpected output. The expected output index variable is industrial added value and the unexpected output index variables are three kinds of major industrial pollutants, which are industrial wastewater discharge, industrial waste gas discharge, and industrial solid waste production. The super-efficient SBM model is used to calculate the green total factor productivity of 11 provinces in the Yangtze River Economic Belt, and the descriptive statistics of the input–output indicators are shown in Table 1. The data mainly come from the China Statistical Yearbooks from 2006 to 2020.

There are other factors that affect green total factor productivity in reality, so control variables are added to this study: the level of economic development (GDP) measured by the regional gross product; the level of science and technological achievement (STL) expressed by the investment intensity of R&D funds; human resources (HR) expressed as the number of colleges and universities; government intervention ability (GOV), which measures government intervention as the proportion of fiscal expenditure in each province to GDP; industrial scale (IES) expressed as the number of industrial enterprises; and foreign investment (FI) expressed as total foreign investment. To accurately measure the impact of carbon constraints and industrial structure upgrading on green total factor productivity, based on the existing studies in the literature, different control variables are added to the benchmark regression model and the intermediary effect model. The data come from the Wind database and the Yangtze River Economic Belt Big Data Platform.

In order to eliminate the influence of dimensions between variables, the logarithm of the above individual variables is first taken. In Table 2, the descriptive statistics of each variable are given.

Equation (2) assumes that there are n decision-making units, that is, 11 provinces in the Yangtze River Economic Belt, each of which contains m input indicators, h₁ expected output indicators, and h₂ unexpected output indicators; x, y^d, and y^u, respectively, represent the corresponding elements in the input matrix, expected output matrix, and unexpected output matrix, and ρ represents the value of GTFP.

For the intermediary effect test, the stepwise regression method is adopted following Wen & Ye (2014). First, we check whether the total effect coefficient β₁ is significant, and if so, whether there is a tentative intermediary effect. If it is not significant, the test is complete, and we can conclude that the intermediary variable has a masking effect in the entire process.

When conducting an empirical test, the panel model must be chosen according to the problem being solved. If the measurement model is not appropriate, there will be large deviations in the results and the results will not be authentic. There are three models for analyzing panel data: the fixed effect model, the random effect model, and the mixed effect model. The model is chosen by the Hausman test. In this study, the explanatory variables of carbon constraints, industrial structure upgrading, and green total factor productivity are tested with the Hausman test. As shown in Table 3, the results show that the p-value is 0.7924, which is greater than 0.01. The original hypothesis cannot be rejected at the 1% significance level, so this study uses the random effect model.

The green total factor productivity in the Yangtze River Economic Belt shows a fluctuating upward trend. From 2006 to 2013, the overall increase in green total factor productivity was not obvious. From 2013 to 2020, the green total factor productivity of the 11 provinces increased significantly, showing obvious spatial characteristics. In 2020, the spatial pattern of green total factor productivity of the 11 provinces in the Yangtze River Economic Belt was characterized by ‘a gradient decreasing spatial distribution from the Yangtze River Delta to Urban Agglomeration in the Middle Reaches to the Chengdu–Chongqing Economic Zone.’ That is to say, green total factor productivity in the Urban Agglomeration in the Middle Reaches and the Yangtze River Delta is generally higher than that of the Chengdu–Chongqing Economic Zone. There are two main reasons: on the one hand, the economic development strength of the Chengdu–Chongqing Economic Zone is not as good as that of the Yangtze River Delta, and the capital, talents, and other factors that promote industrial green development technology are lacking. On the other hand, the industrial fields of the provinces in the Urban Agglomeration in the Middle Reaches insist on promoting industrial energy conservation and emission reduction as a breakthrough and an important starting point for structural transformation. However, the overall level of industrial green development in the Chengdu–Chongqing Economic Zone has improved significantly in recent years. From the above results, it can be found that regions should play a linkage role, strengthen regional cooperation, promote the formation of a new pattern of regional development, and improve the industrial green total factor productivity of each region. The results are also applicable to areas similar to the development of the Yangtze River Economic Belt.

First, 11 provinces in the Yangtze River Economic Belt are used in the study sample to empirically analyze the internal relationship and impact of carbon constraints, industrial structure upgrading, and green total factor productivity. Model 1 is the effect of carbon constraints on green total factor productivity, Model 2 is the effect of industrial structure upgrading on green total factor productivity, and Model 3 is the effect of synergy between carbon constraint and industrial structure upgrading on green total factor productivity. This study discussed the relationship among carbon constraints, industrial structure upgrading, and green total factor productivity from multiple models, but the previous findings did not consider the synergistic effect of carbon constraints and industrial structure upgrading. Therefore, this study could broaden the research in related fields. Combining the results of Model 1, Model 2, and Model 3, the study can test the overall synergistic effect of carbon constraints and industrial structure upgrading on green total factor productivity. Also, the regression results are shown in Table 4.

Table 4 shows that the regression results of Model 1–Model 3 are all significant. The coefficient of carbon constraints intensity index is negative, indicating that the higher the intensity of carbon constraints, the higher the green total factor productivity. The coefficient of industrial structure upgrading is positive, indicating that industrial structure upgrading has a positive role in promoting green total factor productivity. However, careful perusal shows that the results of Model 1 and Model 2 are different from those of Model 3. In Model 3, when both carbon constraints and industrial structure upgrading are included in the model, the absolute value of the coefficient of carbon constraints intensity index and the coefficient of industrial structure upgrading variable all increase, which shows that the synergistic effect of carbon constraints and industrial structure upgrading is stronger for improving green total factor productivity.

Specifically, carbon constraints have a positive effect on green total factor productivity. Both Model 1 and Model 3 are significant at 5% and 1% levels, respectively, which verifies Hypothesis 1. Currently, under China's double carbon strategy, strengthening carbon constraints has become an important goal to improve the quality of industrial development and move away from the traditional industrial growth model of high pollution, high emissions, and high energy consumption in the Yangtze River Economic Belt. According to the above analyses, the development of green industry and the transformation of green and low carbon can be accelerated by reducing CO₂ emissions, reducing the intensity of energy consumption, and increasing the use of clean energy, which can lead to an improvement in green total factor productivity. In Model 1, the control variables, economic development level (GDP), and scientific and technological level (STL), have a positive role in promoting green total factor productivity; the higher the level of economic development and scientific and technological achievement is, the higher the level of industrial green development is. The level of economic development improves the level of industrial green development through the renewal of industrial equipment and comprehensive resource utilization. The improvement in scientific and technological levels is reflected in more research and development, and the use of clean energy. Actively carrying out green low-carbon technology research and development aids industrial green development, reduces the consumption of coal, oil, natural gas, and other traditional energy sources, promotes industrial green transformation, and improves green total factor productivity. The scale of industrial enterprises (IES) and foreign investment (FI) has a negative effect on green total factor productivity; when the industrial scale expands, carbon emissions and energy consumption increase, which is not conducive to the green development of industry and has a negative effect on green total factor productivity. Foreign investment in China is mainly concentrated in the industrial sector, so foreign investment increases the output value of the industry, but in the process, energy consumption and pollution also increase, inhibiting the green development of industry. The influence of government intervention (GOV) on green total factor productivity is not significant, which may be due to the lag in government intervention.

Industrial structure upgrading has a positive effect on green total factor productivity. Both Model 2 and Model 3 pass the test at the 5% significance level, which verifies Hypothesis 2. The upgrading of industrial structure leads to the development of tertiary industry, cleaner and pollution-free industries, and reduced industrial waste, which leads to an improvement in green total factor productivity. Among the control variables, the level of economic development (GDP) has a positive effect on green total factor productivity, as economic development leads to optimization and upgrading of industrial structure, promoting the green development of industry. Government intervention (GOV) and foreign investment (FI) have a negative effect on green total factor productivity; an increase in government intervention leads to de-industrialization (Guo & Yan 2020), which inhibits the promotion of green total factor productivity. Foreign investment increases investment in industry, which is not conducive to the upgrading of industrial structure, so it is not conducive to industrial green development either. In Model 2, the level of STL and the scale of industrial enterprises (IES) have no significant influence on green total factor productivity.

Second, when carbon constraints and industrial structure upgrading are included in Model 3, they play a synergistic role. The regression coefficient of carbon constraints changes from −0.2517 to −0.3155, and the regression coefficient of industrial structure upgrading changes from 0.0653 to 0.0714, which shows that their influence on green total factor productivity increases significantly. Therefore, we should give more attention to their synergistic role, as their simultaneous effects are more conducive to industrial green development. In this model, the level of economic development (GDP), government intervention (GOV), and foreign investment (FI) have a significant effect on green total factor productivity. The level of economic development promotes improvement in green total factor productivity, while government intervention and foreign investment inhibit green total factor productivity. The level of STL and the scale of industrial enterprises (IES) have no significant effect on green total factor productivity. There is an inevitable relationship between the level of science and technology and green development, but in the process of science and technology development, a lot of manpower and funds need to be invested to provide a power source for scientific and technological innovation to promote the green development of industry. So the process takes a long time. Although scientific and technological innovation can promote the green transformation of industry through industrial production efficiency, energy structure optimization, and resource recycling, the promotion effect has not been significantly demonstrated. The expansion of the scale of industrial enterprises is conducive to the increase of industrial output value and plays an important role in the development of industrial economy, but there is a lag effect in this process. So its impact on industrial green total factor productivity has not been brought into play during this period. Based on the above results, it is found that the provinces in the Yangtze River Economic Belt should reduce carbon emissions, build a green and low-carbon product supply system, actively promote the upgrading of industries, and make efforts to improve industrial green total factor productivity.

This study analyzed the basic relationship among carbon constraints, industrial structure upgrading, and green total factor productivity with an econometric model, and the results show that the synergistic effect of carbon constraints and industrial structure upgrading has a greater impact on improving green total factor productivity. In the following sections, we further explore how carbon constraints influence green total factor productivity through the intermediary effect, and consider industrial structure upgrading as an intermediary variable for analysis. The panel regression results of Model 4–Model 6 are shown in Table 5.

In Table 5, Model 4 is the total effect model of carbon constraints on green total factor productivity, Model 5 is the direct effect model of carbon constraints on industrial structure upgrading, and Model 6 is the direct effect model of carbon constraints on green total factor productivity. It is different from previous research; this study made a detailed flow chart for the test of the intermediary effect. As can be seen from Figure 1, the coefficients of carbon constraints and industrial structure upgrading pass the significance test; δ_1,γ₁, and δ₂ are all significant, and γ₁δ₂ has the same sign as δ₁. The regression results of Model 4 show that the total effect coefficient β₁ of carbon constraints on green total factor productivity is −0.1706. The regression results of Model 5 show that carbon constraints have a positive effect on industrial structure upgrading, which verifies Hypothesis 3.

The regression results of Model 6 show that the direct effect of carbon constraints on green total factor productivity is positive. To sum up, the upgrading of the industrial structure shows a partial intermediary effect on industrial green development in terms of carbon constraints; a part of carbon constraints directly improves green total factor productivity, while a part of carbon constraints indirectly improves green total factor productivity through industrial structure upgrading, which verifies Hypothesis 4. According to the intermediary effect γ₁δ₂/β₁, the final intermediary effect is 4.5%; industrial structure upgrading has an intermediary effect of 4.5% in the process of carbon constraint's role in industrial green development. In summary, the important task of energy conservation and carbon emissions reduction should be realized by adjusting the internal structure of the three industries and improving the technical level, especially by promoting the low-carbon upgrading of secondary industry, so as to improve the level of industrial green development.

The regression results shown in Tables 4 and 5 support the hypotheses of this study, showing that carbon constraints and industrial structure upgrading influence green total factor productivity. However, each variable has a different measurement method, and the information conveyed by each method is also different. Therefore, in the robustness test, the measurement methods of some of the variables are first changed and then the empirical test is carried out. Previous findings have done less on the robustness tests, so this study can improve the related research. A robustness test can test the model better and enhance the reliability of the results. Table 6 shows the regression results of the robustness test after changing some of the indexes. The measurement method of carbon constraints is changed to CO₂ emissions per industrial added value, and there is a reverse relationship between carbon constraint intensity and this index, similar to the above measurement standard of carbon constraints. The measurement method of some of the control variables is also changed. As the control variables are not the focus of this study, their results have not been tabulated. Table 6 shows that all the regression results of Model 1–Model 6 pass the significance test, which is consistent with the regression results in Tables 4 and 5, and once again supports the hypotheses. The results of robustness tests showed that the models are not only applicable to the Yangtze River Economic Belt, but also can be generalized to other similar countries or regions, especially to developing countries.

Using data of 11 provinces in the Yangtze River Economic Belt and by calculating green total factor productivity, this study empirically tests the internal relationship among carbon constraints, industrial structure upgrading, and green total factor productivity, as well as how they influence each other, and draws the following conclusions: (1) Over time, green total factor productivity in the Yangtze River Economic Belt fluctuates and rises. In space, the green total factor productivity of provinces in the Yangtze River Economic Belt is characterized by ‘a gradient decreasing spatial distribution from the Yangtze River Delta to the Urban Agglomeration in the Middle Reaches to the Chengdu–Chongqing Economic Zone.’ (2) Carbon constraints and industrial structure upgrading have a positive effect on green total factor productivity. Their synergistic effect also has a positive effect on green total factor productivity. The coefficient of the effect of carbon constraints on green total factor productivity is greater than the coefficient of industrial structure upgrading. (3) The level of economic development and STL promote green total factor productivity, while government intervention, scale of industrial enterprises, and foreign investment inhibit green total factor productivity. (4) Carbon constraints can directly and indirectly promote green total factor productivity. In addition, carbon constraints are conducive to giving impetus to industrial structure upgrading, which, in turn, directly facilitates the green development of industry. Industrial structure upgrading has a partial intermediary effect in the process of carbon constraints promoting green total factor productivity, which accounts for 4.5% of the total effect. The above findings support the four hypotheses of this study.

The results of this study have the following policy recommendations: (1) Reducing carbon emissions and energy consumption is of great significance for improving green total factor productivity, so the publicity of green and low-carbon transformation strategy should be strengthened. Implementing the green and low-carbon transformation strategy can not only promote high-quality economic development, but also be an important measure to deal with global climate change, so as to achieve sustainable economic and ecological development. (2) Provinces in the Yangtze River Economic Belt should accelerate the upgrading of industrial structure, vigorously promote industrial energy saving and carbon reduction, comprehensively improve the efficiency of resource utilization, actively implement the transformation of clean production, and focus on building an efficient, clean, low-carbon, circular green manufacturing system. They should encourage the development of low-carbon industries, comprehensively consider the economic growth effect and carbon emission effect of various industries, and support the development of industries with good economic benefits and high ecological benefits. Also, they should promote the rationalization of industrial structure and the development of industrial structure to tertiary industry to realize the rapid improvement of industrial green total factor productivity. (3) Government departments in the Yangtze River Economic Belt should increase investment in scientific and technological innovation, optimize the investment structure of industrial technological innovation, and give full play to the role of innovation in promoting the green development of industry. Unreasonable government intervention leads to de-industrialization, which is not conducive to industrial development. Decision-makers in the Yangtze River Economic Belt should appropriately reduce the scale of industrial enterprises and foreign investment, and develop economy on the premise of reducing the pressure of carbon emissions and energy consumption on the environment, and promote a green and low-carbon transition in industry to realize the coordinated development of economy and ecology. (4) In the process of carbon constraints promoting green total factor productivity, it is necessary to give full play to the intermediary role of industrial structure upgrading. Cleaner production is an important way to realize the synergy of pollution reduction and carbon reduction, and is conducive to realizing the green and sustainable development of industry and then to promoting high-quality economic development and high-level protection of the ecological environment. (5) The Yangtze River Economic Belt is the largest inland river industrial belt and manufacturing base in China, which has abundant freshwater resources and a large number of energy-consuming enterprises. The abundant water resources are beneficial to the adjustment of the temperature change, and are also important channels for the transportation of goods. In recent years, the provinces have increased investment in clean energy, low-carbon economy, and other fields, promoted the upgrading of industrial structure, and formulated emission reduction plans to slow down the process of climate change. The ecological environment protection of the Yangtze River Economic Belt has undergone a turning change, and economic and social development has made historic achievements. The results are enlightening for the promotion of ecological protection and high-quality development of other river basins in the world, and have guiding significance for developing countries.

The authors would like to thank the editors and reviewers for their valuable time giving comments and support.

This research is supported by the Hunan Provincial Natural Science Foundation of China (Grant No. 2021JJ30304) and Hunan Province Postgraduate Research and Innovation Project (Grant No. CX20221210).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.