Micro input factors affecting industrial energy efficiency in Hebei Province based on granger causal relation test model

Abstract

Reasonably allocating the micro factor investment in the industry and trying to improve the energy efficiency are very important for the green development of industrial economy in Hebei Province, so as to the achievement of its double carbon target. This paper takes industrial energy consumption, assets and labor in Hebei Province as micro-input elements. Based on the panel data of 34 industrial industries in Hebei Province from 2005 to 2016, we identify the correlation of micro-input factors in different industries and conduct Granger causality test. The study shows that: (1) About half of the industry’s combined energy consumption is correlated with the number of employees and assets; About two-thirds of the industry’s asset workers are relevant; (2) There is a two-way causal relationship between the total energy consumption and assets of most industries and the annual average number of employees. (3) The rational allocation of capital, labor and energy consumption depends on the deepening reform of the factor market. The high energy consumption industry should actively improve and innovate the energy technology in this process to realize the coordinated promotion of energy saving and carbon reduction, and thus to promote the healthy and sustainable development of the industry steadily.

Keywords

Energy efficiency micro-elements Granger causality test

1. Introduction

According to the National Economic and Social Development Statistics Bulletin of Hebei Province in 2020, the total industrial added value of Hebei Province in 2020 was 1154.59 billion yuan, and has increased of 4.6% over the previous year, in which the industrial added value above the scale increase by 4.7%. At the same time, energy conservation and consumption reduction are also steadily advancing, industrial energy consumption has kept decreasing, and energy consumption per unit industrial added value of the province has decreased by 4.58%. However, despite the continuous efforts of Hebei Province to accelerate the transformation from a large industrial province to a strong industrial province, the characteristics of high economic growth, high consumption, high emission, low efficiency in Hebei Province are still very prominent, the situation of energy saving and consumption reduction is serious. Based on this, through the research on the driving factors of energy utilization efficiency and its interactive relationship, the paper helps Hebei government to find the key breakthrough to improve energy efficiency and promote industrial transformation, to form the industrial structure strategy for improving Hebei industrial energy efficiency. Realize the high-quality development of industrial industry and build Hebei into a resource-saving society with low resource consumption, less environmental pollution and good economic benefits.

At present, the domestic and international research results on industrial efficiency and its influencing factors are relatively rich, the main research methods can be divided into: Factor decomposition method, Stochastic Frontier Approach, Data Envelopment Analysis and spatial measurement methods. Factor decomposition method emphasizes energy efficiency and intensity [1]. For example, Wu [2] believes that the main factors driving industrial green transformation are energy technology reform, energy structure adjustment and environmental effects of new energy sources. The stochastic frontier production function can calculate the difference of energy efficiency, and decompose the deviation of decision-making unit from the frontier into two parts: Technical efficiency and stochastic disturbance. Wu and Li [3] took the total factor energy efficiency of Shandong province as the research object, estimated the parameters in the stochastic frontier by the Bayesian estimation method of a priori distribution. DEA has been widely used in the field of industrial efficiency research [4]. Taking the inter-provincial industrial energy efficiency of China as the research object, Xu and Li [5] use the empirical study of DEA method to show that the scale of industrial enterprises, foreign investment and technological progress have a positive impact on energy efficiency, among which foreign investment has a greater impact. Wu and Gao [6] used the DEA-Malmquist productivity index to construct an econometric model to test the relationship between energy price, TFP and energy intensity. The impact of efficiency improvement, technological progress and relative energy prices on industrial energy intensity was investigated from the perspective of industry heterogeneity [7]. He et al. [8] combine DEA with rough set theory (RS) and fuzzy artificial neural network (FANN) to classify the provincial regions according to the energy efficiency level, taking into account the non-linear and hysteresis effects between energy efficiency and influencing factors.

In addition, some scholars have discussed the influence factors of industrial efficiency through the construction of spatial model [9, 10]. Using dynamic spatial Durbin model, Han et al. [11] discussed the mechanism behind the influence of urban agglomeration economy on industrial energy efficiency. Dong et al. [12] constructed three new techno-economic correlation matrices. Spatial autoregressive combinatorial model is used to study the spillover effect of industrial integration, which confirms the spillover effect of industrial association.

To sum up, the existing literature is rich in the estimation of industrial energy efficiency and the investigation of the influencing factors, methods and emphasis are different. Some scholars pay attention to the regional differences of the influencing factors of industrial energy efficiency; Some scholars examine the channels through which the factors affecting industrial energy play a role [13]. As we all know, how to improve the efficiency of energy utilization is an urgent problem that needs to be solved in engineering science and technology circles. Especially in the period of accelerated development of industrialization and urbanization, in the face of the country’s strategic task of adjusting the structure and changing the mode, the requirements for industrial energy conservation are higher. Micro-input factors involve all aspects of energy-saving and emission reduction in key energy-consuming industries and production processes. In order to optimize the distribution of factors, improve the allocation efficiency of production factors, actively tap the growth potential, and promote the development of high quality industry, this paper pays more attention to the correlation between the factors of energy efficiency.

2. Correlation analysis of influencing factors of industrial energy efficiency in Hebei Province

2.1 Overview of correlation measurement methods among factors and data sources

Correlation analysis is a kind of data method to judge whether the variables are related to each other, and to judge the degree of correlation among the related factors [14, 15]. When the correlation coefficient is large, the closer the relationship between variables is, the lower the correlation coefficient is, the lower the degree of closeness is. Generally, when the correlation coefficient is less than 0.5, the correlation among variables is low. The variables were significantly correlated between 0.5–0.8; Above 0.8 is highly correlated.

The correlation analysis method is introduced to analyze the correlation between each index, and the correlation coefficient between any two indexes is solved by Eq. (1):

$\displaystyle R=\frac{\sum{(x_{i}-\bar{X})(y_{i}-\bar{Y})}}{\sqrt{\sum{(x_{i}-% \bar{X})^{2}}}\sqrt{\sum{(y_{i}-\bar{Y})^{2}}}}$ (1)

where, $x_{i}$ and $y_{i}$ are the $i$ th sampling value of two indicators $x$ and $y$ respectively, $\bar{X}$ and $\bar{Y}$ are the average values of the two indicators respectively, namely:

$\displaystyle\bar{X}=\frac{{1}}{n}\sum\limits_{i=1}^{n}{x_{i}},\bar{Y}=\frac{1% }{n}\sum\limits_{i=1}^{n}{y_{i}},i=1,2\ldots,n$ (2)

where $n$ is the number of samples. The Eq. (2) is embedded in Eq. (1) to construct a correlation coefficient matrix such as Eq. (3). Where $R_{j}^{ds}$ is the correlation coefficient of the d th and s th index in the index system:

$\displaystyle R=\left[{{\begin{array}[]{*{20}c}1&&&&&\\ {R_{j}^{21}}&1&&&&\\ \\ {R_{j}^{31}}&{R_{j}^{32}}&1&&&\\ \\ {R_{j}^{41}}&{R_{j}^{42}}&\cdots&1&&\\ \\ \cdots&{R_{j}^{m1}}&\cdots&{R_{j}^{ms}}&1&\\ \\ {R_{j}^{d1}}&{R_{j}^{d2}}&\cdots&{R_{j}^{ds}}&\cdots&1\\ \end{array}}}\right]$ (3)

The data used in this paper, energy consumption, average annual number of all employees and total assets are derived from the Economic Yearbook of Hebei Province for 2005–2016. Industrial enterprises in this paper designated SIE are all state-owned enterprises and non-state owned enterprises with annual revenue from principal business over 5 million yuan from1998 to 206, and are industrial enterprise with annual revenue from principal business over 5 million yuan from 2007 to 2010, and are industrial enterprise with annual revenue from principal business over 20 million yuan since 2011. The equation for calculating the annual average number of all employees is:

$\displaystyle L_{i}=\frac{L}{N}\cdot\frac{W_{i}}{W}$

where $L_{i}$ is the number of employees in the i-th industry under the mining industry category, $L$ is the total number of employees in the mining industry, $N$ is the number of industries under the mining industry category, $W_{i}$ is the average salary of employees in the i-th industry under the mining industry category, and $W$ is the average wages of those employed in the mining industry. In the same way, the number of employees in the manufacturing industry category, as well as the production and supply of electricity, gas and water, can be derived.

2.2 Stationary test

If there is a random process $Y$ whose mean, variance and covariance are invariant, then we can call it a stationary process:

$\displaystyle E(Y_{t+k})=E(Y_{t});\textit{Var}(Y_{t+k})=\textit{Var}(Y_{t});% \textit{Cov}(Y_{t},Y_{t+k})=\textit{Cov}(Y_{t+h},Y_{t+k})$

Then for all $t$ , $h$ and $k$ , if the time series $Y_{t}$ is stationary, it is denoted as $Y_{t}\sim I(0)$ . It is noteworthy that if a non-stationary time series is stable after the first order difference, we can refer to this time series as the first order simple whole sequence, denoted as $Y_{t}\sim I(1)$ ; In the same way, when a certain time series passes the difference of order $n$ before showing the stable property, we think that the time series is a simple whole sequence of order $n$ , denoted as $Yt\sim I(n)$ . Most economic time series are difficult to maintain a stable state, but when using cointegration analysis and Granger causality test variables, it is necessary to satisfy the precondition that the time series has a stable state, otherwise it may cause false (pseudo) regression. Therefore, before we do a cointegration analysis of two or more variables and a Granger causality test, we need to verify that the time series of these variables is stable. This paper verifies that the variables are stationary by means of the ADF unit root test, as shown below:

$\displaystyle\Delta Y_{t}=\phi Y_{t-1}+\sum\limits_{i=1}^{n}\delta_{i}\Delta Y% _{t-i}+\varepsilon_{t}$ (4) $\displaystyle\Delta Y_{t}=\alpha+\phi Y_{t-1}+\sum\limits_{i=1}^{n}\delta_{i}% \Delta Y_{t-i}+\varepsilon_{t}$ (5) $\displaystyle\Delta Y_{t}=\alpha+\beta t+\phi Y_{t-1}+\sum\limits_{i=1}^{n}% \delta_{i}\Delta Y_{t-i}+\varepsilon_{t}$ (6)

In the above model, $\alpha$ denote a constant, $t$ denote a time trend variable, and $n$ denote a delay order. Moreover, $\varepsilon_{t}\sim i.i.d.N(0,\sigma^{2})$ . It is obvious that there is neither constant nor time trend in the Eq. (4); Although there is a constant term in the Eq. (5), there is no time trend. The Eq. (6) contains not only constant terms, but also time. Generally speaking, the original hypotheses of the model are: $H_{0}:\phi=0$ , indicating the existence of unit roots in time series; The alternative assumption of the model is $H_{1}:\phi<0$ , there is no unit root in the time series. If the statistic of the ADF test is less than the critical value, the original hypothesis $H_{0}$ can be rejected and the detected variable time series can be considered as stable. If the calculated ADF test statistic is not less than cut-off value, then $H_{0}$ is accepted, we can consider the detected variable time series as non-stationary.

Energy, labor, and assets are the basic production factors of industry, as well as important indicators to measure the operation of industrial enterprises. Three microcosmic influencing factors of industrial energy efficiency in Hebei Province: Comprehensive energy consumption (energy), annual average number of employees (labor) and total assets (asset) were tested. From the perspective of the whole industry, the correlation coefficient between total assets and comprehensive energy consumption is 0.7813, and the results are all significant at the level of 1%. Therefore, total assets are moderately related to combined energy consumption. The test results are shown in Table 1.

Table 1

Summary of correlation analysis results

Correlation	Industry
The average number of employees per year is highly correlated with integrated energy consumption (7 industries)	Paper and paper products industry, education, arts and crafts, sports and entertainment goods manufacturing industry, metal products industry, electrical machinery and equipment manufacturing industry, instrumentation manufacturing industry, waste resources integrated utilization industry, electricity, thermal production and supply industry
Annual average number of all employees is moderately related to integrated energy consumption (12 industries)	Non-metallic mining and dressing, tobacco manufacturing, textiles, textile apparel, wood processing and wood, bamboo, rattan, brown, grass manufacturing, furniture manufacturing, printing and recording media reproduction, pharmaceutical manufacturing, chemical fibre manufacturing, ferrous metal smelting and calendering, general equipment manufacturing, computer, communications and other electronic equipment manufacturing
The average number of employees per year is correlated with low integrated energy consumption (2 industries)	Coal mining and washing, gas production and supply
The annual average number of all employees is largely irrelevant to integrated energy consumption (14 industries)	Ferrous metal mining and dressing industry, agronon-staple food processing industry, food manufacturing industry, wine, beverage and refined tea manufacturing industry, leather, fur, feathers and their products and footwear industry, petroleum processing, coking and nuclear fuel processing industry, chemical raw materials and chemical products manufacturing industry, rubber and plastic products industry, non-metallic mineral products industry, non-ferrous metal smelting and rolling processing industry, special equipment manufacturing industry, other manufacturing industry, water production and supply industry, whole industry
Total assets are highly correlated with integrated energy consumption (13 industries)	Coal mining and washing industry, ferrous metal mining and dressing industry, tobacco products industry, paper and paper products industry, printing and recording medium reproduction industry, culture and education, engineering, sports and entertainment goods manufacturing industry, metal products industry, special equipment manufacturing industry, electrical machinery and equipment manufacturing industry, instrument and meter manufacturing industry, waste resources comprehensive utilization industry, electric power, thermal production and supply industry, industry-wide
Total Assets Moderately Associated with Integrated Energy Consumption (8 industries)	Non-metallic mining and dressing, textile, textile apparel wood processing and wood, bamboo, rattan, brown, straw, furniture, chemical fibre, general equipment, gas production and supply
Total assets are associated with low integrated energy consumption (0 industries)	None
Total assets are largely irrelevant to combined energy consumption (14 industries)	Agro-and non-staple food processing industries, food manufacturing relevance, wine, beverage and refined tea manufacturing industries, leather, fur, feathers and their products and footwear industries, petroleum processing, coking and nuclear fuel processing industries, chemical raw materials and chemicals manufacturing industries, pharmaceutical manufacturing industries, rubber and plastic products industries, non-metallic mineral products industries, ferrous metal smelting and calendering industries, non-ferrous metal smelting and calendering industries, other manufacturing industries, water production and supply industries
Total assets are highly correlated with the annual average number of all employees (20 industries)	Agricultural and non-staple food processing industry, food manufacturing industry, wine, beverage and refined tea manufacturing industry, tobacco manufacturing industry, textile industry, textile apparel industry, leather, fur, feathers and their products and footwear industry, wood processing and wood, bamboo, rattan, brown, grass manufacturing industry, furniture manufacturing industry, paper and paper products industry, printing and recording medium reproduction industry, education, arts and crafts, sports and entertainment products industry, chemical fiber manufacturing industry, rubber and plastic products industry, metal products industry, electrical machinery and equipment manufacturing industry, instrumentation manufacturing industry, waste resource utilization industry, electricity, thermal production and supply industry, gas production and supply industry, water production and supply industry

Table 1, continued
Correlation	Industry
Total assets are moderately related to the annual average number of all employees (5 industries)	Coal mining and cleaning, non-metallic mining and cleaning, pharmaceutical manufacturing, specialized equipment manufacturing, computer, communications and other electronic equipment manufacturing
Total assets are associated with a low annual average number of all employees (0 industries)	None
Total assets are largely irrelevant to the annual average number of all employees (10 industries)	Ferrous metal mining and processing, petroleum processing, coking and nuclear fuel processing, chemical raw materials and chemicals manufacturing, non-metallic mineral products, ferrous metal smelting and rolling processing, non-ferrous metal smelting and rolling processing, general equipment manufacturing, other manufacturing, industry-wide

In conclusion, the annual average number of all employees in about half of the industries is correlated with the integrated energy consumption, and the total assets is correlated with the integrated energy consumption; About two-thirds of the total assets in the industry are correlated with the annual average number of all employees.

Economic and social development depends on energy, but energy is limited, which requires guiding high-energy-consuming enterprises to improve technology, strengthen management, and improve energy efficiency. Whether the relationship between industrial energy efficiency and other factors is scientific and reasonable is directly related to the effectiveness of energy conservation and carbon reduction in key industries. Therefore, continuous optimization and dynamic adjustment are required.

3. Causality between industrial energy efficiency and microscopic influencing factors in Hebei Province

3.1 Summary of Granger causality test method

Granger causality is defined as “the variance of the best least squares prediction that relies on all information at some point in the past” [16]. In the case of time series, the Granger causality between the two variables X and Y can be defined as “if the prediction effect of the variable Y is better than the prediction effect of only the past information of Y under the condition that the past information of the variables X and Y is included, then the variable X is the Granger cause of the variable Y” According to the definition of Granger causality, judge whether there is Granger causality between X and Y. Granger causality expresses statistical correlation, which is the continuity of phenomena in the sense of time.

The Granger causality test for the two variables X and Y requires the construction of a regression equation containing the hysteresis terms (sum) of X and Y, as shown in Eqs (7) and (8).

$\displaystyle y_{t}=\sum\limits_{i=1}^{q}{\alpha_{i}x_{t-i}+\sum\limits_{j=1}^% {q}{\beta_{j}y_{t-j}+u_{1t}}}$ (7) $\displaystyle x_{t}=\sum\limits_{i=1}^{s}{\lambda_{i}x_{t-i}+\sum\limits_{j=1}% ^{s}{\delta_{j}y_{t-j}+u_{2t}}}$ (8)

where, $x_{t-i}$ is the lagging term of $x_{t}$ ; $y_{t-j}$ is the lagging term of $y_{t}$ ; $q$ is the lag length in the regression equation of variable Y; $i$ and $j$ are the number of lag terms; $s$ is the lag length in the regression equation of variable X, and the maximum value of the lag length $q$ and $s$ is the order of the regression model. $\mu_{1t}$ and $\mu_{2t}$ are white noises; $\alpha_{i}$ and $\lambda_{i}$ are coefficient estimates of $x$ ; $\beta_{j}$ and $\delta_{j}$ are coefficient estimates for $y$ . If $\beta_{j}$ ( $j=1,\ldots,q$ ) is statistically significant not zero overall, then X is the Granger cause of Y. Similarly, if $\delta_{j}$ ( $j=1,\ldots,s$ ) is statistically significant not zero overall, then Y is the Granger cause of X.

3.2 Test results

In this section, Granger causality test was conducted among three microcosmic factors of total factor productivity (TFP) of industrial energy in Hebei province, comprehensive energy consumption (energy), annual average number of employees (labor) and total assets (asset). In order to ensure the stability of each variable and prevent the phenomenon of pseudo-regression, the stability of each variable is tested first, and the test results are shown in attached Table S1. According to the test results in Appendix 1, Granger causality test is conducted for stable variables in various industries. See Appendix 2 for the test results. According to the First and Second Schedules, the causal relationship between the comprehensive energy consumption of each industry, the total assets and the annual average number of employees is summarized as shown in Table 2.

Table 2
Summary of causality of energy consumption, labor and asset level in each industry

Causality	Industry
Annual average number of all employees $\to$ Integrated energy consumption (4 industries)	Wood Processing and Wood, Bamboo, Rattan, Brown, Grass Products, Culture, Education, Engineering, Sports and Entertainment Goods Manufacturing, Electrical Machinery and Equipment Manufacturing
Integrated energy consumption $\to$ Annual average number of all employees (9 industries)	Ferrous metal mining and dressing, non-metal mining and dressing, printing and recording medium reproduction, non-ferrous metal smelting and calendering, metal products, special equipment manufacturing, other manufacturing, comprehensive utilization of waste resources, gas production and supply
Annual Average Number of All Employees $\mathbin{\lower 1.29pt\hbox{$\buildrel\textstyle\leftarrow\over{\smash{% \rightarrow}\vphantom{{}_{\vbox to 2.15pt{\vss}}}}$}}$ Integrated Energy Consumption (17 industries)	Agricultural and non-staple food processing industries, the textile industry, the textile and clothing industry, the garment industry, the leather, fur, feathers and their products and footwear industry, the furniture industry, the paper and paper products industry, the petroleum processing, coking and nuclear fuel processing industry, the rubber and plastic products industry, the non-metallic mineral products industry, the general equipment manufacturing industry, the computer, communications and other electronic equipment manufacturing industry, the instrumentation manufacturing industry, the electric power, the thermal production and supply industry, the water production and supply industry
Annual Average Number of All Employees $\mathrel{\hbox to 0.0pt{\hbox to 7.499886pt{\hss/\hss}}\hbox{$\Leftrightarrow$}}$ Integrated Energy Consumption (3 industries)	Industry-wide, pharmaceutical manufacturing, ferrous metal smelting and rolling
Total Assets $\to$ Integrated Energy Consumption (5 industries)	Agronon-staple food processing industry, wine, beverage and refined tea manufacturing industry, rubber and plastic products industry, electricity, thermal production and supply industry, water production and supply industry
Consolidated Energy Consumption $\to$ Total Assets (4 industries)	Ferrous metal ore mining and processing industry, non-metal ore mining and processing industry, non-ferrous metal smelting and calendering industry, metal products industry
Total Assets $\mathbin{\lower 1.29pt\hbox{$\buildrel\textstyle\leftarrow\over{\smash{% \rightarrow}\vphantom{{}_{\vbox to 2.15pt{\vss}}}}$}}$ Integrated Energy Consumption (19 industries)	Textile, textile apparel, apparel, leather, fur, feathers and their products and footwear, wood processing and wood, bamboo, rattan, brown, straw, furniture, paper and paper products, printing and recording medium reproduction, education, education, sports and entertainment, petroleum processing, coking and nuclear fuel processing, chemical fibre manufacturing, non-metallic mineral products, ferrous metal smelting and calendering, general equipment manufacturing, specialized equipment manufacturing, electrical machinery and equipment manufacturing, instrumentation manufacturing, comprehensive utilization of waste resources, gas production and supply
Total Assets $\mathrel{\hbox to 0.0pt{\hbox to 7.499886pt{\hss/\hss}}\hbox{$\Leftrightarrow$}}$ Integrated Energy Consumption (3 industries)	Industry-wide, coal mining and washing, other manufacturing industries
Total Assets $\to$ Annual Average of All Employees (5 industries)	Agricultural and non-staple food processing industry, tobacco industry, textile industry, textile apparel industry, clothing industry, leather, fur, feathers and their products and footwear industry
Annual average number of all employees $\to$ Total assets (3 industries)	Paper and paper products industry, education, arts and crafts, sports and entertainment goods manufacturing, ferrous metal smelting and calendering industry
Total Assets $\mathbin{\lower 1.29pt\hbox{$\buildrel\textstyle\leftarrow\over{\smash{% \rightarrow}\vphantom{{}_{\vbox to 2.15pt{\vss}}}}$}}$ Annual Average of All Employees (21 industries)	Ferrous metal mining and dressing, non-metal mining and dressing, food manufacturing, wood processing and wood, bamboo, rattan, brown, straw, furniture manufacturing, printing and recording medium reproduction, petroleum processing, coking and nuclear fuel processing, chemical raw materials and chemical products, rubber and plastics, non-metal mineral products, non-ferrous metal smelting and calendering, metal products, general equipment manufacturing, specialized equipment manufacturing, electrical machinery and equipment manufacturing, instrumentation manufacturing, other manufacturing, integrated utilization of waste resources, electricity, thermal production and supply, gas production and supply, water production and supply
Total Assets $\mathrel{\hbox to 0.0pt{\hbox to 7.499886pt{\hss/\hss}}\hbox{$\Leftrightarrow$}}$ Annual Average of All Employees (2 industries)	Industry, wine, beverage and refined tea manufacturing

In conclusion, there is a two-way causal relationship between the total comprehensive energy consumption and assets of most industries and the annual average number of all employees, there is a one-way causal relationship between three variables of a small number of industries, and there is no causal relationship between three variables of a very small number of industries.

Ferrous metal smelting and rolling processing industry, non-metallic mineral product industry, electricity and heat production and supply industry, coal mining and washing industry, chemical raw material and chemical product manufacturing industry, petroleum processing coking and nuclear fuel processing industry are the six major energy-intensive industries in Hebei Province. In 2015, the energy consumption per unit of industrial added value in Hebei province dropped by 24% compared with 2010, and the contribution rate of the six major energy-consuming industries to the province’s industrial growth has dropped from 33.7% in 2010 to 29.4% in 2015.

There is a causal relationship between the total assets of these three industries and the comprehensive energy consumption (non-metallic mineral product industry, petroleum processing coking and nuclear fuel processing industry and ferrous metal smelting and rolling processing industry). It is worth emphasizing that the energy consumption of the output value of the above six industries all showed a downward trend during the study period. For instance, the energy consumption per unit GDP of coal mining and washing industry dropped from 2.87 in 2005 to 1.2 in 2016. The same indicator of petroleum processing, coking and nuclear fuel processing industries, chemical raw materials and chemical products manufacturing, non-metallic mineral products industry fall back more than 1.5 times. It may related to the province’s strict implementation of energy consumption limits, special pollutant discharge limits, differential electricity price and water price policies, and the establishment of special funds for eliminating outdated production capacity.

Among the 34 sub-sectors of Hebei’s industry, coal mining, chemical materials and ferrous metal smelting and rolling processing industries such as steel, petrochemical, and chemical industries have significantly higher energy consumption per unit of GDP than other industries. Under the goal of carbon neutrality, those industry in Hebei Province needs to consolidate the results of capacity reduction and resolutely curb the blind development of resource-based heavy chemical industries such as the “High pollution and high energy consumption” projects [17]. Under the policy constraints of emission reduction, the steel, petrochemical, and chemical industries have huge potential to implement industrial low-carbon transformations and coordinate to reduce energy consumption, so as to avoid excessive consumption and expansion of assets to energy inputs. Relevant industries should pay attention to the energy-saving efficiency of asset investment projects, speed up the renewal and iteration of production equipment, and improve the energy efficiency of equipment.

Nowadays, the development of various industries in Hebei Province, especially the development of high-tech manufacturing, is increasingly dependent on innovation. This requires high-quality labor to cooperate with green technology research and development projects invested by enterprises to strengthen the main position of enterprise innovation, thereby accelerating the promotion of green upgrading of manufacturing.

4. Conclusions

(1) About half of the combined energy consumption in industry is related to the number of employees and assets; About two-thirds of the industry’s assets are correlated with the number of employees. Most of the industries which are highly correlated with the integrated energy consumption are light industries in the manufacturing industry classification, showing the demand for labor in the energy consumption process of some industries. In contrast, manufacturing and processing industries such as metals and non-metals show an irrelevance between assets and labor, suggesting that labor mobility among these industries is not sensitive to asset movements.

(2) The total energy consumption and assets of most industries are related to the average annual number of employees. The energy consumption of metal smelting and processing has a significant Granger causal relationship with the assets. Besides, the energy consumption of ferrous metal mining and dressing industry, non-metallic mining and dressing industry, non-ferrous metal smelting and rolling processing industry are all Granger causes of assets. Especially the ferrous metal smelting and rolling processing industry represented by steel products has the same feature, which shows that the traditional heavy industry growth depends on the short board of energy consumption and restricts the other investment behavior of the assets.

(3) Rational allocation of assets, labor force and energy consumption are important component factors to promote the growth of industrial green economy in China. In the future, the green development of China’s industry can not be separated from the steady improvement of energy efficiency, in which the changes of the interrelation of the traditional investment elements on the road of industrial upgrading should be consistent with the requirements of green development and reduction of energy consumption in the development plan of the period of the “14th Five-Year Plan” Enterprises in the energy-intensive industries have the motivation to pay attention to the technological investment in energy conservation, emission reduction and clean technology research, constituting a good development pattern with reasonable allocation of input elements.

Footnotes

Acknowledgments

The authors gratefully appreciate the reviewers for their constructive comments and suggestions. The authors declare no conflict of interest.

Funding

This research was funded by (1) The project of the State Grid Hebei Electric Power Co., Ltd. The topic is “Research on New Power System Development Path”. (2) The American Energy Foundation: “Research on Low-Carbon Consumption Models and Low-Carbon Society Construction in the Context of Carbon Neutrality”.

Appendix A

The Schedule 1 for this paper is shown in Appendix word.

References

Lin

. Quantitative assessment of factors affecting energy intensity from sector, region and time perspectives using decomposition method: A case of China’s metallurgical industry. Energy. 2019; 189: 116280.

. Can the new energy revolution promote the green transformation of china’s industry? Empirical analysis based on factor decomposition. Econ Syst Reform. 2017; (2): 184-191.

. Study on industrial energy efficiency and its influencing factors in shandong province. China Popul Res Environ. 2015; 25(6): 114-120.

Zhu

Zhou

Liang

. Energy and environmental efficiency measurement of China’s industrial sectors: A DEA model with non-homogeneous inputs and outputs. Energy Econ. 2019; 78: 468-480.

. Analysis of industrial energy efficiency and its influencing factors in China. Stat Decis Making. 2015; (18): 132-138.

Gao

. Empirical test on the relationship between energy price, total factor productivity and industrial energy intensity. Stat Decis Making. 2019; 35(16): 122-128.

Zhou

. Effect of energy price, efficiency and technology progress on heterogeneity of energy intensity in industry. Study Quant Econ Tech Econ. 2016; 33: 130-143.

Liao

Zhou

. Analysis on provincial industrial energy efficiency and its influencing factors in China based on DEA-RS-FANN. Energy. 2018; 142: 79-89.

Liu

Lin

Xiao

. Spatial-temporal characteristics of industrial land use efficiency in provincial China based on a stochastic frontier production function approach. J Cleaner Prod. 2021; 295: 126432.

10.

Ouyang

Chen

. Energy efficiency performance of the industrial sector: From the perspective of technological gap in different regions in China. Energy. 2021; 214: 118865.

11.

Han

Xie

Fang

. Urban agglomeration economies and industrial energy efficiency. Energy. 2018; 162: 45-59.

12.

Dong

Zhang

Zhu

Zheng

. How does industrial convergence affect the energy efficiency of manufacturing in newly industrialized countries? Fresh evidence from China. J Cleaner Prod. 2021; 316: 128316.

13.

Wang

Yao

. Evaluation and difference analysis of regional energy efficiency in China under the carbon neutrality targets: Insights from DEA and Theil models. J Environ Manage. 2021; 293: 112958.

14.

Reimann

Filzmoser

Hron

Kynčlová

Garrett

. A new method for correlation analysis of compositional (environmental) data – A worked example. Sci Total Environ. 2017; 607-608: 965-971.

15.

Qin

Liu

. Human capital services, spatial spillovers and economic growth. Quant Econ Res. 2021; 12(3): 112-131.

16.

Vilasuso

. Causality tests and conditional heteroskedasticity: Monte Carlo evidence. J Econ. 2001; 101(1): 25-35.

17.

Granderson

Linvill

. Regulation, efficiency, and Granger causality. Int J Ind O. 2002; 20(9): 1225-1245.

Micro input factors affecting industrial energy efficiency in Hebei Province based on granger causal relation test model

Abstract

Keywords

1. Introduction

2. Correlation analysis of influencing factors of industrial energy efficiency in Hebei Province

2.1 Overview of correlation measurement methods among factors and data sources

3.1 Summary of Granger causality test method

Table 2 Summary of causality of energy consumption, labor and asset level in each industry

Footnotes

Acknowledgments

Funding

Appendix A

References

Table 2
Summary of causality of energy consumption, labor and asset level in each industry