Accounting and Decomposition of Energy Footprint: Evidence from 28 Sectors in China

Abstract

This study aims to clarify the sectoral level of environmental pollution “rights, responsibilities and benefits” and to identify the sectoral pollution “background” to lay the foundation to achieving sustainable economic development. We use input output table data to account for and decompose the sectoral energy footprint. Firstly, based on the principles of producer responsibility and consumer responsibility, the consumption-based energy footprint (CBEF) and the responsible-based energy footprint (RBEF) are accounted for. Secondly, the sectoral energy footprint is decomposed based on energy consumption and responsibility and direct and indirect perspectives. The results show that (1) the distribution of the sectoral CBEF is characterized by a high sector concentration and large inter-sector differences. (2) The distribution of the sectoral RBEF is more balanced, and the difference is smaller. (3) There are also asymmetries and heterogeneity in trends between the sectoral CBEF and the sectoral RBEF. (4) The energy footprint generated by the production of intermediate-use products is an important source of the sectoral energy footprint (EF). The Chinese government should develop differentiated energy saving and emission reduction measures and optimize the sectoral structure to enhance sectoral cleanliness. Policy references for energy saving and emission reduction at the sectoral level and early achievement of carbon-peak and carbon-neutral targets are proposed.

1. Introduction

The ecological footprint is an indicator that affects global climate change [1]. The development of a low-carbon economy and a circular economy and the realization of green recovery have become world trends. As the world’s second-largest economy and the largest carbon emitter, China’s government first proposed an ambitious blueprint of carbon peaking and carbon neutrality at the United Nations Climate Conference and included the promotion of “dual-carbon” in the overall layout of ecological civilization construction from the national strategic level. In order to steadily achieve the goal of carbon-peak and carbon-neutrality goals (referred to as the “dual-carbon” goal), a report by the 20th Party Congress once again proposed the active and steady achievement of carbon peaking and carbon neutrality based on resource endowment, which also made the study of China’s ecological footprint easy to operate [2]. Compared with global ecological and environmental problems, China’s ecological and environmental problems are more prominent. Since the reform and opening up, China has created an “economic miracle”, but at the same time, problems of excessive resource consumption and serious ecological damage have also appeared in the form of “spatial and temporal compression”, and the problems of environmental damage and ecological overload have become increasingly prominent [3,4]. Studies by domestic scholars have also shown that China’s ecological footprint exceeds 89% of its natural renewal capacity, and more than 85% of Chinese provinces are in long-term ecological debt (Xie Heights et al., 2010; Chen Xianpeng et al., 2022) [5,6]. As an essential part of the energy footprint (EF), the effective decomposition of the EF to the sectoral economy is the key to achieving the “dual-carbon” target. The construction of a scientific and reasonable EF reduction assessment index is the basis for the effective decomposition of the “dual-carbon” target. For this reason, there is an urgent need to construct a reasonable energy footprint assessment index based on a sectoral perspective. Studying this issue can provide references for the decomposition of dual-carbon targets, the initial setting of carbon emission rights, and the formulation of sectoral EF reduction policies.

Under the “dual-carbon” goal, China urgently needs to save energy and reduce carbon emissions. On the one hand, in order to measure the negative externalities generated by energy consumption, most previous studies used a single carbon dioxide emission measure to characterize environmental pollution, and a comprehensive indicator to represent the impact of energy consumption on the whole society and ecology is lacking. In this paper, we propose the use of the EF as a comprehensive indicator to represent energy conservation and carbon reduction, and the EF is also an overall measure of carbon emissions. On the other hand, no scholars have analyzed the measurement of how to reduce the EF and, thus, environmental pollution at the sector level is based on a meso perspective. This paper intends to analyze it from a meso perspective to account for and decompose the energy footprints of different sectors and to adopt different approaches to solve the problem of carbon emission reduction for different sectors.

In order to construct a scientific and reasonable environmental responsibility index, reference is made to the sectoral carbon emission assessment index. The main idea of the article is as follows: according to the consumer responsibility principle, consumers should be responsible for the implicit EF emissions in the products they consume, which has the advantage of accounting for both direct and indirect EF emissions. The CBEF and RBEF reflect the sectoral responsibility in terms of EF emissions as a producer and consumer, respectively. Comparing the two by category helps to develop differentiated energy efficiency and emission reduction policies and helps to improve the targeting and implementation results. Both theory and practice have proved that there is a difference between the sectoral CBEF and the sectoral RBEF in China. Therefore, this paper introduces the principle of consumer responsibility into the sector analysis and proposes the concept of the EF to clarify the EF emission responsibility relationship among sectors. At the same time, based on China’s data, the sectoral RBEF and the sectoral CBEF are compared to clarify the emission characteristics and change trends for China’s sector EF.

The possible contributions of this paper are as follows: (1) This paper selects the EF as a comprehensive indicator to measure energy conservation and emission reduction, which can also be used to characterize the impacts of energy consumption on human society and the ecological environment. (2) Based on the producer responsibility principle, the concept of the CBEF is proposed, and the CBEF accounting standard is constructed. Based on the consumer responsibility principle, the concept of the RBEF is proposed, and the RBEF index is constructed. (3) Accounting for the CBEF and the RBEF of each sector in China, we compare the differences, characteristics, and trends between the two in each sector. This determines which sectors produce an EF, how big the EF of each of these sectors is, and who is responsible for generating the EF. We also investigate how big the RBEF of each sector producer and consumer is. (4) This paper decomposes the EF based on the dual perspectives of energy consumption and responsibility, as well as directly and indirectly. We provide empirical support and a policy reference for the identification of key sectors for sectoral structure optimization and energy conservation and emission reduction policies and the design of energy conservation and emission reduction strategies for sectors with clear “rights, responsibilities, and benefits”.

The rest of the paper is organized as follows: The second part reviews existing studies on the ecological footprint, EF, and their relationships with environmental pollution and sectoral structure; the third part builds a theoretical framework containing accounting principles and accounting methods from a theoretical point of view based on the principle of producer responsibility and the principle of consumer responsibility according to the non-competitive input output table and introduces the aggregate data; the fourth part empirically analyzes the results for the sectoral EF and its change trends. The fifth part summarizes the whole text and puts forward policy recommendations.

2. Literature Review

2.1. Accounting for Environmental Pollution

2.1.1. Environmental Pollution and Its Measurement by Traditional Methods

In terms of the accounting and decomposition of environmental pollution, since ecological pollution affects the growth quality of the Chinese economy, Xiao et al. (2011) measured the economic loss caused by environmental pollution in China. They studied the relationship between ecological pollution loss and economic growth [7]. Regarding the selection of environmental pollution indicators, a single indicator was initially used to represent environmental pollution. Chen et al. (2015) used SO2 emissions to describe ecological pollution based on inter-provincial panel data and found an inverted U-shaped relationship between the per capita income and environmental pollution [8]. Deng et al. (2014) used CO₂ emissions to represent environmental pollution and showed that the scale effect was positively related to carbon emissions, while the technology effect was negatively related [9]. Qiu et al. (2011) used wastewater, exhaust gas, and solid emissions to represent pollutant emissions from a provincial perspective. Then, they characterized environmental pollution, verifying that the environmental Kuznets curve conforms to an inverted U-shaped trend in China [10]. Zhang et al. (2020) used carbon emissions and different pollutants to characterize environmental pollution separately. They found differences in the EKC curves formed between the per capita emissions of various emissions and the GDP deflator [11]. Cui et al. (2019) used “three wastes” to calculate the comprehensive index [12]. However, in fact, the above two indicator calculation methods have shortcomings, mainly including the following two limitations (Yuan et al., 2011) [13]: Firstly, because China is in the process of urbanization and industrialization, many kinds of pollutants are growing explosively, and many different types of environmental pollutants show the characteristics of compound pollution. If only one pollutant is selected or if the effects of the three wastes are synthesized, then the process is, in a sense, detached from reality, which makes it difficult to comprehensively reflect the true level of environmental pollution; secondly, because different pollutants have different scopes, pathways, and degrees of impact on the environment, the selection of contaminants that cannot be selected in the same region may lead to diverse conclusions.

Some scholars also choose comprehensive indicators to measure environmental pollution. Cao et al. (2013) selected six major categories of ecological indicators for 20 sectors in China’s manufacturing sector, measured sectoral carbon dioxide emissions, analyzed the trends of environmental pollution levels, and concluded that high pollution comes from capital-intensive sectors [14]. Based on 45 cities along the Yangtze River, An et al. (2022) selected five levels—atmosphere, sound, water, soil, and biology—to construct comprehensive indicators, solving the problem of single and incomplete environmental quality indicators [15]. Drawing on scholars’ ideas, the EF is a complete indicator, so this paper used its characterization of ecological pollution, which has certain advantages.

2.1.2. Ecological Footprint and Its Measurement

Determining how to choose reasonable scientific indicators to accurately measure the status and effect of environmental pollution or energy saving and emission reduction has become the focus of current research. Rees (1992) first proposed the ecological footprint concept, defining it as the total productive land area, including resource consumption and the area needed to absorb waste [16]. Since then, theory and research in the ecological footprint family have gradually become research focuses and hotspots in the field of ecological and environmental economics and have also become important indicator choices and entry points for the scientific measurement of environmental pollution problems (Hong Shunfa et al., 2020) [17]. As a quantitative indicator focused on measuring ecologically sustainable development, the ecological footprint can provide a pooled indicator model that circumvents the shortcomings of single indicators that are simple and isolated, making it possible to analyze ecological characteristics [18] (Feng Yin et al., 2017). The measurement methods used in the existing ecological footprint model can be divided into the most general measurement method, the input output measurement method, the energy-value ecological footprint measurement method, and other extended measurement methods. One is the most general measurement method, which states that the ecological footprint is the land area required to maintain resource consumption and waste absorption for a certain population. Wackernagel (1996) first proposed a method for calculating the ecological footprint and ecological carrying capacity and argued that the land areas used for fossil energy land, forest land, arable land, grassland, built-up land, and watersheds should be included when calculating the ecological footprint [19]. Secondly, the output measurement method, which applies the input output analysis method to the ecological footprint measurement model, argues that the conversion relationship between material inputs and final products obtained through the Leontief inverse matrix can overcome the structural shortcomings of the general ecological footprint model [20] (Bicknell et al., 1998). Thirdly, the energy-value ecological footprint measurement method was developed in response to the drawbacks of the output measurement method model, which fails to analyze the model one-by-one according to the land use type. Ferng (2001) calculated the ecological footprint through the modified input output method, divided the sector into a single sector corresponding to only one land type, and calculated the ecological footprint by using the land complete demand matrix [21]. Fourth, other extensions of the energy footprint calculation method, involving the calculation and comparison of the ecological footprint and ecological carrying capacity, found the same conclusion as the traditional ecological footprint, i.e., the ecological footprint is greater than the ecological carrying capacity [22]. Liu (2010) proposed the measurement of the ecological footprint with the net primary productivity yield factor method. The net primary productivity measurement method uses the yield to reflect the differences in productivity of various types of land, and the results reflect the size of the ecosystem provisioning capacity for consumption [23].

2.1.3. Energy Footprint and Its Measurement

The energy footprint (EF) refers to the ecological footprint generated by the consumption of energy resources, which is an important part of the ecological footprint and the most important factor affecting the degree of change in the ecological footprint, thus becoming a breakthrough in the search for a balance between economic development and the ecological environment (Sun et al., 2016; Fang, 2011; Lan, 2002) [24,25,26], and at the same time, the EF is used as a proxy variable for energy saving and emission reduction. One of the main advantages is that this indicator can comprehensively reflect energy consumption and the resulting environmental pollution and can account for energy saving and emission reduction at the same time. The ecological footprint model generally divides resources into biological and energy categories. Because more than half of the ecological footprint in most regions comes from the EF, the EF is considered to be the primary cause of ecological problems (Fang Kai et al., 2010) [27]. Wackernagel (1999) assumed that the global average carbon uptake rate of forested land is the same and defined the EF as the amount of carbon emissions needed to absorb the carbon emissions generated from the combustion of fossil fuels. The EF can also be defined as the total area of forest land required to absorb carbon emissions from fossil fuel combustion [28]. The EF has been regarded as an important component of ecological footprint evaluation, and special studies on the EF have received widespread attention. Based on previous studies, EF studies can be divided into seven research scales: product, individual, household, organization, prefecture, country, and region. Li et al. (2018) used the life cycle approach to study the product energy carbon footprint based on the product scale [29]; Akbari et al. (2013) used the emissions generated from personal food, clothing, and housing to represent the energy carbon footprint on the individual scale [30]; Gurvich (2021) used the horizontal and vertical synthesis method based on the household scale to analyze the explicit and implicit energy carbon footprints of households [31]; Yang et al. (2021) accounted for the energy carbon footprint within urban organizations on the organizational scale [32]; Hou et al. (2014) created a mathematical model to synthesize and analyze the spatial distribution of the energy ecological footprint on the prefecture and municipal scales [33]; and Wang et al. (2017) analyzed the carbon footprint of the country on the national scale using the input output model of energy and land [34]. Sun et al. (2019) analyzed and compared the spatial and temporal differences in carbon footprints in different regions on the regional scale [35].

2.2. Environmental Pollution and Energy Footprint

As a classical negative externality, the main source of environmental pollution is fossil energy consumption contained in the EF [36]. On the one hand, fossil energy consumption provides important support for China’s economic growth miracle. On the other hand, fossil energy consumption is also the most important source of environmental pollution problems at the current stage in China (Liu et al., 2022; Zhang, 2011) [37,38]. Along with rapid economic development, the consumption of fossil energy sources such as coal, oil, and natural gas has also increased exponentially, and in 2002, China surpassed Japan to become the second-largest oil consumer after the US (Li, 2013; Xu, 2019) [39,40], and the issue of energy consumption and energy security in China has become a focus of both China and the rest of the world. The pursuit of continuous economic growth needs to be balanced with solving the resulting environmental and climate problems, which is a great challenge for China that is still in the late stage of industrialization as a whole at this stage, under the condition that energy saving and emission reduction technologies are still immature (Du, 2021; Huang, 2018) [41,42].

In fact, there are two main ways to reduce the over-dependence on fossil energy consumption. One approach is to replace fossil energy with renewable energy. Given the limitation of fossil energy and ecological and environmental pressures, the world’s major developed countries have put forward a series of policy measures to accelerate the development of renewable energy. In terms of economic, political, and technological aspects (Rej, 2022) [43], it has greatly improved the scale of the renewable energy sector and the technology level. Additionally, it represents a new economic growth point out of the weak state of the economy and a breakthrough way to solve ecological environmental problems. However, in the short term, the proportion of renewable energy consumption is still small, and it is more difficult to realize the substitution of fossil energy consumption on the whole to achieve the goals of energy saving, emission reduction, and green development. Another approach is to optimize the sectoral structure, increase the proportion of clean sectors, reduce the proportion of polluting sectors, improve the energy efficiency, and reduce the pollution per unit of energy consumption. The adjustment of the sectoral structure is an important strategy to achieve energy conservation, emission reduction, and sustainable economic development (Shi, 1999; Wang et al., 2020) [44,45]. Additionally, it is also the most effective and fundamental policy choice to realize ecological environmental protection (Lin, 2013; Han et al., 2015) [46,47].

2.3. Study on the Relationship between Environmental Pollution and Industrial Structure

Research on the relationship between environmental pollution and the industrial structure by scholars at home and abroad started with the emergence and development of industrial economics. The research mainly focused on two aspects based on the different research methods and paradigms.

Firstly, in discussions on the causal relationship between the industrial structure and environmental pollution, the research results focus on an empirical regression analysis of the impact of changes in the industrial structure on environmental pollution (Gong, 2020) [48]. However, in combing this literature, it was also found that different scholars have obtained different research conclusions due to different selections of environmental pollution and energy saving and emission reduction indicators. Shao et al. (2022) found that industrial structure upgrading helps to reduce the carbon emission performance [49]. Zhu (2021) used a dynamic panel model and found that the relationship between industrial agglomeration and environmental pollution is an inverted U-shaped relationship, and that industrial agglomeration currently helps to reduce pollution emissions. Regarding the mediating effect, industrial agglomeration reduces the total amount of pollutants through technological progress and increases pollutant emissions through scale expansion and the inhibition of structural transformation [50]. Dong (2020) used a geographically and temporally weighted regression model and found that the development of industrial agglomeration will, to a certain extent, lead to negative externalities and increase the severity of environmental pollution. At both the national and provincial levels, it was found that industrial agglomeration increased pollution agglomeration [51]. Hu (2020) used PSM-DID to explore the impact of industrial agglomeration policy on regional environmental performance and found that the concentrated emission of pollution formed by industrial agglomeration exacerbates the degree of environmental pollution [52].

Secondly, regarding the use of input output tables to analyze environmental issues, Takase et al. (2005) calculated sustainable household consumption based on discarded input output tables. They found that income rebound should be included when assessing the environmental carrying capacity for different consumption structures [53]. Guo et al. (2008) chose China’s 42 sectoral input output tables to calculate energy consumption coefficients and select priority sectors based on the perspectives of resource and environmental effects [54]. Liao et al. (2011) compiled an input output table of “energy saving and emission reduction”. They compiled an energy input output model for the study of energy saving and an environmental input output model for the study of emission reduction [55]. Chen et al. (2017) accounted for China’s carbon emissions based on the perspectives of sectoral responsible-based carbon emission and consumption-based carbon emission, compared the two types of carbon emissions, and derived carbon emission reduction policies [56]. Huang et al. (2022) compared and analyzed the gap between China’s economy and that of the rest of the world based on the world’s input output table and concluded that interconnections among energy, digital information, and trade should be strengthened [57]. Therefore, decomposing the EF using consumption-based perspective and responsibility-based perspective is of great practical importance.

3. Research Framework and Data Processing

3.1. Research Framework

Constructing scientific and reasonable EF measurement indicators and clarifying the sector EF emission responsibilities are the basis for the effective decomposition of the EF. This part discusses how to construct EF assessment indicators and an accounting framework in a scientific and reasonable way based on a sectoral perspective.

3.1.1. Accounting Principles of Sectoral RBEF and CBEF

The construction of scientific and reasonable environmental responsibility indicators must follow certain principles associated with environmental responsibility allocation, and the accounting principle methods of Peters (2008), Zhang (2013), and Zhang (2014) and other scholars have been used to construct sectoral EF measurement indicators based on the principles of producer responsibility and consumer responsibility [58,59,60].

The principle of allocating the sectoral CBEF is the producer responsibility principle. According to this principle, producers should be held responsible for the EF of their production processes. The economic logic of the producer responsibility principle is the idea of the Polluter Pays Principle (PPP), which requires polluters to pay for the environmental pollution and damage caused by their production process, reflecting the “territorial principle”. Existing literature studies also refer to the environmental pollution generated by economic agents in their production due to fossil energy consumption as energy consumption environmental pollution. Thus, the EF generated in the production process, for which each sector is responsible, is called the consumption-based energy footprint (CBEF). The CBEF was further measured by drawing on the idea of “electric heat sharing” proposed by Zhang (2009) and Tu (2012) [61,62]. This paper defines the CBEF as follows: the sum of the energy footprints of the intermediate outputs that a producer needs to provide to satisfy the need to produce the full range of outputs in all sectors.

The emissions allocation principle for the sectoral RBEF is the consumer responsibility principle. The principle of consumer responsibility means that the sector as a consumer should be responsible for the implied environmental pollution from its intermediate input products from energy, goods, and services, according to which pollution is associated with the final goods and services [63]. Based on this idea, sectors must also be responsible for the EF generated in the production process of the intermediate products they use. Drawing on the ideas and methods of Fan et al. (2010) to measure the carbon emission responsibility, the sectoral RBEF can be measured based on the principle of consumer responsibility in order to analyze the inter-sector EF responsibility relationship and clarify the inter-sector EF sources [64]. This principle attributes responsibility for energy footprint emissions to the final product or service user and holds the consumer responsible for the implied energy footprint. This paper defines the sectoral RBEF: if an industry purchases the output produced by another industry, then the sector should be held responsible for the corresponding implied carbon.

Based on the producer and consumer responsibility principles, the sectoral CBEF and the sectoral RBEF can be calculated separately. Questions that deserve further investigation are whether there are differences between the two and what characteristics the differences show.

3.1.2. Accounting Methodology and Comparison of Sectoral RBEF and CBEF

• Non-competitive input output table

This paper uses non-competitive input output tables as the basis for EF accounting and decomposition studies for the following reasons: at first, the fossil energy consumption provided by the National Bureau of Statistics of China only involved domestic data, which needed to be matched with non-competitive input output tables; secondly, imports as intermediate and final goods have different impacts on the economic system, and the competitive input output method cannot be used directly for analysis and measurement [65]. The basic structure of the non-competitive input output table is shown in

Table 1. The basic structure of a non-competitive input output method.

		Intermediate Use			Final Use		Total Domestic Output or Imports
		1, 2, …, n	Consumption	Investment	Export	Total
Intermediate inputs for domestic products	1, 2, …, n	$x_{ij}^d$	$y_{ic}^d$	$y_{in}^d$	$y_{ie}^d$	$Y_i^d$	$X_i$
Imported intermediate inputs	1, 2, …, n	$x_{ij}^m$	$y_{ic}^m$	$y_{in}^d$		$Y_i^m$	$M$
Added Value		$V_i$
Total Inputs		$X_i$

.

It can be seen that the following quantitative relationships exist in the horizontal direction for the data related to non-competitive input output tables:

$${\displaystyle \sum _{j=1}^n{x}_{ij}^d}+Y_i^d=X_i, i=1,2,\cdots ,n$$

(1)

$${\displaystyle \sum _{j=1}^n{x}_{ij}^m}+Y_i^m=M_i, i=1,2,\cdots ,n$$

(2)

where $x_{ij}^d$, $x_{ij}^m$ denote the products of domestic and foreign sector ($j$) consumed in the production of sector $i$ and the intermediate inputs include both domestic and foreign components; $Y_i^d$, $Y_i^m$ denote the domestic and foreign end-use quantities of the sector; and $X_i$, $M_i$ denote the total domestic output and import quantity statuses of sector $i$.

We write Equations (1) and (2) in matrix form. In turn, this leads to the following:

$$X={(I-A^d)}^{-1}{Y}^d$$

(3)

$$M=A^m{(I-A^d)}^{-1}{Y}^d+Y^m$$

(4)

where $A^d$, $A^m$ indicates that the sectoral production processes of other domestic and foreign products directly consume the matrix. $X$, $M$ indicate the total domestic and foreign end-use statuses for all sectors. $L^d={(I-A^d)}^{-1}$ represents the inverse Leontief matrix of a country’s product production. $L_{ij}^d$ indicates the total consumption of domestic products of sector $i$ per unit of domestic production for sector $j$.

2.

Responsible-based energy footprint measurement methodology

The Lyontief inverse matrix of a country’s products reveals the intrinsic mechanism of interdependence among sectors, and the matrix form can be expressed as:

$$L^d={(I-A^d)}^{-1}=\left[\begin{array}{ccc}{L}_{11}^d& \cdots & L_{1n}^d\\ \cdots & L_{ij}^d& \cdots \\ L_{n1}^d& \cdots & L_{nn}^d\end{array}\right]$$

(5)

The sum of the column vectors ${\displaystyle \sum _{i=1}^n{L}_{ij}^d}$ in column $j$ is the complete consumption coefficient of sector $j$, which represents the sum of all sector inputs consumed by sector $j$ to produce one unit of the final domestic demand. Bringing Equation (5) into Equation (3) yields:

$$X={(I-A^d)}^{-1}{Y}^d=\left[\begin{array}{ccc}{L}_{11}^d{Y}_1^d& \cdots & L_{1n}^d{Y}_n^d\\ \cdots & L_{ij}^d{Y}_j^d& \cdots \\ L_{n1}^d{Y}_1^d& \cdots & L_{nn}^d{Y}_n^d\end{array}\right]$$

(6)

Vertically, the matrix element $L_{ij}^d{Y}_j^d$ on the right-hand side of Equation (6) represents the demand for intermediate goods from sector $i$ by the production process of sector $j$; the sum of the vectors ${\displaystyle \sum _{i=1}^n{L}_{ij}^d}{Y}_j^d$ in column $j$ represents the sum of all sectors’ intermediate inputs required by the production process of sector $j$. Considering the CBEF as the implied EF of total output, it is obtained that:

$$C{E}^e=eX$$

(7)

where $C{E}^e$ denotes the total CBEF generated in the sectoral production process; $e$ denotes the direct EF coefficient of the sector; and $e_i=C{E}_i^e/X_i$ denotes the CBEF per unit of the output of sector $i$. Bringing Equation (6) into Equation (7) yields:

$$C{E}^r=e{(I-A^d)}^{-1}{Y}^d=e{L}^d{Y}^d=\left[\begin{array}{ccc}{e}_1{L}_{11}^d{Y}_1^d& \cdots & e_1{L}_{1n}^d{Y}_n^d\\ \cdots & e_i{L}_{ij}^d{Y}_j^d& \cdots \\ e_n{L}_{n1}^d{Y}_1^d& \cdots & e_n{L}_{nn}^d{Y}_n^d\end{array}\right]$$

(8)

where $C{E}^r$ is the responsible-based energy footprint, $e_i{L}_{ij}^d$ denotes the EF of sector $i$ resulting from the production of one unit of the final domestic demand for sector $j$. It can be further defined as:

$$C{E}_j^r={\displaystyle \sum _{i=1}^n{e}_i{L}_{ij}^d{Y}_j^d}$$

(9)

Equation (9) represents the implied EF of sector $j$, which is defined as the responsible-based energy footprint of sector $j$ according to the consumer responsibility principle, for which sector $j$ is responsible.

3.

Accounting methods for the consumption-based energy footprint

Using the “Electricity and Heat Sharing Method”, the EF of thermal power generation and heat supply is apportioned to the electricity and heat-using sectors to account for the sectoral EF. According to the producer responsibility principle, sector $i$ is responsible for the energy footprint of the quantity $e_i{L}_{ij}^d{Y}_j^d$, which leads to the following:

$$C{E}_i^e={\displaystyle \sum _{j=1}^n{e}_i{L}_{ij}^d}{Y}_j^d$$

(10)

In Equation (10), $L_{ij}^d{Y}_j^d$ denotes the output provided by the sector to satisfy the production of all sectors $i$; the sum of the row vectors ${\displaystyle \sum _{j=1}^n{L}_{ij}^d}{Y}_j^d$ in row $i$ denotes the intermediate output that sector $i$ needs to provide in order to satisfy the production of all sectors.$e_i{L}_{ij}^d$ denotes the energy footprint of sector $i$ that is needed to satisfy the production of one unit of the final demand of sector $j$.

4.

Comparison between sectoral CBEF and the RBEF

The responsible-based energy footprint can be regarded as the redistribution of the CBEF, or it can be regarded as the result of the sectoral CBEF formed after taking into account the EF transfer effect formed by sector linkages, which are equal in terms of the total amount. Using Equations (10) and (9), respectively, the CBEF and the RBEF of all sectors are summed up, which leads to the following:

$${\displaystyle \sum _{i=1}^{n}C}{E}_i^e={\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{e}_i}{L}_{ij}^d}{Y}_j^d, {\displaystyle \sum _{j=1}^{n}C}{E}_j^r={\displaystyle \sum _{j=1}^n{\displaystyle \sum _{i=1}^n{e}_i}{L}_{ij}^d}{Y}_j^d$$

(11)

From Equation (11), it is clear that the sum of the sectors’ CBEF is equal to the sum of the sectors’ RBEF. However, because the sectoral producer and consumer responsibilities are determined based on different economic theories and ideas, the CBEF and the RBEF of the sectors are also necessarily asymmetric and inconsistent (Lenzen, Murray, 2010) [66]. If the sectoral CBEF is larger than its RBEF, the sector plays the role of the producer (consumer); if the difference is not significant, the sector plays both producer and consumer roles.

3.1.3. Decomposition Method of Energy Footprint

In this paper, based on decomposing the EF of each sector into the CBEF and the RBEF based on the above method, we further decompose the EF of the sector into the direct EF and indirect EF to corroborate the two with the two decomposition methods. Based on the principles and ideas of producer responsibility and consumer responsibility, it is known that the EF of the sector comes from two aspects: one is generated by the energy consumption of the sector for the production of the final products and services; the other is generated by the energy consumption of intermediate products used in the production process of the sector. The two are defined as the sectoral direct EF and the indirect EF and are used to realize the decomposition of the EF.

The CBEF is equal to the RBEF, and here only the CBEF is studied as an example, and by considering it as the EF of the total output, the following equation exists:

$$C{E}^c=C{E}^e=eX$$

(12)

where $C{E}^e$ is the sectoral CBEF, $e$ denotes the direct EF coefficient, and $e_i=C{E}_i^e/X_i$ is the direct EF coefficient of sector $i$, which indicates the CBEF due to the production of one unit of the total output, reflecting the direct EF intensity of sector $i$ in the production process.

According to Equation (3), the total output can be considered as a function of the end-use, which is brought into Equation (12) to obtain the following equation:

$$C{E}^c=e{(I-A^d)}^{-1}{Y}^d=e{L}^d{Y}^d=e^c{Y}^d$$

(13)

where $C{E}^c$ is the complete sectoral EF, and $C{E}_j^c$ denotes the complete EF of sector $j$, which can be considered as the implied EF due to the production of the final demand $Y_j^d$, and $L_{ij}^d$ denotes the total consumption of the sectoral product in $i$ by sector $j$ due to the production of one unit of the final demand. $e^c=e{(I-A^d)}^{-1}=e{L}^d$ is the total EF factor. $e_j^c={\displaystyle \sum _{i=1}^n{e}_i{L}_{ij}^d}$ denotes the full EF factor of sector $i$, which is the complete EF produced by sector $j$ to produce one unit of the final demand.

A Taylor expansion of the inverse Lyontief matrix $L^d$ yields:

$$L^d={(I-A^d)}^{-1}=I+A^d+{(A^d)}^2+{(A^d)}^3+{(A^d)}^4+\cdots$$

(14)

By bringing Equation (14) into Equation (3), we obtain the following equation:

$$X=Y^d+A^d{Y}^d+{(A^d)}^2{Y}^d+\cdots {(A^d)}^t{Y}^d+\cdots$$

(15)

Equation (15) implies that the total output can be decomposed into the sum of polynomials consisting of a matrix of the final demand and direct consumption. The first term to the right of the equal sign represents the total output effect induced by the final demand $Y^d$ and is called the direct effect. The remaining terms are called indirect effects and represent the outputs that must be produced to produce the final demand $Y^d$, where $A^d{Y}^d$ denotes the intermediate output that must be produced to produce the final demand and also denotes the indirect effect of the first round; ${(A^d)}^t{Y}^d=(A^d)[{(A^d)}^{t-1}{Y}^d]$ denotes the output that must be produced to produce ${(A^d)}^{t-1}{Y}^d$ and is the indirect effect of round $t$; and the total indirect effect is the sum of the first round, the second round, and the subsequent round $t$, denoted as $A^d{Y}^d+{(A^d)}^2{Y}^d+\cdots {(A^d)}^t{Y}^d+\cdots$.

We bring Equation (15) into Equation (13). The sectoral complete EF can be decomposed as shown in Equation (16).

$$C{E}^c=[e{Y}^d]+[eA{Y}^d+e{A}^2{Y}^d+\cdots e{A}^{t-1}{Y}^d+\cdots ]$$

(16)

where the first term to the right of the equal sign is the direct EF, which represents the EF caused by the final demand $Y^d$, and the second term is the indirect EF, which represents the EF caused by the outputs (intermediate inputs) necessary to produce the final demand $Y^d$.

3.2. Data Processing

This paper mainly uses input output table data, energy consumption data by sector, and EF measurement data to analyze the characteristics of EF changes in each sector and their influencing factors. The input output table data were selected from the national input output table data from 2007, 2012, and 2017. Before the specific analysis, the raw data needed to be processed accordingly.

At first, the sectors in the input output data tables were merged. In order to make them correspond to the sectoral energy consumption data in the China Energy Statistics Yearbook, the sectors in the input output tables for 2007, 2012, and 2017 were combined into 28, as shown in

Table 2. Sector classification and codes.

Sector Classification	Code	Sector Classification	Code
Agriculture	1	Metal products sector	15
Coal mining and washing sector	2	Machinery sector	16
Oil and gas extraction sector	3	Transportation equipment manufacturing	17
Metal Mining	4	Electrical machinery and equipment manufacturing	18
Non-metallic mineral and other mineral mining sector	5	Electronic and communication equipment manufacturing	19
Food manufacturing and tobacco processing sector	6	Instruments and cultural and office supplies machinery manufacturing	20
Textile sector	7	Other Manufacturing	21
Clothing leather and other fiber products manufacturing	8	Electricity, heat production and supply sector	22
Wood processing and furniture manufacturing	9	Gas production and supply sector	23
Paper printing and stationery manufacturing	10	Water production and supply sector	24
Petroleum processing and coking sector	11	Construction	25
Chemical sector	12	Transportation, storage, post and telecommunications	26
Non-metallic mineral products sector	13	Wholesale and retail accommodation and catering	27
Metal smelting and rolling processing sector	14	Other sectors	28

. The original data were obtained from the input output table published by the National Bureau of Statistics and the China Input Output Institute.

Secondly, based on the input output table data of the base period, the price index was used to deflate the input output table data afterwards to eliminate the influence of prices. The specific process was as follows: firstly, according to the sample selection used to determine 2007 as the base year, the 2007 input output table data were used as the base data; secondly, the price index of each sector was selected, and the price indices of each sector for each year during 2007–2017 were accumulated and multiplied to calculate the price indices for 2012 and 2017. For the selection of the price indices of specific sectors, the agricultural production price index was selected to deflate the agricultural data (1); the sub-sectoral ex-factory price index was selected to deflate the coal mining and washing sector (2) and the water production and supply sector data (24); the construction and installation price index, transportation and communication price index, commodity retail price index, and consumer price index were selected to deflate the construction sector data (25) and data from other sectors (28) for price deflation. All price indices were obtained from the China Statistical Yearbook of previous years.

Thirdly, the input output table was converted from a comparable input output table to a non-competitive comparable input output table to eliminate the impact of imports on the analysis. Based on the competitive input output table data published by the National Bureau of Statistics of China, the non-competitive comparable input output table was prepared according to the “proportional equivalence method”, assuming that each sector was using equal proportions of domestic and foreign intermediate and final products.

Fourth, the EF data were calculated based on the energy consumption data from each sector. Based on the availability of data, four major types of energy, namely, coal, oil, natural gas, and electricity, were selected as samples, among which coal included coal and coke and oil included crude oil, fuel oil, gasoline, kerosene, and diesel, and these were further converted according to the equivalent standard coal coefficients of each subdivision of energy to finally determine the consumption levels of the four types of energy. The original data on energy consumption were obtained from the relevant years from the China Energy Statistical Yearbook. The original data on energy consumption were obtained from the China Energy Statistical Yearbook for the relevant years. The specific measurement method for the energy footprint was borrowed from the idea of Zhang and Hao (2020) [67] and is not repeated here.

4. Results

This part calculates China’s sectoral CBEF and RBEF, respectively, based on the 2017 input output table and analyzes the results of measuring the EF rate and the trend of change for 28 sectors in 2007, 2012, and 2017.

4.1. Sectoral Energy Footprint Accounting

We take the input output table data from 2017 as an example. The CBEF and RBEF of 28 sectors were measured using the above-given equations. The results are shown in

Table 3. Consumption-based energy footprint and responsible-based energy footprint of Chinese sectors in 2017.

Sector Code	Consumption-Based Energy Footprint		Responsible-Based Energy Footprint
	Value 1	Rate 2	Value 1	Rate 2
1	1240.489	0.739	3060.540	1.824
2	8121.564	4.841	2326.814	1.387
3	343.258	0.205	497.147	0.296
4	187.520	0.112	1918.468	1.143
5	384.343	0.229	1380.653	0.823
6	629.264	0.375	3046.887	1.816
7	1013.843	0.604	1886.311	1.124
8	30.510	0.018	1277.942	0.762
9	64.954	0.039	1054.920	0.629
10	1238.147	0.738	2256.027	1.345
11	38,724.304	23.081	4749.752	2.831
12	11,421.618	6.808	20,250.470	12.070
13	10,233.253	6.099	10,176.159	6.065
14	24,470.831	14.585	17,175.162	10.237
15	162.489	0.097	3953.223	2.356
16	226.859	0.135	3245.716	1.935
17	117.857	0.070	3286.871	1.959
18	51.798	0.031	2105.979	1.255
19	47.984	0.029	1935.673	1.154
20	6.961	0.004	307.785	0.183
21	147.373	0.088	374.812	0.223
22	59,531.934	35.482	22,024.459	13.127
23	111.215	0.066	381.791	0.228
24	13.046	0.008	3162.214	1.885
25	44.223	0.026	15,351.269	9.150
26	7010.139	4.178	12,933.311	7.709
27	882.566	0.526	3200.944	1.908
28	1320.777	0.787	24,457.820	14.577
Total	167,779.119	100	167,779.119	100

Note: ¹ The unit is million hectares; ² the unit is %. The units in the table below are the same.

and Figure 1. In 2017, the EF of 28 sectors in China was 167,779.119 million hectares. The CBEF measured based on the producer responsibility principle was the same as the RBEF measured based on the consumer responsibility principle. However, specifically among the 28 sub-sectors, there were significant asymmetries and inconsistencies in each sectoral CBEF and RBEF.

Specifically, among the 28 sectors, the three sectors with enormous CBEF values are the electricity and heat production and supply sector (22), the petroleum processing and coking sector (11), and the metal smelting and rolling processing sector (14), and the CBEF of these three sectors are 59,531.934 million hectares, 38,724.304 million hectares, and 244,708.31 million hectares, respectively. These three sectors have the most severe energy consumption and pollution problems in the production of products and are specific polluting sectors with high levels of energy consumption and high pollution. The three sectors with the smallest CBEF are instrumentation and cultural and office supplies machinery manufacturing (20), water production and supply (24), and garment leather and other fiber product manufacturing (8). The CBEF of these three sectors are 69.61 million hectares, 130.46 million hectares, and 305.1 million hectares, respectively. The ratio of the total CBEF of the three sectors is only 0.03%, and these three sectors have the lightest energy consumption and pollution problems during product production. They are typically clean sectors with low energy consumption and pollution levels. In addition, the construction sector (25), electronic and communication equipment manufacturing sector (19), electrical machinery and equipment manufacturing sector (18), wood processing and furniture manufacturing sector (9), transportation equipment manufacturing sector (17), and metal products sector (15) are also specific sectors with high levels of cleanliness. Their CBEF rates are below 0.1%. It can be seen that the CBEF of the 28 sectors has two typical characteristics: Firstly, the distribution of the sectoral CBEF has a high degree of sector concentration. The three sectors with the largest consumption-based energy footprints accounted for nearly three-quarters of the total CBEF. Secondly, the rate of the sector CBEF has a large sectoral difference. For example, the production and supply of electricity and heat (22) energy footprint is more than 8550 times the energy footprint of the instrumentation and cultural and office supplies machinery manufacturing sector (20).

Among the 28 sectors, the three sectors with the most significant responsible-based energy footprints are other sectors (28), the electricity and heat production and supply sector (22), and the chemical sector (12), and the responsible energy footprints of these three sectors are 244,578.2 million hectares, 220.2445 million hectares, and 202.504 million hectares, respectively, which together account for a total CBEF ratio of 39.774%. These three sectors have the most severe energy consumption and pollution levels generated by intermediate products used in product production and are implicitly high energy-consuming and polluting sectors. The three sectors with the smallest responsible-based energy footprints are instrumentation and cultural and office supplies machinery manufacturing (20), other manufacturing (21), and gas production and supply (23), and the responsible-based energy footprints of these three sectors are 3077.85 million hectares, 3748.12 million hectares, and 3817.91 million hectares, respectively, with a combined ratio of 0.634% for the total CBEF of these three sectors. The three sectors have minor pollution problems caused by intermediate-use products and are implicitly clean sectors with low energy consumption and emission levels. The distribution of RBEF sectors is more dispersed than the energy consumption footprint, and the sector variability is relatively small.

From a comprehensive perspective, among the 28 sectors, the electricity and heat production and supply sector (22), metal smelting and rolling processing sector (14), chemical sector (12), and non-metallic mineral products sector (13) are typical “double-high” sectors with a high CBEF and high RBEF. During the “14th Five-Year Plan” period, and even in the longer term, China’s energy saving and emission reduction goals are focused on the sector. Instruments and cultural and office supplies machinery manufacturing (20), other manufacturing sectors (21), and gas production and supply sector (23) are typical “double low” sectors. The CBEF and RBEF are low, so China should further strengthen and optimize the cleanliness of the key sectors. In addition, petroleum processing and coking (11) is a typical high CBEF sector. Other sectors (28), construction (25), and transportation, storage, post, and telecommunications (26) are typical high RBEF sectors. Achieving green production and efficient use of these sectors is also a focus of China’s energy saving and emission reduction strategy.

4.2. Trends in the Sector’s Energy Footprint

The results and trends of the measured EF rates of 28 sectors in 2007, 2012, and 2017 are shown in

Table 4. Energy consumption-based (responsibility-based) energy footprint rates for 28 sectors, 2007–2017.

Year	2007	2012	2017	Amount of Change 1	2007	2012	2017	Amount of Change 1
Energy Footprint	125,136.055	186,095.122	167,779.119		125,136.055	186,095.122	167,779.119
Sector Code	Consumption-Based Energy Footprint Rate				Responsible-Based Energy Footprint Rate
1	1.125	0.641	0.739	−0.386	1.780	2.143	1.824	0.044
2	4.812	7.547	4.841	0.028	1.632	2.128	1.387	−0.245
3	0.559	0.295	0.205	−0.355	1.407	0.775	0.296	−1.111
4	0.109	0.168	0.112	0.003	1.543	1.379	1.143	−0.399
5	0.195	0.309	0.229	0.034	0.607	0.612	0.823	0.216
6	0.482	0.730	0.375	−0.107	1.885	1.692	1.816	−0.069
7	0.694	0.533	0.604	−0.090	1.270	1.180	1.124	−0.145
8	0.044	0.041	0.018	−0.025	1.225	0.614	0.762	−0.464
9	0.088	0.108	0.039	−0.049	0.734	0.614	0.629	−0.105
10	0.872	0.788	0.738	−0.134	1.160	1.447	1.345	0.184
11	19.856	18.647	23.081	3.225	2.886	2.978	2.831	−0.055
12	5.167	6.480	6.808	1.641	10.720	11.671	12.070	1.350
13	5.286	6.411	6.099	0.814	4.521	5.633	6.065	1.544
14	16.381	17.192	14.585	−1.795	13.002	13.470	10.237	−2.765
15	0.134	0.159	0.097	−0.037	3.121	3.190	2.356	−0.765
16	0.282	0.202	0.135	−0.146	4.305	3.671	1.935	−2.371
17	0.163	0.118	0.070	−0.093	2.086	2.291	1.959	−0.127
18	0.051	0.096	0.031	−0.020	2.993	3.115	1.255	−1.738
19	0.038	0.035	0.029	−0.010	1.550	1.301	1.154	−0.396
20	0.005	0.007	0.004	−0.001	0.223	0.161	0.183	−0.040
21	0.093	0.159	0.088	−0.005	0.611	0.284	0.223	−0.387
22	38.398	34.304	35.482	−2.916	17.291	15.714	13.127	−4.164
23	0.304	0.119	0.066	−0.237	0.128	0.186	0.228	0.100
24	0.014	0.012	0.008	−0.006	0.316	0.217	1.885	1.569
25	0.015	0.027	0.026	0.011	9.193	10.325	9.150	−0.043
26	4.003	3.621	4.178	0.175	6.817	6.189	7.709	0.892
27	0.405	0.447	0.526	0.121	2.024	1.389	1.908	−0.116
28	0.426	0.803	0.787	0.361	4.970	5.629	14.577	9.608
Total	100	100	100	-	100	100	100	-

Note: amount of change ¹ is the change in the increase or decrease in the energy footprint rate in 2017 compared to 2007.

. The overall energy footprint shows a trend of rising and then falling, with the sector EF reaching 186,095,122,000 hectares in 2012 and then falling to 167,779,119,000 hectares in 2017, which indicates that since the 18th National Congress, with the continuous promotion of China’s ecological civilization construction, supply-side structural reform, and innovation-driven strategy, energy conservation and emission reduction have occurred at the sector level. However, the EF was still higher than 125,136,055,000 hectares in 2007, and the potential energy conservation, emission reduction, and EF reduction in China’s sector still need to be further explored. In terms of the CBEF, during 2007–2017, the CBEF rate increased in 10 sectors. The three sectors with the most significant increases were petroleum processing and coking (11), the chemical sector (12), and non-metallic mineral products (13), which increased by 3.225, 1.641, and 0.814, respectively. The CBEF rate decreased in 18 sectors. The three sectors with the most significant declines were the production and supply of electricity and heat (22), the metal smelting and rolling processing sector (14), and agriculture (1), which were, respectively, down by 2.916, 1.795, and 0.386. In terms of the RBEF, its rate increased in nine sectors during 2007–2017. The three sectors with the most significant increases were other sectors (28), water production and supply (24), and non-metallic mineral products (13), which increased by 9.608, 1.569, and 1.544, respectively. The RBEF rate fell in 19 sectors. The three sectors with the most significant declines were the production and supply of electricity and heat (22), the metal smelting and rolling processing sector (14), and the machinery sector (16), which were down by 4.164, 2.765, and 2.371, respectively.

It can be seen that, on the one hand, the overall effect of energy conservation and emission reduction in China’s sector has been remarkable, thanks to China’s coal sector supply-side structural reform, the energy revolution strategy, the implementation of the Made in China 2025 plan, coal-to-gas projects, and the implementation of the moderate development of renewable energy to replace fossil energy strategy. The CBEF and RBEF of the electricity and heat production and supply sector and the metal smelting and rolling processing sector have shown significant downward trends, and the “cleanliness” of these sectors has been increasing. On the other hand, some sectors, such as the petroleum processing and coking sector, the chemical sector, other sectors, and the water production and supply sector, have also shown an increasing trend in terms of the “pollution degree”, probably due to the increases in the total output and energy consumption caused by the increased demands of the sector, which has increased the pollution emissions of other sectors. For example, the total output of the petroleum processing and coking sector was CNY 2,107,456 billion in 2007, while the total output of the sector was CNY 3,765,478 billion in 2017, but at the same time, the implementation of the energy revolution strategy and the clean and efficient use of energy have led to a slight decrease in the EF of its intermediate inputs.

5. Input Output Decomposition of the Energy Footprint of Chinese Sectors

This section statically decomposes China’s sectoral EF from two perspectives using the 2017 input output table as an example. Firstly, the sectoral EF is decomposed into the CBEF and RBEF using the input output method; secondly, the sectoral EF is decomposed into direct and indirect energy footprints using the input output method.

5.1. Sector Consumption-Based and Responsible-Based Energy Footprint Breakdown

As for Equation (11), the CBEF and RBEF can be further subdivided by sector, and the CBEF(rows 2–29) and RBEF(columns 2–29) of 28 sectors decomposed in China in 2017 are shown in

Table 5. Breakdown of China’s consumption-based (responsible-based) energy footprint sector by 28 sectors in 2017.

Sector	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	Sum
1	214.13	0.29	0	0.04	0.03	660.66	102.49	8.39	38.6	26.32	0.01	74.89	0.3	0.05	0.12	0.1	0.05	0.07	0	0	3.2	0.09	0	0	27.65	0.68	54.46	27.86	1240.49
2	21.65	1204.23	3.1	27.68	39.1	100.27	53.71	16.01	10.39	119.9	764.24	811.32	1149.81	847.47	15.68	19.87	2.73	6.15	0.01	0.6	4.8	2748.85	51.42	0	23.33	3.15	3.23	72.85	8121.56
3	0	0.08	1.21	0.06	0.05	0	0	0	0	0.1	265.76	26.38	1.91	1.85	0.9	1.38	0.84	0.1	0	0	0	8.59	33.92	0	0	0	0	0.13	343.26
4	0	0	0	12.12	0	0	0	0	0	0	0.17	4.86	1.53	165.01	3.66	0	0	0	0	0	0.01	0	0	0	0	0	0	0.16	187.52
5	0.01	4.07	63.9	1.47	6.89	2.79	0.05	0.03	0.05	2.77	0.08	37.64	182.6	1.69	0.48	0.61	0.63	1.92	0.28	0.02	0.11	0.89	0	0	74.72	0.28	0.18	0.19	384.34
6	93.01	0.38	0.1	0.73	0.49	262.68	0.88	12.74	1.19	1.65	1.76	32.29	1.6	6.81	2.26	5.78	2.5	3.25	5.14	0.53	0.48	3.68	0.27	0.26	4.74	10.12	115.21	58.71	629.26
7	0.13	0.31	0.07	0.15	0.77	2.05	450.95	421.89	10.78	32.81	0.16	19.97	0.78	0.53	2.69	3.19	3.18	2.23	1.28	0.57	9.3	0.02	0.01	0.02	0.25	4.27	6.6	38.9	1013.84
8	0.04	0.3	0.04	0.06	0.07	0.82	0.44	11.16	0.97	0.51	0.11	0.84	0.79	0.18	0.19	0.67	1.19	0.36	0.29	0.07	0.13	0.13	0.04	0.06	0.65	0.61	0.91	8.89	30.51
9	0.11	1.28	0.07	0.12	0.03	0.36	0.15	0.17	31.45	2.85	0.04	0.74	0.69	0.08	1.53	0.5	2.55	0.33	0.24	0.06	0.34	0.05	0	0	18.37	0.42	0.21	2.18	64.95
10	3.7	0.11	0.05	0.72	1.05	64.45	5.08	9.18	10.08	399.56	0.87	41.68	35.58	1.63	6.68	12.44	6.37	17.56	15.71	1.45	2.56	1.32	0.02	0.04	15.61	71.55	32.89	480.21	1238.15
11	605.76	88.23	65.95	697.85	571.84	165.54	66.39	78.82	127.82	114.38	2847.88	7599.47	2285.65	4726.21	333.66	281.85	177.07	108.21	82.06	30.6	72.21	1021.88	123.12	2.57	3409.35	8862.69	423.33	3753.94	38,724.3
12	706.49	30.55	23.78	47.87	63.63	181.79	332.13	219.49	166.53	390.43	89.41	4721.24	320.52	118.85	129.45	222.87	302.43	381.71	243.06	22.01	58.49	1.8	0.35	18.64	817.21	85.18	53.06	1672.64	11,421.62
13	7.74	8.5	0.88	7.53	27.2	97.76	5.98	2.84	29.47	16.03	44.02	125.54	1988.26	236.12	98.33	106.4	170.74	268.15	225.38	38.05	6.26	22.04	0.06	0.17	6618.89	5.39	8.13	67.38	10,233.25
14	61.98	15.95	4.52	12.67	14.92	6.62	2.41	5.71	33.42	225.92	22.8	140.93	261	4828.53	1090.41	831.61	1460.18	322.5	43.62	63.39	2.76	0.71	0.46	2767.95	5.33	0.39	8.74	12,235.42	24,470.83
15	0.35	2.63	0.03	1.18	1.08	1.97	0.44	0.58	2.66	1.42	0.1	4.1	7.44	0.96	26.68	15.96	9.11	10.03	5.45	1.2	1.01	0.09	0.01	0.27	49.89	0.92	0.27	16.67	162.49
16	4.73	3.7	1.97	2.61	3.05	1.48	1.29	0.72	1.16	2.06	1.18	6.04	8.52	8.77	7.24	93.15	22.39	12.51	11.53	1.8	1.02	1.12	0.04	0.09	12.91	4.38	0.39	11.02	226.86
17	0.81	0.09	0.01	0.21	0.25	0.09	0.04	0.06	0.05	0.06	0.01	0.2	0.58	0.07	0.16	5.56	73.85	0.41	0.09	0.1	0.58	0.01	0	0	0.52	21.03	0.15	12.88	117.86
18	0.03	0.12	0.02	0.08	0.11	0.1	0.04	0.02	0.03	0.22	0.02	0.27	0.19	0.11	0.14	5.67	4.04	12.89	6.12	0.63	0.22	4.33	0.01	0.01	9.98	1.41	1.25	3.75	51.8
19	0.02	0.02	0	0	0	0.05	0.01	0.01	0.01	0.16	0	0.13	0.04	0.04	0.04	2.74	1.13	2.65	33.18	0.01	0.09	0.07	0	0	0.47	2.56	0.47	4.07	47.98
20	0	0.01	0.1	0.01	0.03	0.03	0	0.01	0	0.01	0.05	0.06	0.05	0.03	0.01	0.55	0.52	0.36	0.46	1.09	0.03	1.12	0.01	0	0.15	0.07	0	2.2	6.96
21	0.2	0.5	0.24	0.28	0.22	0.94	0.3	1.48	0.27	14.71	0.41	5.19	3.83	64.4	7.87	1.86	1.44	0.94	1.27	0.15	10.18	3.21	0.11	0.05	3.95	2.26	0.38	20.73	147.37
22	1124.72	914.66	320.7	1079.37	623.25	1124.66	748.59	334.56	505.56	786.55	629.22	6085.65	3702.76	5927.16	2103.6	1378.99	787.93	770.12	990.13	121.58	181.84	18,086.89	138.84	365.61	3285.02	2121.65	1665.29	3627.03	59,531.93
23	0.49	0.09	0.09	0.02	0.27	1.7	0.29	0.33	0.08	1.35	1.67	7.41	0.69	0.89	1.2	3	0.65	0.63	0.35	0.03	0.91	4.86	21.61	0	0.11	41.32	7.44	13.73	111.21
24	0.03	0.05	0.07	0.03	0.06	1.31	0.1	0.13	0.09	0.27	0.09	0.7	0.43	0.31	0.11	0.3	0.2	0.25	0.28	0.05	0.11	0.71	0.02	1.1	1.8	0.44	1.4	2.63	13.05
25	0.32	0.05	0.01	0.01	0.01	0.1	0.04	0.04	0.05	0.04	0.01	0.19	0.06	0.06	0.04	0.09	0.06	0.05	0.18	0.01	0.03	0.82	0.01	0.01	30.77	0.82	0.86	9.48	44.22
26	174.52	33.33	5.85	18.56	19.81	276.71	95.19	120.19	60.84	84.27	67.89	397.61	174.73	188.35	93.58	190.03	177.84	139.14	206.61	17.81	13.33	79.52	8.58	2.64	661.21	1502.58	623.57	1575.85	7010.14
27	22.77	5.06	0.89	2.76	2.58	66.13	14.02	25.34	19.02	21.66	7.16	63.74	27.55	24	15.65	36.11	53.67	29.39	38.52	3.86	2.73	10.83	1.12	0.4	101.09	46.68	30.77	209.07	882.57
28	16.82	12.24	3.49	4.29	3.88	25.82	5.31	8.06	4.34	10.01	4.61	41.4	18.26	24.99	10.86	24.42	23.59	14.08	24.42	2.13	2.09	20.84	1.78	2.32	177.3	142.45	161.74	529.25	1320.78
Sum	3060.54	2326.81	497.15	1918.47	1380.65	3046.89	1886.31	1277.94	1054.92	2256.03	4749.75	20,250.47	10,176.16	17,175.16	3953.22	3245.72	3286.87	2105.98	1935.67	307.78	374.81	22,024.46	381.79	3162.21	15,351.27	12,933.31	3200.94	24,457.82	167,779.12

.

Combined with Table 3, we can observe that the CBEF of the high CBEF sectors mainly originates from the EF generated by producing intermediate input products required by high RBEF sectors; the RBEF of high RBEF sectors mainly originates from the EF generated by intermediate input products provided by high CBEF sectors for them. There is a clear correspondence between the two. For example, the production and supply of electric power and heat (22), which has the most significant CBEF, has the largest CBEF caused by the energy consumed for the production of the intermediate product inputs required by its sector, followed by the non-metallic mineral products sector (12), the metal smelting and rolling processing sector (14), and other sectors (28), which are 180,868,900 hectares, 60,856,500 hectares, 59,271,600 hectares, and 36,270,300 hectares in size, respectively. The petroleum processing and coking sector (11) has the most significant CBEF due to the inputs required for the production of transportation, storage, post, and telecommunications (26) and the chemical sector (12): 88,626,900 hectares and 75,994,700 hectares, respectively. Therefore, when formulating energy conservation and emission reduction policies and measures, we should consider the sectoral EF and, more importantly, the EF of other sectors that are “caused” by this to clarify the “rights, responsibilities, and benefits” of energy conservation and emission reduction.

5.2. Direct and Indirect Energy Footprint Decomposition

We take the input output table in 2017 as an example. The EF of 28 sectors was decomposed into direct and indirect energy footprints. The results are shown in

Table 6. China sectoral direct and indirect energy footprint breakdown in 2017.

Sector Code	Direct Energy Footprint	Sector Share	Indirect Energy Footprint	Sector Share	Sectoral Energy Footprint	Energy Footprint Share
1	350.021	11.437	2710.519	88.563	3060.540	1.824
2	132.316	5.687	2194.498	94.313	2326.814	1.387
3	9.517	1.914	487.630	98.086	497.147	0.296
4	1.890	0.099	1916.578	99.901	1918.468	1.143
5	6.026	0.436	1374.627	99.564	1380.653	0.823
6	336.570	11.046	2710.317	88.954	3046.887	1.816
7	190.751	10.112	1695.561	89.888	1886.311	1.124
8	20.549	1.608	1257.393	98.392	1277.942	0.762
9	23.817	2.258	1031.102	97.742	1054.920	0.629
10	315.032	13.964	1940.995	86.036	2256.027	1.345
11	4711.011	99.184	38.741	0.816	4749.752	2.831
12	1619.001	7.995	18,631.469	92.005	20,250.470	12.070
13	545.840	5.364	9630.319	94.636	10,176.159	6.065
14	1874.240	10.913	15,300.922	89.087	17,175.162	10.237
15	31.318	0.792	3921.905	99.208	3953.223	2.356
16	122.760	3.782	3122.957	96.218	3245.716	1.935
17	68.503	2.084	3218.368	97.916	3286.871	1.959
18	21.035	0.999	2084.943	99.001	2105.979	1.255
19	25.685	1.327	1909.989	98.673	1935.673	1.154
20	2.943	0.956	304.841	99.044	307.785	0.183
21	22.052	5.884	352.760	94.116	374.812	0.223
22	3709.163	16.841	18,315.296	83.159	22,024.459	13.127
23	38.702	10.137	343.089	89.863	381.791	0.228
24	5.326	0.168	3156.888	99.832	3162.214	1.885
25	42.306	0.276	15,308.964	99.724	15,351.269	9.150
26	2650.551	20.494	10,282.761	79.506	12,933.311	7.709
27	327.695	10.237	2873.249	89.763	3200.944	1.908
28	719.447	2.942	23,738.374	97.058	24,457.820	14.577
Total	17,924.065	10.683	149,855.054	89.317	167,779.119	100

and Figure 2. It can be seen that the direct EF of the 28 sectors in 2017 was 179,240,650,000 hectares, and the indirect EF was 149,855,054,000 hectares, accounting for 10.683% and 89.317% of the total, respectively. The indirect EF was much higher than the direct EF. Specifically, among the subdivided sectors, except for petroleum processing and coking (11), the indirect energy footprints of all other sectors were much higher than the direct energy footprints, especially for metal mining (4), water production and supply (24), construction (25), non-metallic mineral and other mineral extraction (5), metal products (15), instrumentation and cultural and office supplies machinery manufacturing (20), electrical machinery and equipment manufacturing (18), and other sectors, whose direct energy footprints accounted for less than 1% of the total. So, the energy consumed by sectors is minimal to produce the final products and the resulting EF. A greater part of the EF comes from the energy consumption of intermediate-use products used in the production of the final products and services and the resulting EF, which is consistent with the conclusion obtained from the decomposition based on the CBEF and RBEF perspectives. Therefore, energy saving and emission reduction in the production and utilization of intermediate-use products is an important focus and force point of energy saving and emission reduction policies for the whole sector.

Combined with Figure 2, it can be seen that the sectors with large indirect energy footprints include other sectors (28), electricity and heat production and supply (22), the chemical sector (12), the metal smelting and rolling processing sector (14), and the construction sector (25), and the indirect energy footprints of these five sectors account for about 60% or more of the total indirect energy footprints, while these five sectors are also the five sectors with the largest energy footprints based on the decomposition of the direct and indirect energy footprints, and their energy footprints account for about 50% of the total energy footprints of the sector.

The trends of the direct and indirect energy footprints of 28 sectors in 2007, 2012, and 2017 were further measured and analyzed, and the results are shown in Figure 3. Consistent with the findings of the analysis in the previous section, overall, the EF composition of the sectors in the three years is mainly derived from the indirect EF, and the direct EF accounts for a relatively small amount. Its dynamic changes are also mainly determined by the changes in the indirect EF. It should be noted that the petroleum processing and coking sector (11) shows the most significant change in the indirect EF, decreasing from 52.490% in 2007 to 0.816% in 2017, and its EF composition also changes from relatively equal direct and indirect footprints to mainly being composed of the direct EF.

6. Conclusions and Policy Recommendations

Input output tables were characterized as a comprehensive reflection of the correlation of inputs and outputs between the sectors of the national economy, showing the interdependent and constraining economic linkages between the productive sectors. One shows the realization of the producer and consumer principles, and the other shows the state of the intermediate input products and initial inputs that each sector obtains from other sectors to produce. The accounting function encompasses direct economic linkages in the production process and indirect economic linkages between sectors. Input output tables provide a basis for studying the sectoral structure, including the formulation of national economic plans, price decisions, and quantitative analyses.

6.1. Conclusions

This paper measures and decomposes the environmental pollution problem in China’s sectors by using input output table data from 2007, 2012, and 2017 as the research object and characterizing environmental pollution by the EF, which can characterize the dual meaning of energy saving and emission reduction. The main contents and conclusions include the following two aspects:

Firstly, based on the principles of producer responsibility and consumer responsibility, the CBEEF and RBEF of the sectors were measured, and the results show that, on the one hand, the consumption-based energy footprints of the sectors varied greatly, while the distribution of the consumption-based energy footprints was more concentrated, and the consumption-based energy footprints of three sectors, namely, the electricity and heat production and supply sector, the petroleum processing and coking sector, and the metal smelting and rolling processing sector accounted for more than 70% of the consumption-based energy footprints. The distribution of the RBEF was more dispersed, and the variability among sectors was relatively small. On the other hand, the production and supply of electricity and heat, the metal smelting and rolling processing sector, the chemical sector, the non-metallic mineral products sector, and other sectors are “double-high” sectors in terms of their consumption-based and responsible-based energy footprints and produce serious environmental pollution. There should be a focus on these high emission sectors over a longer period of energy conservation and emission reduction in China. Instruments and cultural and office supplies machinery manufacturing, other manufacturing sectors, gas production, and supply sectors are typical “double-low” sectors, but also, in the future, China should further strengthen and optimize the cleanliness of the key sectors. In addition, the petroleum processing and coking sector is a typical high EF sector, and in addition to other sectors, including construction, transportation, storage, post, and telecommunications, is typically a high RBEF sector. In order to achieve green production in these sectors, efficient use is also a focus of China’s energy saving and emission reduction strategy.

Secondly, the energy footprints of the sectors were decomposed based on the dual perspectives of energy consumption and responsibility, both direct and indirect. The results show that, on the one hand, the energy footprints of sectors with high consumption-based energy footprints mainly come from the energy footprints of the intermediate input products required for the production of sectors with high responsible-based energy footprints, while the responsible-based energy footprints of sectors with high responsible-based energy footprints also mainly come from the energy footprints of sectors with high consumption-based energy footprints for the provision of intermediate input products. The correspondence between the two is clear. On the other hand, the EF is very small, in spite of China’s sectors producing the final products and services. Additionally, a greater portion of the EF comes from the EF required for the production of intermediate-use products. Specifically, the sectoral EF consists of nearly 90% of the indirect EF. Except for the petroleum processing and coking sector, the indirect energy footprints of the remaining 27 sectors are much larger than the direct energy footprints, and the direct energy footprints of some sectors are less than 1% of the total. Taken together, the clean, low-carbon production and the scientific and efficient use of intermediate-use products are the focuses of energy conservation and emission reduction policies for the sector as a whole.

6.2. Policy Recommendations

Based on the above findings, this paper proposes the following policy references:

Firstly, we need to do a good job of clarifying the responsibility of the sectoral environmental pollution subjects and clarify the sectoral energy saving and emission reduction “rights, responsibilities, and benefits” of the top-level design. Each sectoral energy consumption and environmental pollution level is playing a different role. Some sectors are not major sources of energy consumption but are major sources of environmental pollution. Therefore, the government should provide sectoral environmental pollution accounting accuracy based on the polluter pays principle (PPP), clarify the main environmental pollution management strategies, and make environmental pollution management responsibility clear. A practical method is to improve and optimize China’s sectoral energy input output balance sheet and build a national environmental pollution balance sheet to account for environmental pollution responsibilities and clarify the “rights, responsibilities, and benefits” associated with sector energy conservation and emission reduction and path selection.

Secondly, based on the environmental pollution characteristics of the sector, differentiated and targeted energy saving and emission reduction policies and measures need to be developed. The pollution from the CBEF and RBEF are inconsistent across sectors, requiring the development of differentiated energy saving and emission reduction policies. On the one hand, sectors with a high CBEF mainly include the production and supply of electricity and heat (22), petroleum processing and coking (11), and metal smelting and rolling processing (14) sectors. These sectors consume the most energy due to the inputs used to produce their products and generate the most serious environmental pollution problems. These sectors should focus on implementing carbon emission reduction policies based on the producer responsibility principle. Firstly, to enhance the proportion of renewable energy in total energy consumption, there should be a focus on clean and low-carbon production in the sector, specifically through digitalization, intelligence, and other methods of technological upgrading and transformation, to improve the sectoral green production technology, and environmentally friendly and low-carbon methods of production should be used to achieve the energy saving and emission reduction goals. Secondly, low-carbon production technology should be used to reduce the consumption of fossil fuels and other sources of high emission energy. Third, clean energy should be developed to reduce the share of fossil energy in the total energy used. On the other hand, sectors with high RBEF mainly include other sectors (28), the production and supply of electricity and heat (22), and the chemical sector (12). This sector consumes the most energy from the intermediate products used to produce its products and generates the most serious pollution problems, making it a hidden high-polluting sector. Therefore, it should focus on implementing carbon reduction policies based on the principle of consumer responsibility. Specific policies include enhancing the sectoral production efficiency, solving the problem of efficient use of low-energy-consuming intermediate inputs, improving the production efficiency, and finding intermediate substitutes to reduce implicit environmental pollution to realize the sectoral energy saving and emission reduction goals.

Thirdly, we should further optimize the sectoral structure, reduce the proportion of polluting sectors, and increase the proportion of clean sectors. Adjusting and optimizing the sectoral structure, vigorously developing strategic emerging sectors, and promoting the emergence of new sectors and new business models are important methods to achieve high-quality economic development and energy conservation and emission reduction. Specific measures to optimize the sectoral structure and improve sectoral cleanliness include further promotion of the structural reform on the supply side, eliminating the backward production capacity, especially the hidden high environmental pollution sectoral capacity. The government can impose certain constraints on production by such sectors through a series of price, tax, finance, and trade policies and use the opposite policy to encourage the development of sectors with high cleanliness to optimize the sectoral structure and achieve clean, efficient, and high-quality development. On the other hand, we should further strengthen the innovation-driven strategy, use new technologies and methods to upgrade and transform traditional sectors, promote the emergence of new sectors and new models, increase the proportion of strategic new sectors, promote the cleanliness and optimal adjustment of the sectoral structure, and improve the overall green development of sectors.

In addition, the root cause of environmental pollution in the sector is a problem associated with energy consumption, so the development of energy saving and emission reduction policies and measures in the sector consider the energy consumption aspects of the development of corresponding policy measures. Specifically, this can start from the following three aspects: Firstly, we should improve the development of new energy sectors and new energy consumption substitution ratios. In order to improve the efficiency of energy use, on the one hand, cleaner use of traditional coal, oil, gas, and other fossil energy sources should be promoted. On the other hand, we should actively promote the development of renewable energy sources and the use of technology research and vigorously develop solar energy, wind power, bio-energy, tidal energy, and other clean energy technologies. Secondly, we should adhere to systematic thinking and building a new situation of joint participation between multiple subjects of interest. The environment is a systemic issue that requires representatives from different sectors, residents’ organizations, and experts to form a synergy and jointly participate in reducing energy footprints and achieving energy conservation and emission reduction goals. Specifically, stakeholders are continuously organized to encourage the expression of different viewpoints through consultations, public hearings, and collaborative platforms, build consensus, and to ultimately incorporate the views of all parties into the policy framework to provide systematic guidance for the resolution of environmental issues. Thirdly, we should promote the synergistic development of the environment, the economy, and society. Environmental problems should not jeopardize economic development and social employment. Therefore, addressing environmental pollution problems should be balanced with the maintenance of economic growth, ensuring that there are employment opportunities and minimizing social disruption. In order to further enhance the acceptability of policies, employment transition policies, economic incentives, and social support schemes for sectors with high levels of energy consumption and responsible energy footprints should be developed on an ongoing basis.