Spatio-Temporal Evolution of Carbon Emission in China’s Tertiary Industry: A Decomposition of Influencing Factors from the Perspective of Energy-Industry-Consumption

Abstract

The development of the tertiary industry is of great significance for promoting industrial structure, optimizing and upgrading it, and achieving regional energy conservation and emission reduction goals. This study adopts a quantitative method to analyze the spatio-temporal pattern of carbon emissions from China’s tertiary industry from 2004 to 2019. In order to analyze emissions from aspects such as energy structure, energy intensity, energy carrying capacity, industrial structure, level of industrial development, income level, consumption capacity, energy consumption intensity, and population size, this study establishes a hybrid factor decomposition model called the “energy-industry-consumption” research framework. The study shows that carbon emissions from China’s tertiary industry have been increasing year by year from 2004 to 2019, with a growth rate of 353.10%. Transportation is the largest contributor to the increase in carbon emissions from China’s tertiary industry. The carbon emissions from the tertiary industry in each province show four types: high-speed growth, low-speed growth, fluctuating growth, and stable growth. During the study period, carbon emissions produce a spatial heterogeneity with the highest emissions in the south and lowest in the northwestern part of China. The spatial pattern of per capita carbon emissions is not significant. Guangdong has the highest carbon emissions, and Shanghai and Beijing have higher per capita carbon emissions. Industrial factors and consumption factors have a positive effect on carbon emissions in China’s tertiary industry, while energy factors have a negative effect. The leading factor of carbon emissions in China’s tertiary industry has gradually shifted from energy to industry.

1. Introduction

The issue of climate change and global warming has gradually become an important international social and political issue since the release of the third climate assessment report by the Intergovernmental Panel on Climate Change (IPCC) in 2001 [1,2,3,4]. The increase in atmospheric CO₂ concentration triggered by human industrial activities and increasing energy demands are the primary causes of global climate change [5,6,7]. As a result, carbon emissions and global environmental changes have become an important concern for sustainable development [8]. Developing a low-carbon economy and addressing climate change are now crucial tasks for every country globally [9,10]. China, the world’s largest carbon emitter, pledged in the 2015 Paris Agreement that its carbon dioxide emissions will peak by 2030. In 2020, the UN General Assembly went even further and proposed achieving carbon neutrality by 2060 [11]. In the same year, China announced at the 75th Session of the United Nations General Assembly that it would enhance its national contribution and take more decisive policies and measures to reduce carbon dioxide emissions, with the goal of peaking emissions before 2030 and achieving carbon neutrality before 2060 [12]. Developing a low-carbon economy has become an important development path for China. In this scenario, China must find a way to meet its carbon emission reduction goals in order to balance stable and rapid economic development with coordinated progress in energy and society, making it an urgent issue to solve.

As the main driving force of national economic development and the main source of carbon emissions [13,14], how to correctly handle the relationship between industrial development and carbon emission reduction is the primary consideration in the development processes of all provinces, cities, and autonomous regions in China [15]. The research on carbon emissions in China, other provinces, and autonomous regions mainly focuses on the following two categories: one is the research on carbon emission measurement with spatial and temporal differentiation and the other is the decomposition of factors influencing carbon emissions.

The research on China’s carbon emissions mainly focuses on the following two aspects. The first aspect is the measurement of carbon emissions and spatial and temporal differentiation research. For instance, Wang analyzed the overall changes, regional differences, spatial and temporal pattern, and agglomeration characteristics of county per capita carbon emissions in China from 2000 to 2017, and used panel quantile regression to explain the dynamic impact of social and economic development on county per capita carbon emissions [16]; Ma analyzed the regional differences and spatial correlation of provincial carbon emission efficiency in China [17]; Pan used the exploratory spatio-temporal data analysis (ESTDA) framework system and the pattern- and spatio-temporal-dependent dynamic evolution of carbon footprints from 2001 to 2013 from the perspective of spatio-temporal interaction [18]; Liang analyzed the carbon footprint pressure of 30 provinces in China from 2006 to 2015, and explored the driving factors [19]; Liu analyzed the carbon emissions and driving factors of 30 provinces in China using data from the transportation sector, and found that the energy intensity has a crucial impact on the carbon emissions of China’s transportation sector [20].

The second aspect is the study on the decomposition of factors influencing carbon emissions in China, and the common decomposition methods include SDA (structural decomposition analysis), STIRPAT (stochastic impacts by regression on population, affluence, and technology model) and IDA (index decomposition analysis) [21]. Based on the extended Kaya identity and additive LMDI method, Li decomposed the total CO₂ emissions into four influencing factors. Of those, the economic active effect is the most influential factor driving CO₂ emissions, with a contribution rate of 43.92%. The second driving factor is the population effect. On the contrary, the energy intensity effect is the most inhibiting factor, followed by the carbon energy structure effect [22]. From the perspective of the extended KAYA formula, Ren calculated the long-term equilibrium and short-term fluctuation relationship between carbon emission intensity and its influencing factors between 1980 and 2010 in China [23]. Zhu analyzed the direct and brief influencing factors of implicit carbon emissions in China’s construction industry based on the extended STIRPAT model. The study found that controlling building scale, reducing emissions from building material suppliers, improving technology, and preventing rebound is key to alleviating building energy efficiency [24]. Ling used the IO-SDA method to decompose the influencing factors of China’s thermal electricity and heating industry, and the results showed that energy demand had the most significant impact on carbon emissions, while the energy intensity and energy structure had limited effects [25].

The above-mentioned research works have played an important role in promoting the future development of carbon emission reduction in China. However, when they have decomposed the influencing factors of carbon emission, they do not take the impact of the industrial structure and the energy structure into account, so there will be double calculation which makes the calculation results inaccurate. At the same time, current research is mostly focused on the industrial field, while the tertiary industry’s possibility and speed of reducing the carbon emission intensity through structural upgrading is high [26], which has the characteristics of low carbon emissions [27].

China’s tertiary industry has developed rapidly, accounting for about 50% of GDP for many years. It can be said that the tertiary industry has become the pillar of China’s economic development. Certainly, from the perspective of carbon emissions, the carbon emissions from the tertiary industry are significantly lower than those from the secondary industry (approximately 10%) [14]. The carbon emissions from the secondary industry still account for a significant proportion of the total emissions. At the same time, compared with the secondary industry, the tertiary industry is far more likely and faster to expand its scale, upgrade its structure, and then reduce its carbon intensity. This also indicates that studying the carbon emissions of the tertiary industry can provide future impetus and support for the better development of the industrial economy. This is one of the important reasons why the tertiary industry has been chosen as the research subject in this study. Compared with developed countries, the proportion of China’s tertiary industry is still too low, especially the underdeveloped modern service industry, which is also an urgent problem in China’s economic structure. How big is the tertiary industry in carbon emissions in comparison with national emissions What is the trend of carbon emissions in the tertiary industry? Such data are required to justify the study.

There are few studies on the calculation and factor decomposition of carbon emissions in the tertiary industry despite the fact that this research is equally important to realize the innovation, with coordination among the actors to create green and shared industrial development. Therefore, this study corrects and expands the shortcomings of the traditional carbon emission model and constructs a factor decomposition model of carbon energy emissions for China’s tertiary industry on the basis of quantitative calculation of the carbon emissions in various provinces and cities in China. This study utilizes the perspective of “energy-industry-consumption” in the tertiary industry sector. The impact of energy structure, energy intensity, energy intensity, energy carrying capacity, industrial structure, level of industrial development, income level, consumer capacity, energy consumption intensity, and population size on carbon energy emissions of tertiary industries in various provinces and cities in China was quantitatively analyzed from 2004 to 2019 (Figure 1). The carbon emission was compared with the base year of 1995 and its spatial and temporal differences were discussed. The hypothesis of this study is that, from the perspective of the connotation of the tertiary industry, the three factors of energy, industry, and consumption correspond to different sectors within the tertiary industry. Therefore, carbon emissions from the tertiary industry should be closely related to these three factors. However, the specific impact and relationship are still unclear. This research has theoretical and practical implications to deeply adjust the tertiary industry in China to build a new environmentally friendly modern industrial system and achieve the carbon emission reduction target.

2. Data Sources and Methods

2.1. Data Sources

The “Energy-Industry-Consumption” decomposition of carbon emission factors in China’s provincial tertiary industry includes three main categories of data: energy data, economic data, and population data. However, due to the lack of data, the Tibetan Autonomous Region, Hong Kong Special Administrative Region, Macao Special Administrative Region, and Taiwan Province of China were excluded from the study.

(1) Energy data: All data are sourced from the “China Energy Statistical Yearbook (1991–1996, 2005–2016)” and the statistical yearbooks of various provinces, cities, and autonomous regions, such as the “Beijing Statistical Yearbook”. The energy consumptions in the tertiary industry in China’s provinces, cities, and autonomous regions for 1995, and yearly consumption from 2004 to 2019 (without calculating the consumption of electricity and heat and estimating only the carbon emissions generated by fossil fuel) were collected. Energy consumption data include raw coal, clean coal, coke, crude oil, gasoline, kerosene, diesel oil, fuel oil, naphtha, lubricating oil, paraffin, solvent oil, petroleum asphalt, petroleum coke, liquefied petroleum gas, other petroleum products, and natural gas. Due to the difference in the statistical content of the yearbook, the data of clean coal, naphtha and solvent oil in 2016–2019 are estimated based on the statistical data of 2004–2015 and the consumption of other energy in 2016–2019. The data required to measure carbon emissions for each type of energy are obtained from the “Reference Coefficients for Converting Various Energy Sources into Standard Coal” reported in the China Energy Statistics Yearbook 2008 and the IPCC report [28].

(2) Economic data: This includes the output value of the tertiary industry (GDP of the tertiary industry, transportation, storage, postal service, wholesale, retail, accommodation, catering, and others) for various provinces and autonomous regions in China from 1995 and 2004 to 2019. According to the basic division of the tertiary industry and the basis of previous studies, it can be divided into 3 categories: transportation, storage, and postal services; wholesale, retail, accommodation, and catering; and others (finance and real estate). The output value data are sourced from the “China Statistical Yearbook on the Tertiary Industry” for the years 2005 to 2020, which also includes the GDP and sub-item output value of the tertiary industry for different regions in 1995. The total income and expenditure of each province in China is derived from the “China Statistical Yearbook” for 1996 and 2005 to 2020, and missing data are supplemented with data from the “China Statistical Yearbook”.

(3) Population data: According to the “China Statistical Yearbook”, the total population of each research area was collected from 1995 and 2004 to 2019. The missing population data were supplemented using the population and employment statistical yearbooks of provinces, cities, and autonomous regions.

2.2. Research Methods and Models

2.2.1. Model for Measuring Carbon Emissions from the Tertiary Industry

This study calculated the carbon emissions for 1995 and yearly emissions from 2004 to 2019 from fossil fuel consumption by the tertiary industry. In order to improve the accuracy of measuring carbon emissions from the tertiary industry, this study disregards the carbon emissions from electricity and heat generated by the re-consumption of fossil fuels. Based on the real energy consumption of the tertiary industry in China, the actual combustion efficiency of all kinds of energy sources are distinguished. At the same time, the carbon emissions of the tertiary industry in China are calculated according to the carbon emission calculation guidelines recommended by IPCC and previous research [29,30,31]. The carbon energy emission measurement model of the tertiary industry is shown in Formula (1):

$$TC={\sum }_i{\sum }_i{\sum }_q{C}_{ijq}=E_{ijq}\times {LCV}_i\times {CF}_i\times \frac{12}{44}$$

(1)

where i, j, and q are different types of energy, different industries, and different regions, respectively. TC is the total carbon energy emissions from the tertiary industry (10⁴ t). E is the energy consumption (10⁴ t). LCV is the average low-level heat output (KJ/kg). CF is the carbon dioxide emission factor (kg CO₂/TJ); 0.2727 is the molecular weight of the C atom in carbon dioxide, and the factors are shown in

Table 1. Carbon emission factors for energy consumption.

Types of Energy		Carbon Dioxide Emission Factor (kgCO2/TJ)	Average Low Level Heat Generation (KJ/kg)	Discount Factor for Standard Coal (tce/t)
Coal	Raw Coal	94,600	20,908	0.7143
	Cleaned Coal	94,600	26,344	0.9000
	Coke	107,000	28,435	0.9714
Oil	Crude Oil	73,300	41,816	1.4286
	Gasoline	69,300	43,070	1.4714
	Kerosene	71,900	43,070	1.4714
	Diesel Oil	74,100	42,652	1.4571
	Fuel Oil	77,400	41,816	1.4286
	Naphtha	63,100	41,816	1.5000
	Lubricating Oil	56,100	41,816	1.4331
	Paraffin	73,300	41,816	1.3648
	Solvent Oil	94,600	41,816	1.4672
	Petroleum Asphalt	94,600	41,816	1.3307
	Petroleum Coke	107,000	41,816	1.0918
	Liquefied Petroleum Gas	73,300	50,179	1.7143
	Other Petroleum Products	69,300	41,816	1.4000
Gas	Natural Gas	71,900	38,931	1.3300 (tce/103 m3)

.

The integrated energy consumption is measured as shown in Equation (2).

$$TSE={\sum }_i{\sum }_j{\sum }_q{SE}_{ijq}=E_{ijq}\times {CSC}_i$$

(2)

TSE is the combined energy consumption (10⁴ tce). CSC is the coefficient of conversion of physical energy into standard coal.

2.2.2. “Energy-Industry-Consumption” Decomposition Model of Carbon Emission in the Tertiary Industry

The decomposition of energy emissions has been studied for a long time, and different models have different focuses in the decomposition of energy emissions; the Divisia index decomposition focuses on energy and industry [26], while the LMDI factor decomposition and the Kaya equation introduce the effect of population size into the decomposition [32], compensating for the absence of the main consumption. The extended Kaya constant equation further decomposes economic factors to investigate the influence of energy emissions at a deeper level [33,34]. However, in terms of industry characteristics, the tertiary industry is more significantly influenced by the consumption endpoint than the primary and secondary sectors. Therefore, this study considers the influence of consumption on the carbon emissions of the tertiary industry, and further decomposes energy and industry to analyse the mechanism of carbon emissions of the tertiary industry on a deeper level. The decomposition model is shown in Equation (3).

$$TC={\sum }_i{\sum }_j{\sum }_q{C}_{ijq}=\frac{{SE}_{\mathrm{i}\mathrm{j}\mathrm{q}}}{{TSE}_{\mathrm{j}\mathrm{q}}}\times \frac{{TSE}_{jq}}{Y_{jq}}\times \frac{P_q}{{TSE}_{jq}}\times \frac{Y_{jq}}{Y_q}\times \frac{Y_q}{P_q}\times \frac{I_q}{P_q}\times \frac{O_q}{I_q}\times \frac{{TSE}_{jq}}{O_q}\times P_q$$

(3)

where $Y$ is the value of tertiary industry output (10⁸ CNY), $P$ is the population size (10⁴ people), and $I$ is the total income of the people’s livelihoods (10⁴ CNY).$O$ is the total expenditure of people’s livelihoods (10⁴ CNY). The currencies involved in this study are all Chinese yuan. The indicators for each factor are defined as shown in

Table 2. Decompositions and definitions of impact factors.

Type		Formula	Definition
Energy factors	Energy mix	${ES}_{\mathrm{i}\mathrm{j}\mathrm{q}}=\frac{{SE}_{\mathrm{i}\mathrm{j}\mathrm{q}}}{{TSE}_{\mathrm{j}\mathrm{q}}}$	Share of different types of energy consumption within the tertiary industry;
	Energy intensity	${EI}_{\mathrm{j}\mathrm{q}}=\frac{{TSE}_{jq}}{Y_{jq}}$	Resources consumed per unit of output within the tertiary industry;
	Energy carrying capacity	${EC}_{\mathrm{j}\mathrm{q}}=\frac{\frac{1}{{TSE}_{\mathrm{j}\mathrm{q}}}}{P_q}$	Inverse of energy consumption per unit of population within the tertiary industry.
Industrial factors	Industrial structure	${IS}_{\mathrm{j}\mathrm{q}}=\frac{Y_{jq}}{Y_q}$	Share of different sectors within the tertiary industry;
	Level of industrial development	${IDL}_{\mathrm{q}}=\frac{Y_q}{P_q}$	Tertiary industry output per unit of population.
Consumption factors	Income level	${IL}_{\mathrm{q}}=\frac{I_q}{P_q}$	Disposable income per capita;
	Consumer capacity	${CA}_{\mathrm{q}}=\frac{O_q}{I_q}$	Consumption expenditure as a proportion of disposable income;
	Energy consumption intensity	${ECI}_{\mathrm{j}\mathrm{q}}=\frac{{TSE}_{jq}}{O_q}$	Amount of energy consumed per unit of consumption;
	Population size	$P_q$	Population size.

.

As shown in Table 2, Equation (3) can be deformed into Equation (4).

$$\mathrm{T}\mathrm{C}={\sum }_i{\sum }_j{\sum }_q{C}_{ijq}={ES}_{\mathrm{i}\mathrm{j}\mathrm{q}}\times {EI}_{ijq}\times {\mathrm{E}\mathrm{C}}_{\mathrm{j}\mathrm{q}}\times {\mathrm{I}\mathrm{S}}_{\mathrm{j}\mathrm{q}}\times {IDL}_{\mathrm{q}}\times {IL}_{\mathrm{q}}\times {CA}_{\mathrm{q}}\times {ECI}_{\mathrm{i}\mathrm{j}\mathrm{q}}\times P_q$$

(4)

The change in tertiary industry carbon emissions in the study period t compared to the beginning of the study can then be expressed as (5).

$$\begin{gathered} ∆\mathrm{T}\mathrm{C}={TC}^t-{TC}^0={\sum }_i{\sum }_j{\sum }_q{ES}_{ijq}^t\times {EI}_{ijq}^t\times {EC}_{jq}^t\times {IS}_{jq}^t\times {IDL}_q^t\times {IL}_q^t\times {CA}_q^t\times {ECI}_{jq}^t\times P_q^t \\ -{\sum }_i{\sum }_j{\sum }_q{EF}_{ijq}^0\times {ES}_{ijq}^0\times {EI}_{ijq}^0\times {EC}_{jq}^0\times {IS}_{jq}^0\times {IDL}_q^0\times {IL}_q^0\times {CA}_q^0\times {ECI}_{jq}^0\times P_q^0 \end{gathered}$$

(5)

It can be further transformed into (6).

$$∆\mathrm{T}\mathrm{C}=∆{TC}_{ES}+∆{TC}_{EI}+∆{TC}_{EC}+∆{TC}_{IS}+∆{TC}_{IDL}+∆{TC}_{IL}+∆{TC}_{CA}+∆{TC}_{ECI}+∆{TC}_P$$

(6)

Of these, the effect of each indicator on carbon emissions is calculated as follows (7)–(15):

$$\mathrm{Energy} \mathrm{mix}. ∆{TC}_{ES}={\sum }_{\mathrm{i}}{\sum }_j{\sum }_q\frac{{TC}_{ijq}^t-{TC}_{ijq}^0}{ln{TC}_{ijq}^t-ln{TC}_{ijq}^0}\times ln\left(\frac{{ES}_{ijq}^t}{{ES}_{ijq}^0}\right)$$

(7)

$$\mathrm{Energy} \mathrm{intensity}. ∆{TC}_{EI}={\sum }_j{\sum }_q\frac{{TC}_{ijq}^t-{TC}_{ijq}^0}{ln{TC}_{ijq}^t-ln{TC}_{ijq}^0}\times ln\left(\frac{{EI}_{jq}^t}{{EI}_{jq}^0}\right)$$

(8)

$$\mathrm{Energy} \mathrm{carrying} \mathrm{capacity}. ∆{TC}_{EC}={\sum }_j{\sum }_q\frac{{TC}_{jq}^t-{TC}_{jq}^0}{ln{TC}_{jq}^t-ln{TC}_{jq}^0}\times ln\left(\frac{{EC}_{jq}^t}{{EC}_{jq}^0}\right)$$

(9)

$$\mathrm{Industrial} \mathrm{structure}.∆{TC}_{IS}={\sum }_{\mathrm{j}}{\sum }_q\frac{{TC}_{jq}^t-{TC}_{jq}^0}{ln{TC}_{jq}^t-ln{TC}_{jq}^0}\times ln\left(\frac{{IS}_{jq}^t}{{IS}_{jq}^0}\right)$$

(10)

$$\mathrm{Level} \mathrm{of} \mathrm{industrial} \mathrm{development}. ∆{TC}_{IDL}={\sum }_q\frac{{TC}_q^t-{TC}_q^0}{ln{TC}_q^t-ln{TC}_q^0}\times ln\left(\frac{{IDL}_q^t}{{IDL}_q^0}\right)$$

(11)

$$\mathrm{Income} \mathrm{level}.∆{TC}_{IL}={\sum }_q\frac{{TC}_q^t-{TC}_q^0}{ln{TC}_q^t-ln{TC}_q^0}\times ln\left(\frac{{IL}_q^t}{{IL}_q^0}\right)$$

(12)

$$\mathrm{Consumer} \mathrm{capacity}. ∆{TC}_{CA}={\sum }_q\frac{{TC}_q^t-{TC}_q^0}{ln{TC}_q^t-ln{TC}_q^0}\times ln\left(\frac{{CA}_q^t}{{CA}_q^0}\right)$$

(13)

$$\mathrm{Energy} \mathrm{consumption} \mathrm{intensity}. ∆{TC}_{ECI}={\sum }_{\mathrm{i}}{\sum }_j{\sum }_q\frac{{TC}_{ijq}^t-{TC}_{ijq}^0}{ln{TC}_{ijq}^t-ln{TC}_{ijq}^0}\times ln\left(\frac{{ECI}_{ijq}^t}{{ECI}_{ijq}^0}\right)$$

(14)

$$\mathrm{Population} \mathrm{size}. ∆{TC}_{CA}={\sum }_q\frac{{TC}_q^t-{TC}_q^0}{ln{TC}_q^t-ln{TC}_q^0}\times ln\left(\frac{{CA}_q^t}{{CA}_q^0}\right)$$

(15)

3. Results and Analysis

3.1. Spatio-Temporal Distribution Characteristics of Carbon Emissions in the Tertiary Industry

3.1.1. Characteristics of the Time Change in Carbon Emission in the Tertiary Industry

The carbon emissions of China’s tertiary industry showed a yearly increasing trend between 2004 and 2019. The emission was 108.4677 million tons in 2004, which was increased to 491.4707 million tons in 2019, with a growth rate of 353.10%. At the same time, the carbon emissions of the tertiary industry in different regions of China exhibit a growing trend over time (Figure 2).

The tertiary industry in East China has the highest carbon emissions, increasing from 32.2539 million tons in 2004 to 117.3588 million tons in 2019, with transportation emission accounting for more than 50% of the growth in tertiary industry emissions during the study period. Between 2004 and 2019, the carbon emissions of the tertiary industry in East China accounted for more than 24% of the total carbon emissions of the tertiary industry in China. This is mainly because the economically developed provinces such as Shanghai, Zhejiang, and Jiangsu belong to East China, where the development of the tertiary industry was found to be relatively high. At the same time, the carbon emissions of the tertiary industry in East China were mainly concentrated in the transportation, storage, and postal services. The region has the most convenient comprehensive transportation network in China, with internationally important ports such as Shanghai, Ningbo-Zhoushan, and Qingdao. As a result, a substantial amount of carbon emissions is generated. The slow economic and physical infrastructure development was responsible for the lowest carbon emissions of the tertiary industry in Southwest China. The emissions were 8.6293 million tons in 2004 which were increased to 35.3708 million tons in 2019. The carbon emissions in Southwest China were relatively low due to the complex topography hindering industrial development, which further complicated transportation services. The region has unique natural landscapes and cultural attractions. If the transportation conditions can be improved, it is possible to attract a large number of tourists and develop the accommodation and catering industries.

From 2004 to 2019, the carbon emissions of the tertiary industry in provincial administrative units in China showed different development trends over time (Figure 3). In Figure 3, blue represents growth rates lower than the national average, while green represents growth rates higher than the national average. Based on the degree of carbon emission changes, the growth rate of the tertiary industry, and the standard deviation of data, it can be divided into four types:

(1) High-speed growth: The average annual growth rate of carbon emissions of the tertiary industry in these provincial administrative units was significantly higher than the national average (11.82%). Yunnan (22.90%), Guizhou (17.17%), Hunan (16.61%), Anhui (16.49%), Qinghai (14.77%), Chongqing (14.03%), Ningxia (13.95%), Hubei (13.69%), Heilongjiang (13.67%), and Henan (13.41%) had experienced this type of growth.

(2) Low-speed growth: The tertiary industry’s average annual growth rate of carbon emissions in these provincial administrative units was lower than the national average. Those administrative units were Tianjin (1.96%), Shanghai (8.08%), Beijing (8.10%), Fujian (8.33%), Hainan (8.52%), Zhejiang (8.82%), Jiangsu (9.36%), Guangxi (9.39%), Liaoning (9.47%), Shaanxi (9.54%), and Jilin (9.82%).

(3) Fluctuating growth: The carbon emissions in the tertiary industry have fluctuated in some provincial administrative units. The average annual growth rate of carbon emissions in their tertiary industries is similar to the average growth rate in China. The carbon emissions in Shandong (the standard deviation is 29.26%), Shanxi (27.96%), Hebei (22.14%), Inner Mongolia (15.02%), and Xinjiang (14.26%) had shown fluctuating growth.

(4) Stable growth: The average annual growth rate of carbon emissions in some provincial administrative units’ tertiary industries is neither fast nor slow, and there is no apparent fluctuating pattern. It shows a stable growth trend. Sichuan (12.69%), Gansu (11.29%), Jiangxi (10.46%), and Guangdong (10.04%) exhibited this characteristic.

3.1.2. Spatial Characteristics of Carbon Emission in China’s Tertiary Industry

Significant spatial differences have been observed in the carbon emissions in the tertiary industry of China during the study period. The spatial pattern showed the highest carbon emission in southern China and the lowest in the northwest (Figure 4). From the perspective of the cumulative carbon emissions in the tertiary industry, Guangdong has the highest cumulative carbon emissions among all provinces and cities from 2004 to 2019. During the study period, the carbon emissions from the tertiary industry in Guangdong are 318.9811 million tons. Shandong comes in second position with accumulated carbon emissions of 314.6317 million tons. Similarly, Inner Mongolia ranked third with cumulative carbon emissions from the tertiary industry of 251.4127 million tons. For the same period, the combined tertiary industry emissions in the first three provinces accounted for 20.15% of the national carbon emissions. Guangdong is located in the Pearl River Delta region of China. It is bordered to the south by the Hong Kong Special Administrative Region, Macau Special Administrative Region, and the South China Sea. Guangdong has a favorable geographical location, well-developed transportation by land and water, prosperous foreign trade, and a high overall quality of regional development. Guangdong possesses unique Cantonese culture, and its tourism industry is well developed. Additionally, it is the province with the highest permanent resident population in China. As people’s living standards gradually improve, the carbon emissions generated by the tertiary industry have also significantly increased. On the other hand, the lowest cumulative carbon emissions in the tertiary industry were observed in Ningxia, Qinghai, and Hainan, with the cumulative carbon emissions of 22.9614, 24.7500, and 31.9367 million tons, respectively, during the study period. This situation is mainly related to the low level of economic development in these provinces.

From the spatial distribution of per capita carbon emissions in China’s tertiary industry, it can be observed that it generally exhibits a characteristic of high density in the eastern part of China and low density in the western (Figure 5). However, compared to the spatial characteristics of total carbon emissions, the spatial clustering pattern is not significant. Among them, Shanghai and Beijing had higher per capita carbon emissions, and the per capita carbon emissions of the tertiary industry increased from 0.49 tons/person and 0.42 tons/person in 2004 to 1.12 tons/person and 0.88 tons/person in 2019, respectively. This is mainly because Shanghai and Beijing, as the center of economic development and politics in China, had a high level of economic development, high living standards, and high demand for transportation, hotels, and restaurants, making their tertiary industry footprint stand out among other provinces.

3.2. Decomposition of Carbon Emission Factors in China’s Tertiary Industry under “Energy-Industry-Consumption”

3.2.1. Decomposition of Carbon Emission Factors in the Tertiary Industry at the National Scale

In this study, energy, industry, and consumption are taken as the three major factors affecting the carbon emissions in China’s tertiary industry, and the influence degree of these three factors is calculated according to the decomposition model shown in Equation (3). The specific calculation results are shown in

Table 3. Contribution values of the three major elements to the carbon emissions of China’s tertiary industry from 2004 to 2019.

Year	ΔTCE Contribution Value (Billion Tons)	ΔTCTI Contribution Value (Billion Tons)	ΔTCFC Contribution Value (Billion Tons)
2004	−0.87	0.90	0.63
2005	−1.03	1.09	0.99
2006	−1.21	1.29	1.14
2007	−1.41	1.51	1.33
2008	−1.69	1.82	1.61
2009	−1.93	2.09	1.81
2010	−2.21	2.42	2.05
2011	−2.51	2.76	2.27
2012	−2.80	3.09	2.49
2013	−2.90	3.21	2.45
2014	−3.00	3.32	2.52
2015	−3.17	3.63	2.69
2016	−3.33	3.79	2.75
2017	−3.44	3.96	2.80
2018	−3.59	4.17	2.92
2019	−3.74	4.38	3.01

.

Industrial factors and consumption factors have a positive effect on the carbon emissions of China’s tertiary industry, while energy factors have a negative effect on the carbon emissions of the tertiary industry. Industrial factors are the main factors causing the growth of carbon emissions in China’s tertiary industry. China’s tertiary industry has developed rapidly, with the output value of the tertiary industry rising from 6609.7 billion yuan in 2004 to 52,852.5 billion yuan in 2019. Although compared with the secondary industry, the tertiary industry itself is characterized by low carbon intensity and its rapid expansion of the industrial scale has brought about the growth of the carbon emissions of the tertiary industry. The consumption factor plays an important role in the growth of carbon emissions in China’s tertiary industry. With the continuous growth of China’s population and the continuous improvement of people’s living standards, people’s demand for transportation, accommodation, and catering is increasingly strong, which brings about the growth of carbon emissions in China’s tertiary industry. The energy intensity of China’s tertiary industry decreased from 0.16 tons of carbon/10,000 yuan in 2004 to 0.08 tons of carbon/10,000 yuan in 2019. The decline of energy intensity has inhibited the rise of carbon emissions in China’s tertiary industry.

In order to explore the impact of the three factors on carbon emissions from China’s tertiary industry in more detail, this study further decomposes the energy factor into three aspects: the energy structure, the energy intensity, and the energy carrying capacity. The industrial factor is decomposed into the industrial structure and the level of industrial development. The consumption factor is decomposed into the income level, the consumer capacity, the energy consumption intensity, and the population size. The contribution of each factor to the carbon emissions from China’s tertiary industry is shown in Figure 6.

As can be seen from Figure 6, the level of industrial development, the income level, the energy structure, and the population size play driving roles in the growth of carbon emissions in China’s tertiary industry, while the energy carrying capacity, the energy intensity, the energy consumption intensity, the industrial structure, and the consumer capacity play restraining roles in the growth of carbon emissions in China’s tertiary industry. From 2004 to 2019, the growth of carbon emissions of the tertiary industry caused by driving factors was always greater than the reduction in carbon emissions of the tertiary industry caused by restraining factors, which was an important reason for the continuous growth of carbon emissions of China’s tertiary industry from 2004 to 2019.

The level of industrial development and the income level contribute the most to the growth of carbon emissions in China’s tertiary industry. From 2004 to 2019, the carbon emissions from the tertiary industry, driven by the improvement of development level and per capita income, increased from 87.7773 million tons and 66.2922 million tons in 2004 to 619.9311 million tons and 515.6615 million tons in 2019. The carbon emission growth brought by the improvement of the level of industrial development and the income level accounted for more than 92.07% of the carbon emission growth of the tertiary industry. This could be due to the China’s level of tertiary industry development from 5800 yuan/person in 2004 to 38,158 yuan/person in 2019, and the income level from 5400 yuan/person in 2004 to 30,733 yuan/person in 2019. The rapid development of the tertiary industry, people’s living standards, and consumption ability for the increase in China’s third industry’s carbon emissions plays a leading role.

Energy carrying capacity and energy intensity are the main factors inhibiting the growth of carbon emissions in the tertiary industry. From 2004 to 2019, due to the increase in the tertiary industry, the carbon emission reduction in the energy carrying capacity increased from 45.1314 million tons in 2004 to 358.2147 million tons in 2019. The reduction in the carbon emissions caused by the tertiary industry increased from 41.5686 million tons in 2004 to 174.7349 million tons in 2019. Due to the increase in energy carrying capacity and energy intensity, the reduction in carbon emission of the tertiary industry accounts for more than 69.81% of the total carbon emission reduction in the tertiary industry. However, this proportion decreased annually from 85.73% in 2004 to 69.81% in 2019. The energy consumption per unit population of China’s tertiary industry increased from 0.11 tons of standard coal/person in 2004 to 0.61 tons of standard coal/person in 2019. The increasing material demand leads to increasing energy consumption, indicating that reducing the energy consumption per unit population is the main way to achieve low-carbon emission reduction. At the same time, the energy intensity of China’s tertiary industry dropped from 0.28 tons of standard coal per 10,000 yuan in 2004 to 0.16 tons in 2019.

3.2.2. Decomposition of Carbon Emission Factors in the Tertiary Industry at the Regional Scale

In order to highlight the intensity of carbon emissions from China’s tertiary industry, taking into account the availability of data and the accuracy of calculations, 1995 is selected as the comparative baseline year for carbon emissions from the tertiary industry. The study examines the changes in carbon emissions from the tertiary industry in seven major regions of China from 2004 to 2019 and the magnitude of the impact from various influencing factors. Figure 7 highlights the change in intensity of carbon emissions in China’s tertiary industry according to the geographical regions.

From 2004 to 2019, the carbon emissions of the tertiary industry in the seven major regions of China showed an overall upward trend compared with 1995. The level of industrial development and income level are the main driving factors for the growth of carbon emissions, and energy carrying capacity is the main factor to restrain the growth of carbon emissions. Compared with 1995, the carbon emission growth of the tertiary industry was 126.3426 million tons in East China. The region with the smallest growth in carbon emissions was observed in South China, where the carbon emission increment of the tertiary industry was 31.0105 million tons. From 2004 to 2019, Northeast China experienced the highest carbon growth rate, with an annual growth rate of 18.81%. In contrast, South China had the lowest annual growth rate at 8.76%.

Further analysis of the carbon emission driving factors of the tertiary industry in each provincial administrative unit is shown in

Table 4. Table of carbon emission leading factors of the provincial tertiary industry in China from 2004 to 2019.

Year	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019
Beijing
Tianjin
Hebei
Shanxi
Inner Mongolia
Liaoning
Jilin
Heilongjiang
Shanghai
Jiangsu
Zhejiang
Anhui
Fujian
Jiangxi
Shandong
Henan
Hubei
Hunan
Guangdong
Guangxi
Hainan
Chongqing
Sichuan
Guizhou
Yunnan
Shaanxi
Gansu
Qinghai
Ningxia
Xinjiang

, where the yellow represents the energy factor, the blue represents the industry factor, and the green represents the consumption factor.

From 2004 to 2019, the leading carbon emission factors of the tertiary industry in all provinces and cities gradually changed from energy-leading to industry-leading, indicating that China’s tertiary industry lacks environmental protection while pursuing economic interests. Although the energy intensity in China is gradually declining, the speed of economic growth is relatively fast, especially in industries with higher value-added and greater energy consumption, such as transportation. As a result, the reduction in carbon emissions due to the improvement of energy efficiency is lower than the increase in carbon emissions with industrial development. This means that the main factors affecting carbon emissions in China’s tertiary industry have shifted from energy factors to industrial factors.

4. Discussion

4.1. Reliability of Carbon Emission Calculation Results in China’s Tertiary Industry

This paper relies on a carbon emission measurement model for the tertiary industry to calculate the carbon emissions and changing trends in China from 2004 to 2019. The study’s conclusions are consistent with existing research and possess certain credibility. For instance, Lu calculated the energy-related carbon emissions of China’s tertiary industry using the total carbon emission formula proposed by the IPCC. They discovered a yearly increase in carbon emissions from 2004 to 2009, which aligns with the findings of this paper [26]. Similarly, Zhu also obtained similar results in their research [35]. Additionally, Lu’s calculations of carbon emissions from the regional tertiary industry revealed the total transportation-related carbon emissions in China’s Beijing–Tianjin–Hebei region from 2000 to 2013. The results for the period from 2004 to 2013 are also comparable to those presented in this paper [36]. The aforementioned literature provides evidence supporting the consistent trends in carbon emissions within the field.

4.2. Factors Affecting the Carbon Emissions of China’s Tertiary Industry

Through the further decomposition of energy factors, industrial factors, and consumption factors, this paper finds that the level of industrial development, income level, energy structure, and population size of the province play driving roles in the growth of carbon emissions in China’s tertiary industry. Energy carrying capacity, energy intensity, energy consumption intensity, industrial structure, and consumer capacity have restraining effects on the growth of carbon emissions in China’s tertiary industry. At the same time, Lu found that the economic growth of the tertiary industry plays a driving role in the carbon emission of the tertiary industry, while the industrial structure plays an inhibitory role in the carbon emission of the tertiary industry [26], which is consistent with the results of this paper. However, the current research mainly attributes the influencing factors affecting the carbon emissions of the tertiary industry to industrial factors or energy factors. This paper found that consumption factors will also inhibit the generation of carbon emissions in the tertiary industry. For China, capital investment, expanding consumption, and foreign trade have always been the three main ways to stimulate economic growth. In the context of a volatile international environment and unstable prospects for economic development, it is particularly important to promote the improvement of consumption level for China’s economic development. Therefore, in the future development process, China must pay attention to the inhibition of consumer capacity on carbon emissions in order to achieve green development.

Of course, this article only analyzes the influencing factors from the perspectives of energy, industry, and consumption. Policy factors also have a significant impact on the carbon emissions of the tertiary industry. In 2012, China proposed the strategic decision of “vigorously promoting ecological civilization construction” and implemented a series of policy measures. China pays more attention to environmental protection rather than simply pursuing rapid economic growth. Controlling carbon emissions is also one of the policy contents. From the calculation results in this article, it can be seen that the carbon emissions of China’s tertiary industry significantly decreased in 2013–2014. This characteristic partially reflects the significant impact of policy factors on the carbon emissions of the tertiary industry. Although the influence level of policy factors cannot be further elucidated within a longer time frame due to data limitations, future studies can include policy factors as a component of the research framework for analysis.

4.3. Impact of the Tertiary Industry on China’s Carbon Emissions

China’s tertiary industry is developing rapidly. Since 2014, the output value of China’s tertiary industry has surpassed the secondary industry in GDP for the first time, and it has continued until now. In 2022, the output value of China’s tertiary industry accounted for 52.8% of GDP, which has become the pillar of China’s economic development. Although relevant studies show that the secondary industry has a higher impact on China’s carbon emissions than that of the tertiary industry, the tertiary industry has become the main source of carbon emission increase in some regions [9]. At the same time, this research finds that the carbon emissions of China’s tertiary industry have grown by more than 17% annually, which indicates rapid transformation of the service sector. Therefore, although the proportion of carbon emissions directly caused by the tertiary industry is not high [37], more attention should be paid to the management of carbon emissions in the tertiary industry due to its continuous growth. When studying China’s carbon emission policy, some scholars also proposed formulating a decoupling plan for the tertiary industry [38], which is also considered very necessary in this study.

In recent years, China has placed great importance on innovative development, coordinated development, and green development. These measures involve adjusting the industrial structure and enhancing the tertiary industry’s role in the economy. However, increasing the proportion of the tertiary industry entails more than just raising its output value. It is crucial to focus on improving the quality of economic development, ensuring sustainable and stable growth, and enhancing ecological protection. Therefore, this paper presents the following perspectives: (1) China needs to adjust its industrial structure comprehensively by promoting the upgrading of traditional industries; accelerating the development of new manufacturing sectors; implementing green management throughout the entire life cycle; establishing a green manufacturing system; and supporting the development of industries such as new generation information technology, electric vehicles, biotechnology, low-carbon technology, high-end equipment and materials, and digital creativity. (2) Efforts should be made to promote changes in the way energy is produced and utilized. China should optimize its energy structure, improve the energy efficiency, and establish a modern energy system that is clean, low-carbon, safe, and efficient. (3) Establishing a promotion mechanism that considers process, technology, energy consumption, environmental protection, quality, and safety is essential. Outdated production capacity should be eliminated to drive improvements in these areas.

5. Conclusions

The study has quantitatively analyzed the carbon emissions in China’s tertiary industry using the “energy-industry-consumption” factor decomposition model. The carbon emissions trend from 2004 to 2019 has been presented at national, provincial, and regional levels with temporal and spatial trends. The major conclusions drawn from the study are as follows:

• The contribution of the tertiary industry in carbon emissions in China has shown a gradual incremental trend from 2004 to 2019. During the study period, the tertiary industry has experienced an emission growth of 353.10%, where the transportation sector had the largest impact on China’s tertiary industry’s carbon emissions growth, accounting for over 50% of the total emissions.
• From a regional perspective, the tertiary industry in East China has the largest carbon emissions, whereas the smallest emissions were found in Southwest China. The carbon emissions growth characteristics of provincial administrative units can be classified into four types. Yunnan, Guizhou, Hunan, Anhui, Qinghai, Chongqing, Ningxia, Hubei, Heilongjiang, and Henan experienced high-speed growth. Tianjin, Shanghai, Beijing, Fujian, Hainan, Zhejiang, Jiangsu, Guangxi, Liaoning, Shaanxi, and Jilin showed characteristics of low-speed growth. Carbon emissions in Shandong, Shanxi, Hebei, Inner Mongolia, and Xinjiang exhibited fluctuating growth. Sichuan, Gansu, Jiangxi, and Guangdong showed a trend of stable growth.
• This study found significant spatial differences in carbon emissions in China’s tertiary industry, with the highest emissions in the south and lowest in the northwest. Guangdong had the highest cumulative carbon emissions from the tertiary industry, reaching 318.9811 million tons. And from the perspective of per capita carbon emissions, the clustering pattern is not significant. However, overall, it still exhibits a spatial characteristic of high density in the eastern part of China and low density in the western. The per capita carbon emissions from the tertiary industry in Shanghai and Beijing are much higher than in other regions.
• The improved decomposition model has identified the effect of tertiary industrial sectors on carbon emissions, where industrial factors and consumption factors have a positive effect while the energy factors have a negative effect. Specifically, the level of industrial development, income level, energy structure, and population size play driving roles in the growth of carbon emissions in China’s tertiary industry, and energy carrying capacity, energy intensity, energy consumption intensity, industrial structure, and consumer capacity play restraining roles in the growth of carbon emissions in China’s tertiary industry. From 2004 to 2019, the carbon emissions of the tertiary industry in the seven regions of China showed an overall upward trend compared with 1995. The level of industrial development and the income level were the main driving factors causing the growth of carbon emissions, and the energy carrying capacity was the main factor restraining the growth of carbon emissions. The leading factors of carbon emissions in the tertiary industry in all provinces and autonomous regions of China have gradually shifted from energy-leading to industry-leading.

At the same time, there are still some deficiencies in this paper. By analyzing the leading factors affecting the carbon emissions of the tertiary industry, the key elements can be grasped to a certain extent, and the government can formulate and implement corresponding policies targeted way. However, there are potential interactions between various factors. This effect and how the effect of this effect on the carbon emissions of the tertiary industry need further research.

China has developed rapidly in the tertiary industry, accounting for a high proportion of its economic contribution, and its economic manifestations are becoming more and more diverse and complex. The official statistics can relatively accurately measure carbon-emission-related research. However, with the diversity of data sources, some network data and enterprise survey data can also be applied to relevant research to improve the accuracy and accuracy of measurements.