Regional Differences in the Emission-Reduction Effect of Environmental Regulation Based on the Perspective of Embodied Carbon Spatial Transfer Formed by Inter-Regional Trade

Abstract

On the basis of the latest input–output data, this paper estimates the amount of embodied carbon emissions in inter-regional trade by constructing a multiregional input–output model to evaluate how environmental regulation stringency influences its spatial transfer. We found that environmental regulation stringency had significant positive correlation with transferring out embodied carbon emissions in trade at the national level, and a significant negative correlation with transferring in embodied carbon emissions in trade. In East and Central China, effective environmental regulation observably improves the issue of carbon emissions caused by trade, while in the western region, environmental regulation stringency had significant positive correlation with transferring in and out embodied carbon emissions in inter-regional trade. For that reason, we further use the geographically weighted regression model (GWR) to assess the spatial evolution characteristics of the intensity of environmental regulation on the transfer of embodied carbon emissions in trade; thereby, the above results are verified and show that environmental regulation has failed to play its due role.

1. Introduction

Along with the fast development of regional economies and the implementation of domestic demand policy, the exchange of goods and services between regions has become increasingly close, forming a complex trade network [1]. The formation of an inter-regional trade network promotes economic development and is accompanied by the cross-regional transfer of carbon emissions [2]. In fact, environmental regulation has the force of law, to a certain extent, to reduce greenhouse gas emissions, such as carbon dioxide, so as to improve regional environmental issues. However, due to the different understanding regarding environmental regulation between regional governments and enterprises, many differences exist in the enforcement intensity of environmental regulation [3]. Enterprises that advocate profit maximization, in order to avoid the economic loss caused by environmental regulation, often reduce the cost of regulation through a variety of channels [4,5]. For example, with the help of inter-regional trade networks, enterprises transfer highly polluting industries to areas with a lower cost of violation, that is, areas with weak environmental regulation [6]. This puts forward a new scientific problem for environmental regulation and regional management. That is to say, from the viewpoint of trade embodied carbon, studying the regional characteristics of environmental regulation and emission reduction is generally conducive to providing scientific support for the establishment of optimal emission-reduction policies.

In recent years, greenhouse gases such as carbon dioxide have brought about a series of ecoenvironmental problems that urgently need to be solved around the world. In particular, China, as the country with the maximal total carbon emissions in the world [7,8], has promised to reach the carbon peak by 2030 and faces huge pressure on emission reduction. Figure 1 shows China’s total and per capita carbon emissions from 2005 to 2018 with data compiled by the World Bank (https://data.worldbank.org/, accessed on 7 March 2022), since the latest data in the input–output table are from 2012. The figure shows that the amount of carbon emissions increased from 1.59 billion metric tons to 2.81 billion metric tons during 2005–2018 in China, and per capita carbon emissions rose from 1.22 metric tons to 2.00 metric tons. China has had increasing both total and per capita carbon emissions. Against this background, exploring the emission reduction effect of environmental regulation in various regions of China would provide notable practical support.

For a long time, most of the research results of Chinese scholars have regarded which level of intensity of environmental regulation could significantly affect carbon emissions. For example, Lu et al. [9,10] thought that there existed inverted U-shaped correlation between environmental regulation and carbon emissions by introducing a competitive equilibrium model. However, some studies showed that the impact of China’s environmental regulation in reducing carbon emissions is not obvious or not obvious in some areas. Chai et al. [11,12] believed that there are regional differences in the impact of environmental regulation on carbon emissions in China’s different regions. In addition, some studies have shown that the competition between local governments has an obvious “race to the bottom” effect on the treatment of environmental pollution [13,14,15]. In order to pursue short-term economic growth, attract foreign investment, and promote economic development and other goals, local governments often sacrifice the environment. This kind of “bottom-to-bottom competition” behavior in environmental governance weakens the implementation effect of environmental policy. In general, an indepth exploration of the effect of environmental regulations on carbon emissions provides important theoretical value for the establishment of effective emission-reduction policies. However, with the development of economic integration, trade between regions is becoming closer, which m separates the producers and consumers of products geographically and spatially, and many products are produced outside the region but provided for local consumption, which leads to embodied carbon emissions in regional trade. For the sake of clarifying embodied carbon emissions in inter-regional trade, many scholars have contributed to the study of inter-regional industrial transfer [16], carbon-emission transfer trends [17], and the migration features of embodied carbon emissions in inter-regional trade [18,19]. How does environmental regulation influence the embodied carbon transfer in trade? Does it have different effects in different regions? There is little reference to this in the existing literature. In fact, taking the transfer of trade embodied carbon emissions as a breakthrough point, it is of notable theoretical significance and practical support for regions to analyze the effect of environmental regulation with different intensity levels on emission reduction, so as to formulate more targeted coordinated emission-reduction policies. Therefore, on the basis of the latest input–output data, a multiregional input–output model is used to the calculation of trade embodied carbon. Then, the panel data model and spatial measurement model are used to study and judge the spatial differences in emission reduction to examine how to influence environmental regulation.

On the whole, the contributions of this paper are mainly twofold: (1) It is different from research on traditional environmental regulation and carbon emissions, as the embodied carbon in regional trade is calculated by regional input–output tables. Taking the spatial transfer of trade embodied carbon as the breakthrough point, this paper explores how environmental regulation affects the transfer of trade embodied carbon, hence promoting new ideas for evaluating the emission-reduction effect of environmental regulation. (2) This paper effectively combines multiregional and multisectoral input–output tables, and uses the spatial measurement method of geographically weighted regression to establish the spatial transfer trend of embodied carbon in trade. It also evaluates the characteristics of regional spatial differences in the influence of environmental regulation on embodied carbon in trade to further enrich the analysis of and research on embodied carbon in trade and its interfering factors in research methods.

The rest of this paper is organized as follows. Section 2 describes the research methods and models, and data sources. Section 3 analyzes the trade embodied carbon transfer in 2007, 2010, and 2012, and then explores environmental regulation to establish how to influence regional trade embodied carbon and its spatial differences through panel regression and geographically weighted regression. Section 4 summarizes and analyzes the research results of this paper, and gives the corresponding policy recommendations.

2. Research Methods, Models, and Data

2.1. Measuring Method

According to the UNFCCC, embodied carbon is defined as “carbon dioxide emitted during the whole process of a good’s manufacturing, from the mining of raw materials through the distribution process, to the final product provided to consumers” [20]. When it comes to measuring embodied carbon emissions in inter-regional trade, we referred to the research of Zhou and Zhong [21,22], of which the input–output table was mainly used for calculations. Under the framework of input–output analysis [23], the equation is as follows:

$$X={\left(I-A\right)}^{-1}Y$$

(1)

where A is a direct consumption coefficient matrix, element $a_{ij}=x_{ij}/x_j$ in $A$, $I$ is the identity matrix with the same order as that of $A$, matrix ${\left(I-A\right)}^{-1}$ is the Leontief inverse matrix, $X$ represents the total output column vector, and $Y$ represents the final product column vector. Then, on the basis of this formula, we can obtain the carbon emissions by multiplying the carbon emission coefficient to the left, that is:

$$T=E{\left(I-A\right)}^{-1}Y$$

(2)

where $E$ represents the carbon emission coefficient, and $T$ is the total carbon emissions.

In order to better explain the relationship among regional carbon emission transfers, on the basis of macroeconomic and matrix algebra knowledge, this paper decomposes the multiregional input–output table, which is expressed with the model as follows:

$$X_r=A_{rr}{X}^r+y_{rr}+{\displaystyle \sum }_{s\ne r}{A}^{rs}{X}^s+{\displaystyle \sum }{y}^{rs}$$

(3)

where r and s represent different regions; $A_{rr}$ represents the intermediate demand coefficient matrix of the industrial departments of the region r to the industrial departments of the province; $A^{rs}$ represents the intermediate demand coefficient matrix of the industrial departments of the region s to industrial sectors of region r; $y_{rr}$ represents the final demand produced and supplied to region r by region r; amd $y^{rs}$ represents the final demand produced and supplied to region s.

This paper only takes provinces as the research unit without industry segmentation, and summarizes the carbon emissions of all industrial sectors in every province. Therefore, on the basis of Equation (3), the equation for calculating carbon emissions of region r produced and used for consumption in the province is as follows:

$$T_{rr}=E_r{\left(I-A^{rr}\right)}^{-1}{y}^{rr}$$

(4)

where $T_{rr}$ represents the emissions of province r for production and consumption in the province, and $E_r$ is the carbon emission coefficient of province r. In addition, the calculation formula of carbon emissions $T_{rs}$ used for production and consumption by province r in province s is shown in Equation (5):

$$T_{rs}=E_r{\left(I-A^{rr}\right)}^{-1}\left({{\displaystyle \sum }}_{s\ne r}{A}^{rs}{X}^s+{{\displaystyle \sum }}_{s\ne r}{y}^{rs}\right)$$

(5)

Furthermore, this paper gives the accounting equation of embodied carbon emissions $T_{sr}$ used for production and consumption by province s in province r, as shown in Equation (6):

$$T_{sr}=E_r{\left(I-A^{ss}\right)}^{-1}\left({{\displaystyle \sum }}_{r\ne s}{A}^{sr}{X}^r+{{\displaystyle \sum }}_{r\ne s}{y}^{sr}\right)$$

(6)

In order to calculate the transfer-out or -in amount of trade embodied carbon in province r, the difference was calculated, and the equation is as follows:

$$\begin{array}{c}{T}_{rsr}=T_{rs}-T_{sr}\end{array}$$

(7)

where $T_{rsr}$ represents the trade embodied carbon emissions transfer in or out from province r to province s. If $T_{rsr}$ is greater than 0, $T_{rsr}$ is the amount of embodied carbon emissions in the trade transfer out from province s to province r. If $T_{rsr}$ is less than 0, $T_{rsr}$ is the transferring in of trade embodied carbon from province r to province s (the transfer-in and transfer-out amounts of embodied carbon emissions in trade studied in this paper are both net transfer in and out). Lastly, the amount of embodied carbon in the trade transfer out of province r is expressed as the sum of trade embodied carbon transfer out of province r to other provinces, and the amount of trade embodied carbon transfer in province r is expressed as the sum of embodied in trade carbon emissions transfer in of province r to other provinces.

2.2. Theoretical Model Construction

For verifying the influence of environmental regulation on the transfer of trade embodied carbon, it is necessary to verify the relationship of environmental regulation and the transfer of embodied carbon emissions in trade and the amount of trade embodied carbon. On the basis of the relevant studies of Chen et al. [24], and Guo et al. [25], the control variables of the econometric model are determined comprehensively. The econometric model was constructed as follows:

$$CI={\alpha }_0+{\alpha }_{1}E{R}_{i,t}+{\alpha }_2{X}_{i,t}+{\mu }_i+{\epsilon }_{i,t}$$

(8)

where $CI$ represents the transfer out (in) of embodied carbon in inter-regional trade (ton); $ER$ is environmental regulation in each region, which is expressed by the ratio of regional environmental governance investment to the average of national governance investment [26,27]; $X$ represents the set of control variables. This paper mainly selected four control variables, namely, the level of economic development, population density, domestic technological progress, and foreign direct investment [28]. The level of economic development is expressed in terms of the per capita GDP (CNY) of the area because data used in this paper are cross-sectional, and the regional GDP was calculated at the current year’s price. Population size is expressed in terms of the area occupied by the resident population at the end of the year (person/square kilometers). The total number of patents granted in the region in that year represents domestic technological progress [29]. In addition, the amount of foreign direct investment represents foreign direct investment (FDI) [30]. $\alpha$ is the parameter to be estimated; ${\mu }_{i }$ represents the unobservable regional effect; $\epsilon$ is other interference factors; and subscripts $i$ and $t$ represent region and time, respectively.

Table 1. Variable description and index design.

Category	Symbol	Variable Name	Representative Index
Dependent variable	CI	Embodied carbon in trade transfer in (out)
Independent variable	ER	Intensity of environmental regulation	Regional environmental governance investment/national average governance investment
Control variable	EC	Economic development level	Regional gross domestic product/total population
	PS	Population density	The number of the resident population at the end of the year/the occupied area
	DTP	Domestic technological progress	Total amount of patent authorization
	FDI	Foreign direct investment	Amount of foreign direct investment

describes the specific variables.

The traditional regression model is mostly used to describe the influence of multiple variables on one variable, which has some limitations in dealing with the spatial correlation of independent variables. Therefore, to effectively reflect the spatial difference characteristics of the effect of environmental regulation on the transfer of embodied carbon in inter-regional trade, the geographically weighted regression model (GWR) is used for econometric analysis. The specific model structure is shown in Equation (9):

$$y_i={\beta }_0\left(u_i,v_i\right)+{\displaystyle \sum }_{k-1}^p{\beta }_k\left(u_i,v_i\right)x_{ik}+{\epsilon }_i$$

(9)

As geographically weighted regression is based on cross-sectional data, for the sake of avoiding the contingency and difference caused by the analysis of cross-sectional data, referring to the practices of Zhou and Chen et al. [31,32], this paper takes the mean values of the econometric data of 2007, 2010, and 2012. In Equation (9), $y_i$ represents the three-year average value of the transfer out (in) of trade embodied carbon in province i, $\left(u_i, v_i\right)$ and ${\beta }_k\left(u_i,v_i\right)$ are the coordinate and the $k$ regression coefficient of sample point i, respectively, and ${\epsilon }_i$ is the random error of sample point i.

2.3. Data Sources

The panel data of 30 provinces, municipalities, and autonomous regions in China are used for quantitative inspection (because the data of Tibet, Hong Kong, Macao, and Taiwan are not easily available, these regions are not included in the study). The data needed in this process include the two following aspects. When calculating the amount of trade embodied carbon using input–output tables, referring to Chen et al. [33,34], we sorted and consolidated the industrial sectors in the input–output table into 30 sectors, as shown in

Table 2. Classification of industrial sectors.

ID	Industries	ID	Industries
1	Agriculture	16	General and special equipment manufacturing industry
2	Coal mining and washing industry	17	Transportation equipment manufacturing industry
3	Oil and gas extraction industry	18	Electrical machinery and the equipment manufacturing industry
4	Metal mining and dressing industry	19	Communication equipment, computers, and other electronic equipment manufacturing industry
5	Nonmetallic minerals and other minerals’ mining and selection industry	20	Instrument, cultural, and office machinery industry
6	Food manufacturing and tobacco processing industry	21	Other manufacturing industry
7	Textile industry	22	Production and supply of electric power and hot power industry
8	Textiles, clothing, shoes, hats, leather, feathers, and their products industry	23	Production and supply of gas and water industry
9	Wood processing and furniture manufacturing industry	24	Construction industry
10	Paper printing, cultural and educational sports products manufacturing industry	25	Transportation and warehousing industry
11	Petroleum processing, coking, and nuclear fuel processing industry	26	Wholesale and retail industry
12	Chemical industry	27	Lodging and catering industry
13	Nonmetallic mineral products industry	28	Leasing and business services industry
14	Metal smelting and rolling processing industry	29	Research and experimental development industry
15	Metal product industry	30	Other services industry

. In addition, the required energy consumption and energy balance table of different provinces (including coal, gasoline, coke, kerosene, fuel oil, diesel, natural gas) were derived from the China Energy Statistical Yearbook of 2008, 2011, and 2013. In light of the final energy consumption of industries in the energy balance table, the energy consumption of 30 departments in different regions was established; energy emission factors were obtained from IPCC [35]. In empirical analysis, data on environmental pollution governance investment and the average value of national governance investment in various provinces and cities came from the China Environmental Statistical Yearbook; data on the resident population and regional GDP of each province and city at the end of the year came from the China Statistical Yearbook of 2008, 2011 and 2013; data on the amount of foreign direct investment and the number of granted patents were from the Statistical Yearbooks of provinces and cities of 2008, 2011, and 2013 (data source website of each statistical yearbook: https://data.cnki.net/, accessed on 7 March 2022).

3. Results and Analysis

3.1. Transfer of Embodied Carbon Emissions in Inter-Regional Trade

To more intuitively show the transfer of trade embodied carbon, the three-year transfer out (in) in the corresponding area of the map of China is projected in this paper, and the results are displayed in Figure 2.

Figure 2 reveals the spatial pattern of trade embodied carbon in 2007, 2010, and 2012. In these three years, the spatial distribution of trade embodied carbon was highly consistent with regional economic development stage. The trade embodied carbon in developed eastern coastal areas was generally great than that in the two other regions, and gradually increased from west to east; there was a spatial locking phenomenon. From a regional perspective, the transfer of trade embodied carbon is primarily concentrated in underdeveloped areas in the west and gradually increased from east to west, and there was a great disparity between the east and the west. Meanwhile, the trade embodied carbon emissions in Heilongjiang, Jilin, and other regions were more than 1.1 million tons, which is several or even tens of times that of the more sophisticated eastern regions, for instance, Guangdong, Shanghai, and Jiangsu. Whether in terms of time or space, there were wide differences in the transfer out (in) of trade embodied carbon among different areas in China, especially between east and west. Is the reason for this pattern distribution, in addition to the gap in economic development, related to environmental regulation? To answer this question, the spatial evolution characteristics of trade embodied carbon from the emission-reduction effect of environmental regulation are discussed in this paper.

To further clarify the correlation of the transfer of trade embodied carbon and the intensity of regional environmental regulation, the model was divided into two cases. First, on the national level, the influence mechanism between the intensity of environmental regulation and the transfer out (in) of trade embodied carbon is analyzed; second, according to economic development stage, we divided China into eastern, central, and western economic zones. On the basis of the regional division, we studied the influence mechanism between environmental regulation and the transfer out (in) of embodied carbon in trade.

3.2. Empirical Results

The estimated results (

Table 3. Regression results of environmental regulations and embodied carbon emissions in trade.

Independent Variable	Transfer Out of Embodied Carbon Emissions in Trade	Transfer In of Embodied Carbon Emissions in Trade
ER	0.396 ***	−0.226 *
	(5.319)	(−1.955)
EC	−0.029	−0.475 ***
	(−0.353)	(−3.853)
PS	0.478 ***	−0.488 ***
	(6.475)	(−4.192)
DTP	0.244 ***	0.52 ***
	(2.633)	(2.903)
FDI	0.057	0.323 **
	(0.52)	(1.828)

Note: the value of t is in parentheses; *, **, and *** indicate significance at the levels of 10%, 5%, and 1%, respectively.

) first show that, on the national level, environmental regulation and the transfer out of embodied carbon in trade had a significant positive correlation, which indicates that stricter environmental regulation leads to a higher transfer of trade embodied carbon. It also reveals that stricter environmental regulation advances the transfer of embodied carbon in inter-regional trade. Meanwhile, transferring out trade embodied carbon had a significant positive correlation with population density and domestic technological progress.

Second, the environmental regulation and the transfer in of embodied carbon in inter-regional trade had a significant negative correlation, indicating that effective environmental regulation can restrain the transfer in of trade embodied carbon. In addition, the transfer-in amount of trade embodied carbon had a significant negative correlation with economic development stage and population density, which showed no difference with the results of the spatial distribution map of carbon emissions (as shown in Figure 2), and that the level of economic development was higher; thus, the transfer-in amount of trade embodied carbon emissions was lower, namely, the embodied carbon emissions in inter-regional trade tend to flow to areas with a lower level of economic growth. The transfer-in amount of trade embodied carbon also had a significant positive correlation with domestic scientific and technological progress and foreign investment, which may have been due to the fact that underdeveloped areas tend to attract more foreign investment by reducing requirements in order to accelerate economic development, which leads to an increase in the transferring in of embodied carbon emissions in inter-regional trade.

To further explore the influence of regional characteristics, the 30 provinces were divided into 3 regions according to the characteristics of economic growth, namely, central, eastern, and western. The results of

Table 4. Regression results.

Independent Variable	Eastern Region		Central Region		Western Region
	Embodied Carbon Emissions in Trade Transfer out	Embodied Carbon Emissions in Trade Transfer in	Embodied Carbon Emissions in Trade Transfer out	Embodied Carbon Emissions in Trade Transfer in	Embodied Carbon Emissions in Trade Transfer out	Embodied Carbon Emissions in Trade Transfer in
ER	0.411 ***	−0.067	0.297 *	−0.393 *	0.559 *	0.35 **
	(3.168)	(−0.33)	(1.838)	(−1.985)	(2.552)	(1.973)
EC	−0.116	−0.188	0.081	−0.878 **	−0.351 ***	−0.365
	(−0.745)	(−0.778)	(0.363)	(−2.235)	(−1.638)	(−2.825)
PS	0.544 ***	−0.263	0.965 ***	−0.484	0.27 ***	−0.414
	(3.89)	(−1.394)	(5.035)	(−1.574)	(1.22)	(−2.817)
DTP	0.289 *	0.221	−0.27	1.833 ***	0.578 ***	0.804
	(1.833)	(0.697)	(−1.118)	(4.059)	(1.193)	(4.881)
FDI	0.036	−0.277	−0.004	−0.68 **	−0.542	−0.107
	(0.211)	(−0.899)	(−0.017)	(−2.616)	(−1.126)	(−0.546)

Note: the value of t is in parentheses; *, **, and *** indicate significance at the levels of 10%, 5%, and 1%, respectively.

show that the relationship between environmental regulation and the transferring out of trade embodied carbon was positive in eastern China, which reveals that strong environmental regulation in this region can efficiently advance transferring out trade embodied carbon. Meanwhile, the intensity of environmental regulation in the eastern region had a negative relationship with the transfer-in amount of trade embodied carbon, but the negative correlation was not significant, which may have been due to the fact that the eastern part of China is more developed. Figure 2 shows that this region is the discharge place for trade embodied carbon and not the transfer-in place, and the transfer-out amount of trade embodied carbon was much greater than that of transfer-in carbon. Therefore, environmental regulation plays a notable role in advancing transferring out embodied carbon emissions in inter-regional trade, but it is not obvious in restraining the transfer in of trade embodied carbon from other provinces or cities.

Environmental regulation in the central area had a significant positive correlation with the transfer-out amount of trade embodied carbon, and a significant negative correlation with its transfer-in amount. This reveals that the effective environmental regulation in central area not only promotes the transfer out of embodied carbon in trade in this area, but also suppresses it transferring in from other regions to this region.

Environmental regulation in the western area had a significant positive relationship with the transfer-out and -in volumes of trade embodied carbon. This reveals that the stricter environmental regulation in the western area could promote not only the transfer out of trade embodied carbon, but also the transfer in of embodied carbon in inter-regional trade in other regions, which may have mainly been due to the fact that the western region of China is still at an insufficient economic development stage, and various costs are lower than they are in other regions. In a comprehensive consideration, enterprises are more inclined to move factories to the western region. In addition, competition among governments has caused a “race to the bottom” effect, which leads to environmental degradation in order to accelerate rapid economic growth. Therefore, even if the region strengthens environmental regulation, the transfer-in amount of trade embodied carbon still increases rather than decreases. To some extent, this also shows that the role of environmental regulation in restraining the transfer in of embodied carbon in external trade is not obvious in this area. Therefore, the western area of China should look for more effective emission-reduction policies.

3.3. Robustness Test and Analysis

Furthermore, through the geographically weighted regression model, this paper discusses the characteristics of inter-regional spatial differentiation, and analyzes in detail the environmental regulation and spatial transfer characteristics of trade embodied carbon in different regions of the underdeveloped western region to verify the robustness of the results. On the basis of the results of geographically weighted regression with environmental regulation stringency as the explanatory variables, regional development level and population density as the control variables, between explanatory variables and most of the control variables had a significant correlation, and the symbol of the coefficient was consistent with the expectation. On the basis of the estimated results of the geographically weighted regression model, the following further explores the spatial differences of the influence between regional environmental regulation and the transfer of embodied carbon in inter-regional trade.

The results reveal that environmental regulation had a positive impact when it came to the transfer out of trade embodied carbon in each region, and the regression coefficients of environmental regulation were all negative when it came to the transfer-in amount of trade embodied carbon emissions in the eastern and central regions. However, the regression coefficients between the intensity of environmental regulation and the transfer-in amount of embodied carbon in inter-regional trade in the western region were mostly positive. On the whole, the impact of environmental regulation in different provinces on the transfer-out and transfer-in amounts of trade embodied carbon shows a certain gradient distribution in space. Figure 3a shows that environmental regulation in the eastern area had a more significant effect on the transfer out of trade embodied carbon. Figure 3b shows that the change in environmental regulation to the transfer of trade embodied carbon in eastern area was smaller. This reveals that the stricter the environmental regulation in the eastern and central regions of China was, the higher the transfer-out amount of trade embodied carbon; its transfer-in amounts from other regions were also relatively low. This also shows that the regional trade embodied carbon with higher environmental regulations tended to flow to areas where environmental regulations were lower. Nevertheless, the coefficients of environmental regulation and the transfer out (in) of trade embodied carbon in many areas of West China was positive, which indicates that strengthening environmental regulation can promote the transfer out of embodied carbon emissions in trade in these areas, but still cannot restrain the transferring in of trade embodied carbon emissions from other provinces or cities.

There was a spatiotemporal difference in the impact of the level of economic development on the transfer of trade embodied carbon. When it came to transferring out trade embodied carbon, the per capita GDP in trade had a positive impact, and when it came to transferring in trade embodied carbon, the per capita GDP has a negative impact. Overall, the impact of the level of economic development on the transfer out of trade embodied carbon was mainly concentrated in the eastern part of China (Figure 3c), and the influence on transferring in trade embodied carbon emissions was mainly concentrated in the western region (Figure 3d). This reveals that the eastern region, which economically grew, was the main transfer-out place of embodied carbon emissions in inter-regional trade and mainly flowed to the western region where the resources is relatively scarce.

There was a spatiotemporal difference in the impact of population density on the transfer of trade embodied carbon. The regression coefficient between population density and the transferring out of trade embodied carbon was positive. In addition, the regression coefficient between population density and the transferring in of trade embodied carbon was negative. Overall, the impact of population density on transferring out embodied carbon in trade was more concentrated in the western region (Figure 3e); on the transfer in of trade embodied carbon, it was concentrated in the central and eastern areas (Figure 3f). This shows that the higher the population density is, the higher the transfer-out amounts of trade embodied carbon are, and the lower the corresponding transfer-in amounts of trade embodied carbon are.

There was a spatiotemporal difference in the impact of domestic technological progress on the transfer of trade embodied carbon. Regarding transferring out trade embodied carbon, the regression coefficient of domestic technological progress was positive (the coefficient in Figure 3g); it was negative regarding the transfer in of trade embodied carbon (Figure 3h). On the whole, the impact of technological progress on embodied carbon emissions in inter-regional trade was concentrated in the eastern coastal regions. This shows that technological progress can effectively promote the transfer out of embodied carbon emissions in inter-regional trade while accordingly restraining the transfer in of trade embodied carbon.

Lastly, there was a spatiotemporal difference of the impact of foreign direct investment on the transfer of trade embodied carbon. On the basis of empirical results, the regression coefficients between the amount of foreign direct investment and the transfer-out or -in amount of embodied carbon in trade was negative (Figure 3j). On the whole, the influence of FDI on embodied carbon in inter-regional trade was mostly concentrated in the western region. This shows that the use of foreign investment in the western region had a more notable impact on embodied carbon emissions in inter-regional trade.

4. Conclusions and Analysis

To achieve the purpose of this research, the multiregional input–output model was used to establish the transfer out (in) of trade embodied carbon in each region, and environmental regulation strength was selected as the core independent variable by verifying the relationship between environmental regulation strength and the transfer out (in) of embodied carbon emissions’ inter-regional trade in today’s increasingly close trade network. We clarified the influence mechanism of environmental regulation strength and the transfer of trade embodied carbon to explore the emission-reduction effects of the strength of different environmental regulations in different areas. The results are as follows:

Stronger environmental regulation can promote the transfer out of embodied carbon emissions in inter-regional trade, and effective environmental regulation can also restrain the transfer in of external trade embodied carbon. According to economic development stage, the impact of environmental regulation in the eastern and central regions of China on the transferring out and in of trade embodied carbon is more obvious. Not only did environmental regulation in the western region have a significant positive correlation with the transfer-out amount of embodied carbon in inter-regional trade, but it also had a significant positive correlation with the transfer-in amount. The reason may be that the economy of the western area was less developed than that of the central and eastern areas, as it belongs to a group of polluting enterprises in the trade network formed between regions. At the same time, for the sake of developing the economy, the western government does not hesitate to increase environmental pollution, so even if environmental regulation is strengthened, the transfer-in amount of embodied carbon emissions in trade still increases rather than decreases.

Furthermore, to verify the robustness of the results, a geographically weighted regression model was constructed to explore the influence of environmental regulation on the transfer of trade embodied carbon. The results reveal that environmental regulation stringency in the eastern and central regions of China promotes the transfer of trade embodied carbon, while in the western area, due to the limitations of its own conditions, even if environmental regulation is strengthened, the transfer-in amount of trade embodied carbon increases.

Through the above research results, this paper holds that effective environmental regulation can facilitate the transfer out of embodied carbon in a region’s trade and restrain the transfer in of trade embodied carbon, so as to improve the environmental situation of the region. However, not all regional environmental regulations can contribute to reducing carbon emissions and promoting environmental protection, and the influence of environmental regulation in the relatively less developed western region is not obvious. Therefore, more effective measures should be sought to achieve emission-reduction target. At the same time, due to differences in environmental regulation at different regions, regions with stricter environmental regulation tend to transfer the trade embodied carbon to the areas where the intensity of environmental regulation is weaker, which can easily lead to the total carbon emissions not being effectively suppressed.

Lastly, although this paper discussed the regional emission-reduction effect of environmental regulation on the basis of embodied carbon transfer in trade, which can provide policy makers with some targeted regional carbon emission reduction countermeasures, there are still some deficiencies and limitations that are worthy of further study in the future. First, new energy consumption is an important factor affecting carbon emissions, so it needs to be further explored to improve the proportion of new energy, and optimize the structure of energy consumption to further explore the influence of environmental regulation intensity on embodied carbon transfer in trade. Second, in the context of the new development pattern of domestic and international double circulation, the steady economic growth, the assignment of carbon emission-reduction responsibilities, and the division of trade embodied carbon transfer and regional emission reduction responsibilities based on the degree of internal cycle participation all need to be further studied to help in providing scientific support for the establishment of differentiated emission reduction policies in instances of unbalanced and insufficient regional development.