Functional Upgrading of Value Chains and the Carbon Emissions Reduction Embodied in China’s Exports: From the Perspective of the Improvement in the FDI Quality
-
Fuzhong Chen
Abstract
Improving the quality of foreign investment to help functional upgrading of the value chain is an important starting point for China’s manufacturing industry to achieve low-carbon and high-quality development. Based on the data of the functional activities of each manufacturing greenfield investment project in China from 2003 to 2018, this paper constructs specialization index of the upstream and downstream functional division of labor of value chains, and theoretically explains and empirically tests the impact and mechanism of the functional upgrading of the value chain driven by the improvement of foreign investment quality on the embodied emissions in China’s export trade. The results show that the functional upgrading of the value chain can significantly reduce the embodied emissions in China’s export trade. The mechanism analysis shows that the functional upgrading of the value chain driven by the improvement of the quality of foreign investment mainly plays a role in reducing the embodied emissions in China’s export trade through the optimization effect of factor structure, the effect of human capital accumulation and the effect of service demand creation. Extended analysis finds that high-quality foreign investment engaged in upstream functional activities and inflows into eastern China have more obvious inhibitory effects on the embodied emissions in export trade. At the same time, the functional upgrading of the value chain is more conducive to reducing the embodied emissions in export trade in pollution-intensive manufacturing industries. This paper verifies the “pollution halo” effect of high-quality foreign investment, and provides strong support for China to attract and utilize foreign investment more vigorously and promote high-level opening up under the new situation.
1 Introduction
It is a typical feature of the division of labor in the global value chains (GVCs) that countries engage in related and orderly functional activities. Relying on the natural connection with the GVCs network of multinational corporations, foreign capital evolves into a bridge to promote China’s deep integration into the global production and supply chain, and plays a key role in shaping China’s functional division of labor in integrating into GVCs. In fact, the functional division of labor in the value chain driven by foreign investment is not only related to profit distribution, but also has a profound impact on environmental pollution control. However, for a long time, China has mainly integrated into GVCs with the advantage of low factor cost, and has been “locked in the low-end” in the production and manufacturing function, and the real profit of trade is seriously unbalanced with environmental pollution. In terms of environmental costs, the embodied emissions in export trade reveals the carbon emissions caused by the production of traded goods in exporting countries to meet the consumption of other countries, and depicts the transfer of trade carbon emissions caused by the geographical segmentation of production and consumption in the context of GVCs embeddedness, reflecting the emission reduction burden of countries under the producer responsibility system (Zhao et al., 2014; Meng et al., 2018). Data shows that China is the country with the largest carbon emissions in the world, with more than 40% of carbon emissions coming from manufacturing, and 20%–30% of carbon emissions being implicit in exports (Zhu et al., 2023).
Considering that foreign investment is an important part of China’s domestic production network, there are a lot of empirical explorations in the existing literature on the environmental effects of foreign investment introduction on China at the regional, industrial and enterprise levels. Due to the different data samples, measurement methods, and environmental pollution measurement indicators used, there are two opposing theoretical views in the final conclusion. On the one hand, the “pollution haven” hypothesis argues that the relatively lax environmental regulatory policies of developing countries give developed countries an incentive to relocate polluting-intensive industries and production activities to them, ultimately harming the environment of host countries (Chichilnisky, 1994; Copeland and Taylor, 1994). On the other hand, the “pollution halo” hypothesis states that developed country multinationals tend to enforce strict environmental standards, leading to greener production skills, and that the resulting technology spillovers can help improve the environmental welfare of host countries (Antweiler et al., 2001).
As the “factory of the world”, China produces a large number of intermediate and final goods in the global division of labor network to meet external demand, which to a certain extent accelerates the transfer of foreign investment in energy-intensive production links, and the resulting carbon emissions from trade cannot be ignored. In view of this, many studies use input-output models or value-added trade accounting systems to measure the embodied carbon level of China’s bilateral and multilateral trade, and tried to trace the reasons for the cumulative changes in the embodied carbon emissions in China’s export trade in the context of GVCs embeddedness. Lv et al. (2019) argue that the deepening of GVCs embeddedness will gradually expand the embodied emissions in China’s export trade due to the rapid growth of intermediate goods exports, while Lv and Lv (2019) emphasize that the embodied emissions in export trade can be reduced when GVCs are integrated in a forward-embedded manner. However, the use of export value-added decomposition to reflect the embedding characteristics of GVCs mainly reveals the vertical specialization characteristics of the value chain, but cannot reflect the functional specialization characteristics within the industry (Timmer et al., 2019). The reason for this is that the expansion of gross value-added may come from the accumulation of manufacturing activities or the adjustment of the structure of industries such as primary industries (Vries et al., 2019), but does not directly indicate that China is moving along both ends of the “smile curve” (Wang et al., 2020). In fact, the current demand for upgrading China’s industrial chain is no longer limited to the drive of high-tech industries, but also needs to achieve functional upgrading within the industrial chain. Therefore, the identification of GVCs participation characteristics should not ignore the functional specialization information of the value chain.
In addition to directly accounting for the proportion of domestic value-added caused by undertaking different functional parts (Timmer et al., 2019; Wang et al., 2020; Kordalska and Olczyk, 2023), the country’s functional division of labor in GVCs can also be indirectly portrayed based on the functional activities mainly engaged by foreign investment in the local industrial chain (Stöllinger, 2021; Xiong and Luo, 2023). The reason is that foreign capital, as the governor of the global intra-product division of labor system, not only introduces capital flow into China’s production system, but also drives the accumulation of management experience and advanced technology to China, which plays a key role in the cultivation of China’s integration into the division of labor of GVCs and the extension of the domestic industrial chain (Jiang and Meng, 2021). Therefore, China’s functional role in GVCs is an important reference for evaluating the quality of foreign investment in China. Specifically, the more active foreign investment in the high value-added links of the upstream and downstream of the industrial chain, the more obvious the upgrading of China’s value chain function (Stöllinger, 2021).
At present, the global economic recovery is slow, the localization of the value chain layout is strong, and cross-border investment is facing uncertainties and challenges. Since 2017, China has been “subtracting” for the negative list of foreign investment access for five consecutive years. The 2024 Government Work Report proposes to “completely abolish restrictions on foreign investment in the manufacturing sector”, sending a strong signal that China is firmly promoting high-level opening-up. However, while ensuring the reasonable growth of foreign capital utilization, it is particularly important to use high-quality foreign capital to occupy a more favorable position in the functional division of GVCs. According to the “smile curve”, compared with the manufacturing link, the ability of upstream and downstream functional activities in the value chain to create environmentally friendly added value is more prominent. So, can the functional upgrading of the extension to the mid-to-high-end value chain activities become an ideal path to achieve the low-carbon and high-quality development of China’s manufacturing industry? In order to clarify the above questions, this paper turns the research focus to the quality of foreign investment, theoretically explores and empirically examines the impact of value chain function upgrading on the embodied emissions in China’s export trade and its mechanism, so as to extend the analytical perspective of foreign investment quality on the functional upgrade of value chain, and also provides strong empirical support for its high-end and green development of manufacturing industry and the achievement of the carbon peak and carbon neutrality targets.
The possible marginal contributions of this paper are mainly reflected in the following aspects: First, from the perspective of research, it focuses on the shaping role of foreign investment quality in the functional division of labor in the value chain. In this paper, we construct a specialization index of the upstream and downstream functions of the value chain to reveal the trend of functional upgrading of the value chain and examine its impact on the embodied emissions in China’s manufacturing export trade. Second, in terms of mechanism identification, this paper incorporates the functional division of labor factors of the value chain into the decomposition framework of the embodied carbon emission effect of trade, and comprehensively considers the impact mechanism of the functional upgrading of the value chain on the embodied carbon emission reduction of export trade from the perspectives of factor input, technology spillover and final output structure. Thirdly, in terms of extended analysis, this paper confirms the heterogeneity of the carbon characteristics of greenfield investment and export trade, which is conducive to a more comprehensive analysis of the impact of value chain function upgrading driven by changes in foreign investment structure on China’s environment.
2 Theoretical Model and Mechanism Analysis
2.1 Theoretical Model
Based on the trade and environmental pollution analysis model adopted by Antweiler et al. (2001) and Lv et al. (2019), this paper expands the framework of the decomposition of the embodied carbon emission effect of trade to include the influence of the functional division of labor in the value chain, so as to clarify the relationship between the functional upgrading of the value chain and the embodied emissions in China’s export trade.
For the sake of simplifying the analysis, it is assumed that an open economy satisfies the following basic assumptions: (1) Only two countries, a and b, engage in production activities under the GVCs division of labor, resulting in import and export trade. (2) Two traditional factors of production of capital K and labor L and the factor of productive services S are used to produce two kinds of products X and Y, where X is a pollution-intensive commodity with more energy consumption and carbon emissions, Y is a clean-intensive service product that does not produce any environmental pollution. (3) Focusing only on the embodied carbon emission effects of trade and ignoring other possible environmental effects. (4) Constant returns to scale is assumed. The production cost functions of the two products are CX(w,r,v) and CY(w,r,v), where w,r and v represent the prices of labor, capital, and service factors, respectively, and the prices of commodities are PX and PY, respectively. (5) The market is in a state of perfect competition.
Assuming that the potential output of product X in country a is f(KX,LX,SX), KX, LX, SX are the capital, labor, and service factor inputs required to produce product X, respectively. Among them, the proportion of ϑ of product X is consumed in the country a, then the export output of product X and the corresponding embodied carbon emissions C in export trade in country a are as follows:
φ(ϑ) is the carbon emission function related to ϑ, and generally speaking, the larger the export scale of the product, the greater the resulting carbon emissions. However, since the optimization of production processes driven by technological progress can help curb the growth of embodied carbon emissions in export trade, this paper sets φ(ϑ) as the following formula:
T stands for technical level, φ (ϑ)′ < 0, φ (ϑ)″ > 0, parameter β ∈ (0,1). Combining Equation (3) and Equation (2) gets:
Combining Equation (1) and Equation (4), it can be deduced that the export output of product X is:
TC is the level of embodied carbon emissions in export trade that takes into account the technical conditions of production. Further, considering the factor of functional division of GVCs driven by foreign investment, we know that the main activities engaged in by multinational corporations in the host country’s industrial chain are potential indicators to characterize the position of GVCs functional division in the host country (Stöllinger, 2021). Therefore, when more high-quality foreign investment engaged in upstream and downstream activities enters the production chain of product x in country a, country a is in a relatively stronger link in the value-added capacity of the product division of labor.
Specifically, ρ(f) is used to reflect the ratio of relative functional specialization of the value chain driven by foreign investment f. In the production process of product x, the larger its value, the higher the degree of specialization of the upstream and downstream functional activities of the industrial chain undertaken by country a. At the same time, ω[ρ (f)] is constructed to reflect the impact of the functional division of GVCs on the embodied emissions in export trade, and ω′[ρ(f)] < 0, that is, engaging in productive services industry at both ends tends to produce lower carbon emission intensity (Huang and Xie, 2019). Taking ω[ρ (f) into the determinant of embodied carbon emissions in export trade:
From Equation (6), the actual export output of product X after considering the functional division of labor of the value chain can be deduced:
Suppose that the carbon tax rate levied by the home country (country a) and the carbon tariff rate imposed by the importing country (country b) for product X are μ1 and μ2, respectively, and the trade transportation cost and friction cost are not considered. According to the cost minimization criterion, enterprises need to make the optimal decision on the embodied carbon emissions in production output and export trade at the lowest production and environmental costs, that is:
Construct the Lagrangian function, and take the derivative of the Lagarangian function with respect to the embodied carbon emission C and the output f(KX,LX,SX) of product X respectively to obtain the first-order conditions:
λ is the Lagrange multiplier. Dividing the two equations of Equation (9) obtains the conditional formula for the company’s pursuit of cost minimization:
Under the assumption of the perfectly competitive market, the profit of the enterprise from the production of product X is zero:
Substituting Equation (10) into Equation (11), the actual output of product X is as follows:
Then, the carbon emission function per unit of production is expressed as:
Therefore, the embodied carbon emission function of export trade in Equation (6) is rewritten as:
Among them, B = PXX + PYY indicates the scale of the country’s exports, and S = PXX/(PXX + PYY) indicates the proportion of product X exports in the country’s total exports. Further, substituting Equation (13) into Equation (14) obatins:
Finally, take logarithmas in both sides of Equation (15) at the same time:
Among them, ln (βPX) is a constant. After examining the impact of the functional division of labor in the value chain shaped by foreign investment on the embodied emissions in China’s export trade, it is found that
Hypothesis 1: The functional upgrading of the value chain driven by the improvement of foreign investment quality can help reduce the embodied emissions in China’s export trade.
2.2 Mechanism Analysis
2.2.1 Factor Structure Optimization Effect
The functional upgrading of the value chain driven by high-quality foreign investment will introduce more foreign productive service factors, which can replace some of the more pollution-intensive physical factor inputs (Rothenberg, 2007), optimize the factor input structure of enterprises, and reduce energy consumption (Zhu et al., 2020). On the one hand, the diversification of the supply of service elements enables manufacturing firms to outsource services to more efficient professional service providers, improve the quality of inputs, and increase productivity levels (Arnold et al., 2011). On the other hand, the extension of downstream marketing and after-sales services can help improve the production efficiency of enterprises in response to foreign target markets, reduce redundant inventory, and reduce the embodied carbon emissions in export trade.
Hypothesis 2a: From the perspective of factor inputs, the functional upgrading of the value chain driven by the improvement of foreign investment quality provides more advanced service factors for the development of manufacturing industry, which in turn can reduce the embodied emissions in China’s export trade.
2.2.2 Human Capital Accumulation Effect
According to the theory of endogenous economic growth, human capital is a key factor in driving technological progress. Upgrading the value chain driven by high-quality foreign investment will help cultivate high-level local human capital, boost technology spillovers, improve energy efficiency, and contribute to the reduction of embodied emissions in export trade. In terms of technology spillovers, foreign enterprises that undertake high-end functional links often hire high-quality human resources locally and train them in production skills, management concepts and green environmental protection awareness, which can help local employees directly access technology spillovers. And when these high-level human capital are self-employed or transferred to work in other affiliated enterprises, technology diffusion and spread can be accelerated (Cole et al., 2007). In terms of energy use, human capital can improve energy utilization equipment, optimize pollution control methods, promote the transformation and application of green technologies, thereby optimizing energy use efficiency and reducing embodied carbon emissions in export trade. Based on this, this paper proposes:
Hypothesis 2b: From the perspective of technology spillovers, the functional upgrading of the value chain driven by the improvement of foreign investment quality accumulates more high-level human capital for the development of manufacturing industry, which in turn can reduce the embodied emissions in China’s export trade.
2.2.3 Service Demand Creation Effect
Foreign enterprises engaged in high-end functional activities in the manufacturing industry in China, based on the main characteristics of the market segment and consumer positioning, rely on their core products to provide customers with a business model of product of fully life cycle services (Liu and Wang, 2016). The process of upgrading the value chain carries a large number of advanced value-added services, and a richer variety of service products can improve the availability of certain services to a wider consumer group (Arnold et al., 2011). In turn, the expansion and upgrading of service demand will also guide domestic enterprises to transform to service-oriented products. Unlike the production of physical products, which mostly rely on energy inputs, the intangibility and immediacy of service demand itself determine its consumption characteristics of low energy consumption. Furthermore, the production and consumption processes of service products are carried out simultaneously, and this inseparable feature can incorporate consumers’ personalized service needs into the value creation process, strengthen customer relevance and product identity (Mont, 2002), reduce ineffective production and resource waste, thereby reducing energy loss and inhibiting the expansion of embodied carbon emissions in export trade.
Hypothesis 2c: From the perspective of final output structure, the functional upgrading of the value chain driven by the improvement of foreign investment quality generates more demand for low-energy service products for the development of manufacturing industry, which in turn can reduce the embodied emissions in China’s export trade.
3 Research Design
3.1 Model Setting
In order to test the impact of the functional upgrading of the value chain on the embodied emissions in China’s manufacturing export trade, this paper sets up the following multi-dimensional fixed-effect benchmark model:
Among them, i represents the manufacturing sub-industry, o represents the destination country (region), and t represents the year. CEiot indicates the embodied carbon emissions in China’s manufacturing sector’s export trade to the destination country (region) in year t, RFSIit indicates the relative functional specialization level of China’s manufacturing sector i engaged in upstream and downstream activities of GVCs in year t, Controls is a collection of other control variables. In addition to considering the relevant characteristic variables of industries and destinations that change over time, industry fixed effect vi, national (region) fixed effect vo and time fixed effect vt are introduced into the model. εiot represent the random perturbation term.
3.2 Description of Variable Measures
3.2.1 Explained Variable: Carbon Emissions Embodied in (CE) Export Trade
In this paper, we draw on the embodied carbon decomposition method of Meng et al. (2018) to track the embodied carbon emissions in trade according to the source and destination of trade in value-added under the framework of total trade accounting. Specifically, after classifying and decomposing China’s total exports to destination countries (regions) into eight value-added and double-counting parts, the carbon emission coefficient vector per unit value-added of China’s manufacturing industry is introduced, and finally the embodied carbon emissions in export trade corresponding to the decomposition part of each value-added part is obtained. The calculation is as follows:
The DVA_FINiot, DVA_INTiot, DVA_INTrexiot, and RDViot represent the final goods export, the direct and indirect intermediate goods exports, and the value-added that returned home, respectively. After multiplying the carbon coefficient vector f_ c, the four items on the right estimate the embodied carbon emissions in China’s final goods exports, direct and indirect intermediate goods exports, and the value-added that returned home, respectively. The sum of the four items represents the total amount of embodied carbon emissions generated in China, revealing the environmental pollution costs of China’s embedding in GVCs.
3.2.2 Core Explanatory Variable: Relative Functional Specialization Level (RFSI) in Upstream and Downstream of the Value Chain
Based on the perspective of foreign investment quality, this paper constructs a upstream and downstream functional specialization index of the value chain with the help of global greenfield investment micro project data, so as to quantify and present the trend of China’s functional upgrading in GVCs. Firstly, with reference to the classification method of Stöllinger (2021), each actual manufacturing greenfield investment project in China is matched to different functional modules of the value chain according to the type of activities it is mainly engaged in, including five functional specialization activities: headquarters economy, R&D, manufacturing, logistics and retail services, and after-sale services. Second, the actual effect of inward greenfield investment projects of different scales may be different. Therefore, the proportion of the amount of investment projects flowing to certain functional activities to the total investment in China relative to the corresponding proportion at the world level can objectively reflect China’s comparative advantage in the specific functional modules of the value chain. The specific calculation formula is as follows:
Among them,
Finally, in order to reveal China’s comparative advantages in the upstream and downstream of the value chain in terms of high-end functional modules, this paper further constructs the upstream and downstream relative functional specialization index (RFSIit) of the value chain that reflects the degree of functional upgrading.
Among them, UDFSIit represents the upstream and downstream functional specialization index of China’s value chain, that is, in Equation (19), f corresponds to the
3.2.3 Other Control Variables
Referring to the existing research on the influencing factors of embodied emissions in export trade, this paper also selects two sets of important variables for control. At the country-industry level, two variables are added: export scale (export) and FDI scale (fdi). The control variables at the manufacturing sector level include profit level (profit), capital intensity (capital), R&D innovation level (patent), energy consumption (coal), environmental regulation intensity (regu), GVCs participation (gvcpt), and GVCs division of labor (gvcpo).
3.3 Data Sources and Sample Descriptions
The data for micro greenfield investment projects measuring the core explanatory variables are derived from the fDi Markets database. It should be noted that by the end of 2018, the database had recorded a total of 206,669 global greenfield investment events, and based on the research needs, this paper finally matched 98,242 manufacturing projects, accounting for 47.5%, of which 10,954 manufacturing greenfield investment projects flowed into China, accounting for 11.2% of the world. The raw data for calculating the embodied emissions in export trade are derived from the Organisation for Economic Co-operation and Development Inter-Country Input-Output Tables (OECD-ICIO) and the supporting trade embodied carbon database. The control variable data are mainly from the OECD Trade in Value Added Database (OECD-TiVA), the China Industrial Statistical Yearbook, the China Science and Technology Statistical Yearbook, and the China Environment Statistical Yearbook.
Due to the differences in the subdivision standards of the manufacturing industry among different databases, in order to ensure the relative consistency of the statistical caliber, this paper first refers to the classification standard of Stöllinger (2021) and classifies each greenfield investment project into the manufacturing sector of the General Industrial Classification of Economic Activities of the European Community (NACE Rev.2.0). Subsequently, the International Standard Industrial Classification (ISIC Rev4.0) adopted by the OECD-ICIO is used as a benchmark, and the manufacturing sectors in NACE Rev.2.0 and the Industrial Classification of the National Economy (GB/T 4754–2017) are matched with them at the same time, and finally retaining 16 manufacturing industries.[1] Since the fDi Markets database has been counted since 2003, and the 2021 version of the OECD-ICIO covers 67 countries (regions) around the world, with the Input-Output matrix updated to 2018. Therefore, the scope of this paper is the balanced panel data of 16 sub-manufacturing sectors from 66 economies (including one rest of the world and excluding China) from 2003 to 2018, with a total of 16,896 sample observations.
4 Empirical Results and Analysis
4.1 Benchmark Regression Results
Table 1 reports the baseline regression results. Column (1) consider only the core explanatory variables, columns (2) and (3) include the rest of the control variables, and column (2) does not include fixed effects. The results show that the regression coefficient of the core explanatory variable is always significantly negative. In an economic sense, column (3) shows that when the relative upstream and downstream functional specialization of China’s manufacturing value chain increases by one standard deviation, the embodied emissions in export trade will be reduced by about 5.20%, so the funtional upgrading of the value chain driven by the improvement of investment quality can significantly reduce the embodied carbon emission level of China’s manufacturing export trade. Thus, hypothesis 1 is verified.
Benchmark Regression Results
| Variables | (1) | (2) | (3) |
|---|---|---|---|
| –0.0117*** | –0.0061** | –0.0138*** | |
| RFSI | (0.0033) | (0.0025) | (0.0039) |
| 0.3995*** | 0.2208*** | ||
| export | (0.0294) | (0.0273) | |
| 0.1912 | –0.5447 | ||
| fdi | (0.3244) | (0.3397) | |
| profi t | –0.0309 | –0.0470** | |
| (0.0252) | (0.0181) | ||
| –0.0240*** | –0.0165 | ||
| capital | (0.0074) | (0.0137) | |
| –0.0218*** | –0.0016 | ||
| patent | (0.0057) | (0.0011) | |
| 0.0741*** | 0.0695** | ||
| coal | (0.0210) | (0.0309) | |
| 10.0295*** | –7.1168*** | ||
| regu | (2.5929) | (2.4980) | |
| 1.0228*** | 3.1540*** | ||
| gvcpt | (0.3487) | (1.0242) | |
| –1.9195*** | –4.3285*** | ||
| gvcpo | (0.6429) | (0.9847) | |
| 0.7988*** | 1.3588** | 3.3119*** | |
| Constant | (0.0053) | (0.5579) | (0.6456) |
| Year fixed effect | Yes | No | Yes |
| Industry fixed effect | Yes | No | Yes |
| National fixed effect | Yes | No | Yes |
| N | 16896 | 16896 | 16896 |
| Adj-R2 | 0.2969 | 0.2230 | 0.3295 |
Note: Values in the brackets are the clustering robustness standard errors at the national (regional) level, ***, **, and * indicate that the variables are significant at the level of 1%, 5%, and 10%, respectively. The following tables are the same.
4.2 Robustness Test
The robustness test in this paper includes the following four aspects: (1) Replace the core explanatory variable measurement method. In this part, the following three indicators are used to reflect the characteristics of value chain functional upgrading. They are the Upstream and Downstream Functional Specialization Index (UDFSI), which does not contain information on manufacturing activities, the proportion of investment projects and the proportion of actual investment amount of greenfield investment enterprises in China flow into the upstream and downstream functional links of the manufacturing industry chain. (2) Change the coverage of the explanatory variable measure. On the one hand, the total export trade value data is used instead of the export value added decomposition to re-estimate the embodied carbon emissions in China’s export trade from various manufacturing industries to other countries (regions). On the other hand, this paper no longer focuses on the difference between the use object of the product and the nature of import and export, and estimates the overall carbon emissions generated by the country’s energy consumption based on the carbon emission coefficient published by the IPCC’s Guidelines for National Greenhouse Gas Emission Inventories. (3) Adjusting the regression model settings. Since the change of carbon emission does not occur instantaneously, it has obvious path-dependent inertia characteristics (He and Zhang, 2012). Here, the lagged period of embodied carbon in export trade is added to the explanatory variables of the model, and the adjusted dynamic panel model is regressed by the systematic GMM estimation method. (4) Sample data processing. First, the timber, wood products and softwood products manufacturing industry absorb zero greenfield investment projects in most years, thus this sector is excluded here. Second, due to the impact of the financial crisis, the scale of global capital and trade commodity flows in 2009 was significantly sluggish, thus the sample data for 2009 are excluded here. In sum, the robustness test results obtained according to the above different methods are consistent with the core conclusions of the benchmark, indicating that the results of this paper are basically reliable.
4.3 Endogeneity Discussion
In order to overcome the interference of potential endogeneity problems such as reverse causality and missing variables, appropriate instrumental variables is considered. To this end, this paper refers to the ideas of Liu and Wang (2016) and Xiong and Luo (2023) to select the level of specialization of the upstream and downstream functions of the manufacturing industry in India and Brazil as the instrumental variables, respectively. China, India and Brazil are all BRICS countries, which not only have similarities in the development path of the manufacturing industry, but also have similar status of GVCs function upgrading. As major global investment countries, the functional distribution of the value chain driven by foreign capital is approximately showing an inverted “U” shaped trend. Therefore, there is a high correlation between the functional division of labor in the value chain of India and Brazil and the core explanatory variables of this paper, and there is no obvious correlation between the functional embedding form and the embodied emissions in China’s export trade. In summary, the construction of such instrumental variables satisfies the correlation and exogenity conditions. The test results show that there is no unrecognizable and weakly recognized problem for the two instrumental variables, and the results of the two-stage least squares estimation show that the core conclusion is still reliable.
5 Mechanism Analysis
5.1 Factor Structure Optimization Effect
In terms of investment in the manufacturing sector, the functional upgrading of the value chain reflects the high-end and clean structure of factor use, and more productive service factors penetrate into the entire industrial chain. Based on this, this paper uses the total input coefficient (fserv) of foreign service factor input of various sectors in the domestic manufacturing industry to reflect the optimization effect of the improvement of investment quality on the factor input structure. The underlying data comes from the OECD-ICIO. Table 2 shows the results of the mechanism test from the perspective of factor inputs. Column (1) corresponds to the baseline regression result. The estimation coefficient of RFSI in column (2) is significantly positive, indicating that functional upgrading of the value chain driven by the improvement of foreign investment quality brings more mature and advanced service elements abroad. The regression coefficients of fserv in columns (3) and (4) are significantly negative, indicating that the optimization of factor input structure can reduce the embodied carbon emissions in manufacturing export trade. Furthermore, the Sobel Z-statistic is significant at the 1% level, and the confidence interval of the Bootstrap (1000 times) mediating effect test does not contain 0. In summary, hypothesis 2a is validated.
Mechanism Test I: Factor Structure Optimization Effect
| (1) | (2) | (3) | (4) | |
|---|---|---|---|---|
| Variables | CE | fserv | CE | CE |
| RFSI | –0.0138*** | 0.0001*** | –0.0129*** | |
| (0.0039) | (0.0000) | (0.0039) | ||
| fserv | –12.9774*** (2.6590) |
–12.5847*** (2.6168) |
||
| Control variables | Yes | Yes | Yes | Yes |
| Year fixed effect | Yes | Yes | Yes | Yes |
| Industry fixed effect | Yes | Yes | Yes | Yes |
| National fixed effect | Yes | Yes | Yes | Yes |
| Sobel Z | –4.800*** | |||
| Confidence intervals for bootstrap tests (1000 Times) | [–0.0015, –0.0002] | |||
| N | 16896 | 16896 | 16896 | 16896 |
| Adj-R2 | 0.3295 | 0.9764 | 0.3297 | 0.3299 |
5.2 Human Capital Accumulation Effect
Employees engaged in high-end activities within foreign enterprises can acquire advanced production technology and management experience through skills training and other means, which will accelerate the technology spillover and diffusion to domestic enterprises, and ultimately affect the embodied carbon emissions in China’s manufacturing sector’s export trade. This paper uses the job creation ratio (jobsh) directly created by greenfield investment projects in China to serve the upstream and downstream functional activities of the value chain to reflect the human capital accumulation effect. The relevant job data comes from fDi Markets. Table 3 reports the results of mechanism testing from the perspective of technology spillovers. The results in column (1) are from baseline regression. Column (2) estimates show that the specialization of functions of upstream and downstream of the value chain contributes to the accumulation of high-level human resources. The estimation coefficients of jobsh in columns (3) and (4) are both significantly negative, indicating that human capital accumulation can help reduce the embodied carbon emissions in export trade. Overall, hypothesis 2b is validated.
Mechanism Test II: Human Capital Accumulation Effect
| Variables | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| CE | jobsh | CE | CE | |
| RFSI | –0.0138*** | 0.0044*** | –0.0126*** | |
| (0.0039) | (0.0000) | (0.0036) | ||
| jobsh | –0.3124*** (0.0926) |
–0.2876*** (0.0867) |
||
| Control variables | Yes | Yes | Yes | Yes |
| Year fixed effect | Yes | Yes | Yes | Yes |
| Industry fixed effect | Yes | Yes | Yes | Yes |
| National fixed effect | Yes | Yes | Yes | Yes |
| Sobel Z | –3.312*** | |||
| Confidence intervals for bootstrap tests (1000 Times) | [–0.0024, –0.0002] | |||
| N | 16896 | 16896 | 16896 | 16896 |
| Adj-R2 | 0.3295 | 0.4996 | 0.3296 | 0.3297 |
5.3 Service Demand Creation Effect
The expansion of service demand intuitively reflects the optimization of the output structure of the manufacturing sector, and the satisfaction of consumers’ personalized service needs leads to less resource and energy consumption, which in turn acts on the overall scale of embodied carbon emissions in China’s export trade. Since it is difficult to directly quantify the actual response of each manufacturing sector to the demand for consumer services, this paper uses the factor input ratio of each manufacturing sector embedded in the service industry to realize the separation of the final demand of the service sector. The specific calculation is as follows:
Among them, demandit is the quantity of service demand created by manufacturing industry i in year t, Aist represents the input coefficient of the intermediate input of each service sector s required by manufacturing sector i in year t, and HFCEst, NPISHst and GGFCst represent the final consumption of households, non-profit organizations and government departments satisfied by each service sector s in year t, respectively. The raw data comes from the OECD-ICIO.
Table 4 presents the results of the mechanism test from the perspective of the final output structure. Column (1) is the baseline result, the regression coefficient of RFSI in column (2) is significantly positive, indicating that the functional upgrading of the value chain can derive more demand for service products, and the estimation results in columns (3) and (4) show that more consumer demand for services will help inhibit the expansion of embodied carbon emissions in export trade. Based on the above analysis, hypothesis 2c is verified.
Mechanism Test III: Service Demand Creation Effect
| Variables | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| CE | demand | CE | CE | |
| RFSI | –0.0138*** | 0.3327*** | –0.0103*** | |
| (0.0039) | (0.0001) | (0.0034) | ||
| demand | –0.0111*** (0.0023) |
–0.0106*** (0.0022) |
||
| Control variables | Yes | Yes | Yes | Yes |
| Year fixed effect | Yes | Yes | Yes | Yes |
| Industry fixed effect | Yes | Yes | Yes | Yes |
| National fixed effect | Yes | Yes | Yes | Yes |
| Sobel Z | –4.854*** | |||
| Confidence intervals for bootstrap tests (1000 Times) | [–0.0049, –0.0022] | |||
| N | 16896 | 16896 | 16896 | 16896 |
| Adj-R2 | 0.3295 | 0.8491 | 0.3302 | 0.3302 |
6 Extended Analysis
6.1 Distinguish Specific Functional Activities
Using the upstream and downstream functional specialization index of the value chain to roughly depict the trend of functional upgrading of China’s manufacturing may mask the difference in the impact of specific upstream and downstream activities on the embodied emissions in China’s export trade. The estimation results show that specialization in specific functional activities in the upstream and downstream of the value chain can help reduce the embodied emissions in export trade. Among them, the regression coefficient of relative functional specialization in the upstream is slightly greater than that of the downstream. The reason for this is that the upstream link can start from the source of the industrial chain and spread the environmental protection business philosophy within the enterprise, focusing on the use of cleaning elements and the improvement of production efficiency. The downstream link mainly extends the value creation space and upgrades the market demand structure by providing value-added services, so the upstream link has a relatively more direct role in curbing carbon emissions.
6.2 Distinguish Greenfield Investment Destinations
Different regions in China have different resource endowment conditions and industrial undertaking bases, which to a certain extent causes the uneven quality of foreign capital utilization. Based on the classification of the eastern, central and western regions by the National Bureau of Statistics, this paper examines the differences in the role of greenfield investment in the eastern and central western regions on the embodied emission in China’s export trade. The results show that the upgrading of the value chain driven by foreign investment to eastern region of China can significantly reduce the embodied emission in export trade, but not in the central and western regions. The possible reason is that the eastern region has mature industrial supporting facilities and talent reserves, and the attraction of high-end functional activities of foreign investment and the ability to receive and apply foreign technological achievements are relatively stronger. However, due to its energy enrichment advantages, the central and western regions undertake the transfer of a large number of energy-intensive industries, and the burden of emission reduction is heavier.
6.3 Distinguish Manufacturing Industry Types
In view of the differences in carbon emission intensity among manufacturing industries, this paper examines the impact of the functional upgrading of the value chain on the embodied emission in export trade of pollution-intensive and non-pollution-intensive manufacturing industries according to Busse’s (2004) classification criteria. The results show that the manufacturing industry with different embodied carbon emission scales shows distinct heterogeneity in the process of value chain upgrading. Among them, for pollution-intensive manufacturing industries, the regression coefficient of the core explanatory variable is significantly negative. In other words, the functional upgrading of the value chain shows a strong pertinence of pollution control. The reason for this is that China’s pollution-intensive manufacturing industries are obviously dependent on energy inputs, and they are easier to be constrained in low-end production links by multinational corporations. Therefore, to achieve low-carbon development of the manufacturing industry, the high-energy-consuming manufacturing industries need to make more obvious carbon emission reduction contributions, and there is a large room for energy conservation and carbon reduction in these sectors.
7 Conclusions and Implications
This paper attempts to incorporate the functional division of labor factors of the value chain into the theoretical analysis model of trade and environmental pollution. Based on the high-micro project data of global greenfield investment enterprises in the fDi Markets database, a relative functional specialization index of the upstream and downstream of the value chain driven by the improvement of foreign investment quality is constructed. Quantifying the environmental costs of China’s embedding of GVCs with the help of embodied emissions in export trade contained the connotation of cross-border transfer. Finally, using the panel data of 16 sub-manufacturing industries in 66 economies from 2003 to 2018, this paper empirically explores the impact of the functional upgrading of the manufacturing value chain on the embodied emissions in export trade and its role channels. The results show that the functional upgrading of the value chain promoted by the improvement of the quality of investment has a “pollution halo” effect, which can significantly reduce the embodied emissions in China’s export trade. The mechanism test emphasizes that the functional upgrading of the value chain driven by the improvement of the quality of investment mainly acts on the embodied carbon emission reduction of export trade through the optimization effect of factor structure, the effect of human capital accumulation and the effect of service demand creation. The extended analysis shows that high-quality foreign investment engaged in functional activities in the upstream of the value chain and flowing to the eastern region has a more obvious effect on the reduction of embodied carbon in export trade. At the same time, the functional upgrading of the value chain can help reduce the embodied carbon emissions in trade in pollution-intensive manufacturing.
Based on the above research conclusions, this paper puts forward the following policy implications: First, increase the investment in the manufacturing industry and strive to improve the quality of foreign investment. China should steadily promote a higher level of institutional opening-up and seek to improve the quality and efficiency of foreign investment in the manufacturing industry. Second, optimizing the input structure and circulation environment of manufacturing factors to drive the expansion and upgrading of service demand. China should continue to expand the opening up of the service industry and strengthen the cooperation and interaction between foreign investment in the producer service industry and the domestic manufacturing industry-related sectors. Third, China should guide the flow of foreign investment and fully unleash the effect of environmental optimization. It is necessary to actively encourage foreign investment to set up business headquarters, R&D and innovation centers and supporting service networks in China, support foreign investment to participate in the carbon peak and carbon neutrality target strategy, and focus on driving pollution-intensive manufacturing to achieve low-carbon transformation.
References
Antweiler, W., Copeland, B., & Taylor, M. (2001). Is Free Trade Good For the Environment?. American Economic Review, 91(4),877–908.10.1257/aer.91.4.877Search in Google Scholar
Arnold, J., Javorcik, B., & Mattoo, A. (2011). Does Services Liberalization Benefit Manufacturing Firms? Evidence from the Czech Republic. Journal of International Economics, 85(1), 136–146.10.1016/j.jinteco.2011.05.002Search in Google Scholar
Busse, M. (2004). Trade, Environmental Regulations and the World Trade Organization: New Empirical Evidence. Journal of World Trade, 38(2), 285–306.10.54648/TRAD2004012Search in Google Scholar
Chichilnisky, G. (1994). North-South Trade and the Global Environment. American Economic Review, 84(4), 851–874.Search in Google Scholar
Cole, M., Elliott, R., & Strobl, E. (2008). The Environmental Performance of Firms: The Role of Foreign Ownership, Training, and Experience. Ecological Economics, 65(3), 538–546.10.1016/j.ecolecon.2007.07.025Search in Google Scholar
Copeland, B., & Taylor, M. (1994). North-South Trade and the Environment. Quarterly Journal of Economics, 109(3), 755–787.10.2307/2118421Search in Google Scholar
Eskeland, S., & Harrison, E. (2003). Moving to Greener Pastures? Multinationals and the Pollution Haven Hypothesis. Journal of Development Economics, 70(1), 1–23.10.1016/S0304-3878(02)00084-6Search in Google Scholar
He, X., & Zhang, Y. (2012). Influence Factors and Environmental Kuznets Curve Relink Effect of Chinese Industry’s Carbon Dioxide Emission——Empirical Research Based on STIRPAT Model with Industrial Dynamic Panel Data. China Industrial Economics (Zhongguo Gongye Jingji), 1, 26–35.Search in Google Scholar
Huang, Y., & Xie, J. (2019). Input Servitization of Manufacturing Industry and Carbon Emission Intensity——Empirical Analysis Based on WIOD’s Cross-country Panel. Finance & Trade Economics (Caimo Jingji), 40 (8), 100–115.Search in Google Scholar
Jiang, X., & Meng, L. (2021). Mainly Inner Circulation, Outer Circulation Empowerment and Higher Level Double Circulation:International Experience and Chinese Practice. Management World (Guanli Shijie), 37 (01), 1–19.Search in Google Scholar
Kordalska, A., & Olczyk, M. (2023). Upgrading Low Value-added Activities in Global Value Chains: a Functional Specialisation Approach. Economic Systems Research, 35(2),265–291.10.1080/09535314.2022.2047011Search in Google Scholar
Liu, N., & Wang, N. (2016). Input Servitization of Manufacturing and Dual Margins of Firms’ Export——An Empirical Study Based on the Data of Chinese Micro-enterprise. China Industrial Economics (Zhongguo Gongye Jingji), 9, 59–74.Search in Google Scholar
Lv, Y., & Lv, Y. (2019). The Environmental Effect of China’s Participation in Global Value Chain. China Population, Resources and Environment (Zhongguo Renkou Ziyuan yu Huanjing), 29 (7), 91–100.Search in Google Scholar
Lv, Y., Cui, X., & Wang, D. (2019). GVC Participation and Carbon Embodied in International Trade: Nonlinear Analysis Based on GMRIO and PSTR Model. Journal of Quantitative & Technological Economics (Shuliang Jingji yu Jishu Jingji), 36 (2), 45–65.Search in Google Scholar
Meng, B., Peters, G., Wang, Z., & Li, M. (2018). Tracing CO2 Emissions in Global Value Chains. Energy Economics, 73(1), 24–42.10.1016/j.eneco.2018.05.013Search in Google Scholar
Mont, O. K. (2002). Clarifying the Concept of Product–Service System. Journal of Cleaner Production, 10(3), 237–245.10.1016/S0959-6526(01)00039-7Search in Google Scholar
Rothenberg, S. (2007). Sustainability Through Servicizing. MIT Sloan Management Review, 2(48), 83–91.Search in Google Scholar
Shou, C., Su, D., & Yang, Q. (2021). Environmental Effects of FDI on Domestic Firms in Host Countries: Evidence from China. The Journal of World Economy (Shijie Jingji), 44 (3), 32–60.Search in Google Scholar
Stöllinger, R. (2021). Testing the Smile Curve: Functional Specialisation and Value Creation in GVCs. Structural Change and Economic Dynamics, 56,93–116.10.1016/j.strueco.2020.10.002Search in Google Scholar
Timmer, M., Sébastien, M., & De Vries, G. (2019). Functional Specialisation in Trade. Journal of Economic Geography, 19(1), 1–30.10.1093/jeg/lby056Search in Google Scholar
Vries, G., Chen, Q., Hasan R., & Li, Z. (2019). Do Asian Countries Upgrade in Global Value Chains? A Novel Approach and Empirical Evidence. Asian Economic Journal, 33(1), 13–37.10.1111/asej.12166Search in Google Scholar
Wang, Z., Zhang, Y., Niu, M., & Zhong, Y. (2020). Dynamic Changes of Functional Specialization in China’s Export and International Comparison under Global Value Chains. China Industrial Economics (Zhongguo Gongye Jingji), 6, 62–80.Search in Google Scholar
Xiong, B., & Luo, K. (2023). Input Servitization of China’s Manufacturing and Functional Upgrading of the Value Chain——From the Perspective of Inward Greenfield Investment. Journal of International Trade (Guoji Maoyi Wenti), 2, 126–142.Search in Google Scholar
Zhao, Y., Zhang, Z., Wang, S., & Wang, S. (2014). CO2 Emissions Embodied in China’s Foreign Trade: An Investigation from the Perspective of Global Vertical Specialization. China & World Economy, 22(4),102–120.10.1111/j.1749-124X.2014.12077.xSearch in Google Scholar
Zhu, M., Stern, N., Stiglitz, J., Liu, S., Zhang, Y., Li, J., & Hepburn, C. (2023). Embracing the New Paradigm of Green Development: A Study of China Carbon Neutrality Policy Framework. The Journal of World Economy (Shijie Jingji), 46 (3), 3–30.Search in Google Scholar
Zhu, S., Xie, Y., & Wu, D. (2020). A Study on the Energy-Saving Effect of Manufacturing Servitization and Its Intermediary Mechanism. Finance & Trade Economics (Caimo Jingji), 41 (11), 126–140.Search in Google Scholar
© 2024 Fuzhong Chen, Ke Luo, Kangyin Dong, Published by De Gruyter
This work is licensed under the Creative Commons Attribution 4.0 International License.
Articles in the same Issue
- Frontmatter
- Column: China’s Economic Development
- China’s Dual Structure-Based Growth Accounting—Theoretical and Empirical Analysis of Introducing the Labor Employment Rate
- The Growth Contribution of China’s Regional Coordinated Development Strategy—On the Dispute of Regional Policy Paths
- New Production Relations Driven by New Quality Productive Forces: Trends, Challenges and Countermeasures
- The Theoretical Gap in the Study of New Quality Productive Forces and the Economic Analytical Perspective of “Heterogeneity”
- The Impact of Manufacturing Transfer from China to India on China’s GDP and Employment
- Functional Upgrading of Value Chains and the Carbon Emissions Reduction Embodied in China’s Exports: From the Perspective of the Improvement in the FDI Quality
Articles in the same Issue
- Frontmatter
- Column: China’s Economic Development
- China’s Dual Structure-Based Growth Accounting—Theoretical and Empirical Analysis of Introducing the Labor Employment Rate
- The Growth Contribution of China’s Regional Coordinated Development Strategy—On the Dispute of Regional Policy Paths
- New Production Relations Driven by New Quality Productive Forces: Trends, Challenges and Countermeasures
- The Theoretical Gap in the Study of New Quality Productive Forces and the Economic Analytical Perspective of “Heterogeneity”
- The Impact of Manufacturing Transfer from China to India on China’s GDP and Employment
- Functional Upgrading of Value Chains and the Carbon Emissions Reduction Embodied in China’s Exports: From the Perspective of the Improvement in the FDI Quality
