Carbon Emission Effects of the Coordinated Development of Two-Way Foreign Direct Investment in China

Round  1

Reviewer 1 Report

While this is an interesting study, it has moderate deficiencies. First and the foremost, English language requires slight improvement. Also, it includes almost no substantial policy implications and discussion of empirical findings. My main concern is that it looks more like a hastily performed mechanical econometric exercise than a publishable and thorough academic research. It's value to readers is very limited as it is hard to follow what point they are trying to make starting from the beginning to the end. I would expect to see an in-depth discussion of policy implications and how they relate to the existing theory and the literature. It is clear that they authors are trying to fill a gap in the literature but they are not explaining the background and the conceptual/theoretical framework. This is particularly the case from a policy perspective. I would suggest that the authors spend a bit more time on it to improve it with proper discussion of their findings, what they mean in practice, and what their significance is from a policy perspective. They can perhaps devote a 1-2 page to policy implications and the discussion of empirical results with proper discussions in light of the theory and the existing literature, as well as the policy constraints, environmental concerns and/or socioeconomic factors specific to China.

Author Response

Thank you for the reviewers’ comments. We revised the manuscript point by point according to the reviewer’s comments using the "Track Changes" function in the Word to make the revision easily visible to the editors and reviewers.

Response to Reviewer 1 Comments

While this is an interesting study, it has moderate deficiencies.

Point 1: First and the foremost, English language requires slight improvement.

Response 1: As requested, Dr. Arunima Malik, one of our authors and a native English speaker, has polished the language throughout the manuscript.

Point 2: Also, it includes almost no substantial policy implications and discussion of empirical findings.My main concern is that it looks more like a hastily performed mechanical econometric exercise than a publishable and thorough academic research. It's value to readers is very limited as it is hard to follow what point they are trying to make starting from thebeginning to the end. I would expect to see an in-depth discussion of policy implications and how they relate to the existing theory and the literature. It is clear that they authors are trying to fill a gap in the literature but they are not explaining the background and the conceptual/theoretical framework. This is particularly the case from a policy perspective. I would suggest that the authors spend a bit more time on it to improve it with proper discussion of their findings, what they mean in practice, and what their significance is from a policy perspective. They can perhaps devote a 1-2 page to policy implications and the discussion of empirical results with proper discussions in light of the theory and the existing literature, as well as the policy constraints, environmental concerns and/or socioeconomic factors specific to China.

Response 2: In the last part, we removed conclusions in previous edition and added 2 pages to detail discussion of empirical results and policy implications in this paper.

The detailed discussion of empirical results is as follows, and in green color is the new added.

6.1 Discussion

Under the background of paying equal attention to the “introduction of foreign investment and investment in foreign countries”, based on the analysis framework of Copeland and Taylor, this paper examines effects of China's spatial spillover andtwo-way FDI coordinated development on domestic carbon emissions whichis realized through three channels — the scale, composition, and technique effects. 

Carbon emissions in China have a significant spatial spillover effect and remarkable spatial agglomeration phenomenon among provinces. This result coincides with the existing studies using spatial econometric models^[42,64-66], which indicates that spatial correlation characteristics of environmental pollution are of existence. If researchers and policy makers ignore these linkages, it will likely lead to unreliable results and subsequently incorrect policy guidance.

The two-way FDI coordinated development has a significant effect on emission reduction. Previous studies only focused on effects of IFDI or OFDI on environment^[21,37,38], however this study is the first to conclude that the findings agree with China’s strategy of IFDI and OFDI coordinated development and guidance for China’s emission reduction, and thus sustainable development goal of opening up and environmental protection can be integrated effectively.

We found that the two-way FDI coordinated development has a positive scale effect on carbon effect, which is consistent with existing studies that the scale effect can increase emissions^[13-15]. Composition effect, which has a negative effect on emissions is different from previous studies, which considered that capital-intensive production will increase emissions,^[18,67]but it supports that higher capital/labor ratio implies more advanced technology level^[37], which will help in reducing emissions. The key driver of the emission reduction effect is due to the leading role of the technique effect, which has far exceeded the scale and the composition effect. Therefore, the total effect of the two-way FDI on carbon emissions is consistent with the direction of the technique effect.

The detailed policy implications of this paper are shown in session 6.2 and they are as follows (green color).

6.2 Policy implications

This research not only provides useful guidance for policy makers to effectively control emissions by implementing economic means, but also provides a practical example for countries that are undergoing similar development stage as China.

The regional emissions have significant positive correlation at a spatial level. This indicates that carbon emissions produced in one area not only affect the local ecological environment but also probably diffuse towards surrounding areas due to industrial transfers, factor mobility, and interregional trade. Thus, environmental governance should focus on strengthening the coordination and cooperation between local governments. Any unilateral governance behaviors can only generate limited effects on local environment. The interregional joint-prevention and joint-control strategies should be adopted to effectively resist the diffusion of emissions and emission leakages. For example, carbon emission trading system can be applied to interregional industries to reduce the diffusion and leakages among regions.

The two-way FDI coordinated development has a significant effect on emission reduction, which implies that the effective coordination of the process of introducing foreign investment and overseas investment can help in mitigating carbon emissions. We have elaborated the interactive mechanism of IFDI and OFDI in the theoretical part of this paper and empirical studies have already verified that IFDI can drive OFDI^[26]and OFDI can promote IFDI in turn^[27]. Previous studies primarily use only IFDI or OFDI to show how governments’ decisions can maximize performance of resources and environment^[37,38], whereas this paper provides a new perspective to merge IFDI and OFDI to form an interactive mechanism to achieve emission reduction. We present results that can used for informing policy making, in particular for promoting the coordination of IFDI and OFDI for achieving the aim of emission reduction and pollution regulation. In current governmental system, IFDI and OFDI related institutes include National Development and Reform Commission, Ministry of Commerce, and China Customs etc. The institutes have a different role to play in foreign investment because of their specific responsibilities. When making policies, cooperative mechanism among the decision institutes could be a way of enhancing their connections to promote the coordination of IFDI and OFDI and to play a more active role in the reduction of carbon emissions.

The two-way FDI coordinated development results have shown that the leading emission reduction pathway should be focused on the technique effect. This reminds us that during the development of two-way FDI, attention should be directed towards improving the quality of FDI. During expansion of the size of the two-way FDI, it is of utmost importance to identify two-way FDI with cleaner technologies and to bring IFDI’s technology diffusion and OFDI’s inverse diffusion. On the one hand, IFDI can introduce advanced and cleaner production technologies and produce green technological diffusing effects in upstream and downstream industries in local and inter-regions to help in improving China’s environmental quality. Thus, local governments should pay attention to environmentally friendly FDI, for enhancing the requirements for China’s market access, and to bring IFDI into full play of technical advantages on emission reduction and spillover effect. On the other hand, technology-seeking OFDI scale should be further enlarged to achieve reverse technical diffusion and drive China’s manufacturing industries’ promotion with cleaner technologies. In addition, OFDI should merge with China’s industrial structure upgrade and strengthen international industries’ coordination by the Belt and Road Initiativeto promote domestic industry transfers to developing countries. During this process, environmental pollution in destination countries caused by intermediate input from domestic industries should also be resolved using the same cleaner technology of production.

Author Response File: Author Response.docx

Reviewer 2 Report

I recommend the following amendments:

·        The implications for research, theory, practice and society are not clear though I can see that these aspects can be elaborated further. The authors need to conduct and familiarise with the academic literature surrounding the subject matter to answer the implications for research, practice and/or society much clearer.

·        Methodology is appropriate for the study. However, some of the statements required more clarifications.

·        The results do make sense but they are not presented in a proper format. I would like to know how the results were similar to or different from other studies.

Author Response

Thank you for the reviewers’ comments. We revised the manuscript point by point according to the reviewer’s comments using the "Track Changes" function in the Word to make the revision easily visible to the editors and reviewers.

Response to Reviewer 2 Comments

I recommend the following amendments:

Point 1: The implications for research, theory, practice and society are not clear though I can see that these aspects can be elaborated further. The authors need to conduct and familiarise with the academic literature surrounding the subject matter to answer the implications for research, practice and/or society much clearer.

Response 1: In the Introduction and Literature Review, we elaborated the implications of our study for research, theory, practice and society, respectively.

In the introduction session, we added the elaborated implications (in green) in the second and third paragraphs.

However, China's expansion of opening-up process has also highlightedobvious climate change issues. Both “race to the bottom line”^[6]and “dirty industries transfer” theories^[7,8]believe that, with promotion of free trade, countries relax the standards of their environmental regulations in order to maintain or enhance their international competitiveness, which results in a deterioration of the global environmental quality^[6]. Thus, the impacts of international economic activities on the environment have prompted governments and scholars in various countries to consider the interrelationship between cross-border economic activities and climate change. IFDI and OFDI, one of the main forms of cross-border capital flows, serve as an important way of international economic activities and induce thereof the climate change issues. This has become a hot topic of constant concern and high controversy among researchers and policy makers. Figure 1 shows that China's IFDI, OFDI, and carbon emissions maintain the continuous growth tendency at the same time. As the largest carbon emitter, China has made an arduous emission reduction commitment to respond to global climate issues. China has committed to reduce its carbon intensity (emissions per unit of GDP) by 40-45% in 2020 compared to the level of 2005 in the 2009 Copenhagen Summitand has put forward its CO₂emission peak by 2030 within the framework of Paris Agreement.

Therefore, China’s faces the growing challenge of maintaining its growing economic system, whilst at the same time ensuring economic, social and environmental sustainability. IFDI and OFDI are primary avenues for opening to the world and embedding into the global value chain, hence their coordinated development has become one of the important strategies for China’s current and future opening up strategy. To achieve win-win scenarios for the two-way FDI coordinated development and the reduction of carbon emissions has been a difficult problem, and is a challenge for China’s economic transition and continuous growth of its opening to the outside world. The questions we need to answer are:What is the relationship between the two-way FDI and carbon emissionsin China? Should the two-way FDI be responsible for China's carbon emissions? To end this, in the context of the cross-border capital flow characterized by the coordinated development of the two-way FDI, the study on potential factors influencing the carbon emissionsis of great theoretical and practical significance to the construction of China's ecological civilization, and thus alsoto the mitigation of local pollution diffusion towards the global range because the flows of international capital and industrial transfer are accompanied byeconomic globalization.

In the Literature Review session, we added the detailed implications (in green) in the several paragraphs in the session 2.2.

Only a few studies have analysed the influence of OFDI on IFDI. Nie and Liu (2019) found that the dual transmission mechanism of China's OFDI has a significant positive impact on the scale and quality of its IFDI. For host countries with abundant natural resources and high technical proficiency, the structural transmission mechanism of China's OFDI affecting the quality of IFDI is particularly obvious. And for low-income host countries, the exchange rate transmission mechanism of China's OFDI affecting the scale of IFDI is more obvious^[27].

Furthermore, the interaction between two-way FDI is also beginningto be widely recognised. Based on the new open economy macroeconomic framework, Tian and Wang used cross-border panel data to build simultaneous formulas model to investigate the interaction relationship between IFDI and OFDI. The results found that there is a significant interaction between IFDI and OFDI, and the higher the national overall income level, the more sustainable interactive development of the IFDI and OFDI^[28]. Based on data on Chinese manufacturing industries and the panel vector autoregressive model, Huang verified the interaction effect of the two-way FDI and the global value chain drive model, the form for division of labor and the governance model has promoted the interactive development of the two-way FDI^[29].

In nutshell, most of the previous studies have investigated the positive and negativeimpacts of IFDI on host country’s environment. Some of them support the Pollution Heaven Hypothesis^[30-33], whilst others focus on the Pollution Halo Hypothesis^[34-36]. In particular, these studies have only focused on the total effects of IFDI on the host country’s environment and do not discuss the mechanism of IFDI on the environment^[37]. Another stream of research has been undertaken on the mechanism investigation^[13,15], but such literature has also ignored the possibilities of pollution diffusions across regions because of interregional economic activities.

It is important to note that,there is a knowledge gapon the effect of OFDI on host country’s environment. Although some studies have undertaken effect analysis between OFDI and environment^[38], further mechanism and spatial correlation characteristics still need to be investigated.Prior research has verified that there existsan interactive mechanismbetween IFDI and OFDI^[26,27,29]. Ignoring such interactions provide incomplete results, and these results obtained from a single aspect of IFDI or OFDI cannot provide an informed assessment for policy making.

Considering this evident gap in literature,the novelty of this paper lies in three aspects: (1) based on the objective fact of interactive development of the two-way FDI, IFDI and OFDI are integrated as ameasure for representingthe level of coupling and coordination development to master the internal interaction mechanism of the two-way FDI; (2) The mechanism of two-way FDI on emissions, and the overall effect of the two-way FDI on carbon emissions is divided into scale effect, composition effect and technique effect. (3) The spatial diffusion characteristics of emissions across regions areexamined with the construction ofthe spatial econometric modeland China's national andprovincial carbon emissions caused by the two-way FDIaremeasured, during 2004-2016 to investigate the impact mechanism of the two-way FDI’s coordinated development on carbon emissions.

Point 2: Methodology is appropriate for the study. However, some of the statements required more clarifications.

Response 2: We checked throughout the Methodology related parts, i.e., Session Theoretical Model and Session 4 Econometric Model, Variable Definitions and Data Descriptions. We clarified some statements. The details are as follows in green.

3. Theoretical Model

Following the analytical framework of Copeland and Taylor^[9], this paper attempts to introduce the coordinated development level of IFDI and OFDI to describe the functioning mechanism on carbon emission effect by coordinated development of two-way FDI, which sets the theoretical foundation for the empirical analysis presented in this paper.

3.1. Basic settings

It is assumed that only two products are produced in an economic system, one is a pollution-free clean product, the other is a contaminated product.Carbon emissions  are generated when the product  is produced simultaneously, which leads to a negative externality effect and generates social costs. In the case of clear delimitation for property rights, enterprises pay certain fees for pollution discharge to make up for the negative impact on the environment. These expenses are usually considered environmental taxes, pollution charges, or pollution discharge permission fees. The optimal decision-making for an enterprise with the goal of maximizing its profit is such that a part of production factors isused to reduce the emissions.

This paper assumes that the production function of the product  is ,where capital Kand labor Lare factors for production input, respectively.The potential output of the product is. The proportion of input used by the enterprise for emission reduction to the total factors is .If, the enterprise inputs all production factors into the production of product, which means that the enterprise totally disregards emission reduction.If ,the enterprise uses the proportion of all production factors into emissiondischarge is , the real output of  is  and there is a certain amount of carbon emission , such that:

where is the carbonemission function about  for the enterprise, a decreasing function of . is production technology and parameters. The production function of  is obtained by combiningthe Equation (1) - (3):

3.2 Production decision-making

Based on the Equation (4), when implementing the production decision-making, the enterprise, should select an appropriate capital to labor ratioaccording to the interest rate of external capital price and the salary of the labor price  to minimize the cost  for producing theunit of potential output . Given the cost  for producing the unit of potential output  and carbon emission cost , an optimal combination between potential output and carbon emission  is selected by the enterpriseto minimize the unit cost  of the production of product, then                                                                   

The optimal first order conditions of the Equations (5) and (6), respectively.

3.3. Pollution discharge decision-making

Assuming that in a fully competitive market, the profit of the enterprise is 0 and the price of the product is . Then:                                                                                                                        

Obtained from Equations (8) and (9):                                                                                                                     

Further, Equation (10) is transformed as:

Defining, , Equation (11) is sorted and log-linearized as:  

Equation (12) shows that, when the enterprise produces product , its carbon emissions arerelated to the production scale , production composition ,and the production technique , that is, the carbon emissions areinfluenced by the scale factor, the composition factor, and the technique factor.

3.4. Impact mechanism of the two-way FDI on emissions

The impacts of China's IFDI on domestic environment shows that ①IFDI has expanded production scale and output level of domestic enterprises in China, while the pollution discharge as derivatives in the production process increased along with the expansion of production scale, thus, the scale effect of IFDI has a positive effect on China's pollution discharge. ②The technology spillover effect caused by the IFDI has promoted the technological progress of Chinese enterprises. Technological progress usually is directional in practice and can be divided into progress tending to production technology and progress tending to emission reduction technology. The former can increase the production scale and thus increase emissionsby improving productivity; the latter can reduce emissions by improving emission reduction technology. The direction of technological progress determines the direction of impact on emissions. Thus, the technique effect of IFDI has uncertainty on increasingor decreasingimpact of emissions in China. ③IFDI has led to the adjustment of the production composition of Chinese domestic enterprises. The IFDI has rapidly expanded the relative proportion ofthe capital elements of enterprise and thus improved the ratio of capital to labor. The higher ratio of capital to labor in the production means higher technical efficiency and thus the ratio of capital to labor can also reflect the level of the technical level to a certain extent. In general, capital-intensive production has a higher technology level and labor-intensive production technology level is lower^[39]. It is worth noting that capital-intensive industries may have cleaner production technology^[40]and also carry higher energy consumption and emission demand^[18]at the same time. Therefore, the composition effect of the IFDI has uncertainty on the impact of domestic emission in China.

The impacts of China's outward FDI on domestic environment represent that ①the OFDI of Chinese enterprises can influence emissions by adjusting its production composition. On the one hand, the OFDI of Chinese enterprises can change the proportion of its capital to labor in domestic production, which is similar to the IFDI to change the domestic ratio of capital to labor, inducing uncertainty on direction of pollution discharge; however on the other hand, in the process of OFDI, Chinese enterprises can transfer some high-pollution and high-energy-consumption production lines to foreign countries to reduce the domestic emissions. Therefore, the comprehensive effect of two aspects also makes the direction of China's FDI on domestic emissions uncertain. ②The OFDI of Chinese enterprises can also promote the domestic technological progress through reverse technology spillover effect, which is consistent with the technology spillover effect of IFDI, and the deviation of the technological progress determines the direction of its impact on emissions. Therefore, the technique effect of Chinese enterprises’ OFDI has the same uncertainty on the impact direction of emissions. ③Chinese enterprise’ OFDI also contributes to the promotion of motherland's economic growth and the expansion of the domestic output scale. Due to the greater proportion of China's second industry in national economy in the past decades, the higher economic scale means a higher level of industrialization and more emission generation. Therefore, China's OFDI has a positive impact on domestic emissions.

Considering the above effects of IFDI and OFDI on emissions and the interaction between the two-way FDI, we incorporate the IFDI and OFDI into a system to investigate the effect of integrated FDI on the economic and social development. Coupling originally is a concept in physics, which means thatthe phenomenon of synergy caused by interactions amongtwo or more systems. In other words, coupling refers to a phenomenon in which two or more systems or forms of motion interact with each other through various interactions.The degree of coupling reflects the measurement for such kind of synergy. To describe the interaction feedback mechanism between IFDI and OFDI, referring to the model of coupling degree in physics^[41], the coupling relationship between IFDI and OFDI can be expressed as:

Because of the difference between IFDI and OFDI and the unbalanced characteristics of the two-way FDI development in each region, lower values of IFDI and OFDI may have higher degree of coupling. In addition, the coupling degree model can only indicate the existence of interaction between systems, and cannot reflect the level of coupling and coordination between systems.Thus, the degree of coordination is introduced to avoid confusion.

where . ,a large  value refers to a high degree of coordination for IFDI and OFDI; conversely, a small value refers to a low level of coordination for both cases. Similar to the impacts of both IFDI and OFDI on enterprises’ emissions through production scale, production technique, and production composition, the coupling coordination system of IFDI and OFDI also influence enterprises’ emissions through the above three aspects.

An increase of domestic research and development (R&D) expenditure enhances the independent innovation ability of enterprises and promotes the progress of enterprise production technology to a certain extent ^[18]. Thus, the technique effect function can be expressed as:

In addition, factor endowment can influence the production structure of enterprises. Emissions are significantly related to the ratio of capital to labor KL^[18]. Therefore, the composition effect function can be expressed as:

This paper also assumes that the scale of production is a function of COOR^[18], the scale effect function can be expressed as:

In combination with the Equations (14) - (17), the Equation (12) can be rewritten as:

Equation (18) shows that, the impacts of the coordinated development of two-way FDI on the carbon emission is realized by the scale effect, the composition effect, and the technique effect. The total effect depends on the relatively strong and weak relationship among three effects.

Since domestic R&D expenditure RDis an exogenous variable, the COORderivation is made on both sides of Equation (18) and multiplied by COORto obtain:    

Also, this paper assumes that there exists effect relation of the two-way FDI development on the per capita capital stockKL:. It is assumed that the labor supply is exogenous and this paper does not consider the effect of the two-way FDI on the intrusion or crowding of the domestic capital.Referring to Sheng and Lyu’s methodology, according to the Equation (19), the effect of the two-way FDI development on emissions can be written as ^[15]:                                                                 

In Equation (20),  and  is the capital output elasticity.

4. Econometric Model, Variable Definitions and Data Descriptions

4.1. Basic model and estimation method

Carbon emissions, as an externality factor in economic development, can transfer not only through atmosphere diffusions due to a change of natural climate conditions but also through industrial shifts and element flows in human activities, and generate the phenomenon of spatial transmission. Thus, carbon emissions may have a significant correlation effect in space^[42]. The omission of such spatial correlation will necessarily result in an error in the estimation of models and false parameter tests^[43]. This paper introduces a special weight matrix, andusesthe spatial econometric model to control the possible spatialassociation.

Expanding Equation (18) to construct the econometric model as follows:

where,  is a spatial weight matrix reflecting the spatial relationship among the units;  is a set of control variables;  represents the individual effect;  is a random disturbance term; both  and  are the spatial lag coefficients reflecting the space-dependent relation between variables;  is the spatial error coefficient reflecting the spatial relation existing in the random disturbance item. If both  and  areequal tozero, Equation (21) is a Spatial Error Model (SEM), which means that the spatial relations are influenced by unobservable factors in different regions, that is to say, the spatial correlation features are reflected in random perturbation terms;If both  and  areequal tozero, Equation (21) is a Spatial Lagged Model (SLM) which means that spatial relations come from explained variablebetween different regions; If  isequal tozero, Equation (21) is Spatial Dubin Model (SDM) which means that spatial relationships are not only derived from explained variables between different regions, but also from explanatory variables between different regions.The specific model settings are required to be further inspected.

Because of the spatial lag term  is included in the model, and mutual causality between dependent and independent variables exits, which leads to endogeneity problems. So that, the OLS, fixed effects, and random effects estimators have a large probability to be biased. But Arellano and Bond present a differential generalized moment estimation method (DIFF-GMM)^[44]that can solve those problems effectively.The authors suggestusingthe lagging term of endogenous explanatory variable as the instrumentalvariable of difference term to control the endogeneity. Therefore this paper choosesthe difference GMM to estimate Equation (20). As a consistent estimate, GMM requires instrumentalvariables to be strictly exogenous and highly correlated with endogenous variables, andthere is no auto-correlation in the disturbance term. These require Hansen or Sargan instrumental variable test and Arellano-Bondauto-correlation test^[45].

4.2. Construction of spatial weight matrix

The common spatial weight matrices include 0-1 adjacent weight matrix, the geographic distance weight matrix, and the economic distance weight matrix. Considering that the 0-1 neighboring weight matrix does not reflect spatial connections between individuals that are geographically adjacent but not bordering. Although the geographic distance weight matrix or the economic distance weight matrix respectively reflect the relationship between individuals from geospatial and economic behavior, the spatial correlation between actual regions may not only come from one aspect of geography or economy. Therefore, this paper adopts a nested form of geographic distance and economic distance, which not only takes the spatial influence of the geographical distance into account but also reflects the objective factof economic factor overflow in the inter-regions^[46].

This paper sets the spatial weight matrix as:

where  represents the proportion of the economic distance space weight to the geographical distance space weight. This paper chooses . is the economic distance weight matrix with the definition as:

In Equation (23),  is the economic distance between two regions, the difference of two regions’ per capita GDP is used as a proxy of in this paper. is the geographic distance weight defined as:

where  is the geographical distance between two capital cities of provinces. 

In addition, the global spatial auto-correlation tests are carried out for China's carbon emissions using Moran’s I and its scatter diagram to describe the overall spatial relationship among all units within the study scope.Moran’s Iis a measure of spatial autocorrelation developed by Pratrick Alfred Pierce Moran^[47],reflecting the correlation between the observed value and the spatial lagging term, that is, certain property value in one region is related to the same attribute value in nearby region^[48]. Moran’s I is defined as:

where  is the element of the spatial weight matrix.  , if the valueofMoran’s I is greater than zero, it means that there is apositive spatial correlation and vice versa.

4.3. Definitions of variables

The factors affecting carbon emissions are complicated. In this study, we study the effects at the provincial level because regional differences between provinces is an important issue that can't be ignored. To this end, in this paper, the variables with regional difference characteristics are selected as far as possible to describe the regional heterogeneity and to avoid the bias error caused by the model estimation due to the missing variable. The specific variables are defined as follows:

1.  Explained variables: Carbon emissions ()

The carbon emission data from 30 provinces, municipalities directly under the Central Government, and the autonomous regions are from China Emission Accounts and Datasets (CEADs)^[49,50]. It refers to the emissions generated in the process of combustion for 17 kinds of fossil fuels and cement production. CH4, N20 and other greenhouse gases are not included.The latest carbon emission data currently published by the database is only up to 2015. This paper uses the quadratic exponential smoothing algorithm and China's provincial carbon emission data from 2004 to 2015 to estimate the missing data for 2016. Carbon emissions in Chinese provinces show an obvious agglomeration phenomenon (Figure 2). Inner Mongolia, Shanxi, Henan, Hebei, Shandong, Liaoning, Jiangsu were the highest carbon emitting provinces in 2004, while Xinjiang, Gansu, Yunnan and Guangxi were the low ones. Similarly, in 2015, Inner Mongolia, Shanxi, Henan, Hebei, Shandong, Liaoning, Jiangsu were still the highest carbon emitting provinces, while Yunnan, Guangxi, Guizhou, Chongqing, Jiangxi and Fujian were the lowest ones, indicating that the carbon emissions in Chinese provinces show relatively stable gathering characteristics.

2.    Core explanatory variable: coordinated development level of two-way FDI (COOR)

This paper selects the actually utilized value of foreign direct investment by each province and overseas direct investment of each province to measure the flow of inward FDI and OFDI, respectively. The coordinated development level of IFDI and OFDI is calculated by Equations (13)-(14) with ① is usually considered as low-degree coupling coordination; ② is moderate coupling coordination; ③ is high-degree coupling coordination; ④ is extreme coupling coordination.. Figure 3, the ordinate representing the coupling coordination level, shows an imbalance phenomenon among regions for the coordinated development of China's two-way FDI, eastern regions of the two-way FDI’s coupling coordination level ranking the highest, central regions being the middle, western regions being the lowest.

3.     Other control variables

Output scale (S). Carbon emissions, as derivatives from production process, are directly linked to the scale of an economy's output. Carbon emissions in the production process will also increase with the expansion of the output scale. Gross domestic product (GDP) of each province in this paper is selected as the proxy variable of output scale. The real GDP is obtained using GDP index forconverting GDP at current price into constant price at the base year 2004.

Environmental regulation intensity (REGU). In general, the greater the government's investment in environmental pollution control, the stricter the punishment for pollution discharge against rules, the pollution discharge can be controlled to a certain extent to improve the environmental quality. This paper uses the proportionof investment value for environmental pollution control to provincial GDP as a proxy variable for environmental regulation intensity.

Capital-labor ratio (KL). This paper uses the ratio of fixed capital stock to the annual average employees in each province to measure the input composition of production factors for enterprises, where fixed capital stock is obtained using the perpetual inventory method for the base year 2004 with the depreciation rate of 9.6% taken from Zhang et al. (2004)^[51].

Industrial structure (ISS). The industrialization promotion will generate more emissions. At the early stage of economic development, extensive economic growth has a negative impact on the environment, the mode of economic growth will transfer into the intensive type with the economic development to a certain extent. In this paper, the proportionthat the value added of the second industry to GDP of each province is selected as the proxy variable for industrial structure.

R&D expenditure (RD). The impact of R&D expenditure on technological progress is directional and thus the impact on the environment is also uncertain. R&D expenditure includes investment to research and development of more advanced environmental technologies and cleaning equipment and also to increase productivity and production output. The proportion of the R&D expenditure to GDP of each province in this paper is selected as the proxy variable for R&D investment.

4.4. Data sources

This paper covers 30 provinces, autonomous regions, and municipalities directly under the Central Government (excluding Tibet) in Chinese Mainland and the period of 2004 - 2016. Table 1 shows the main data sources and related explanations. Table 2 presents the descriptive statistics of variables. Variables in absolute terms are transferred into logarithmic values to eliminate the differences in order of magnitudes.

Table 1．                                        Data description

Point 3: The results do make sense but they are not presented in a proper format. I would like to know how the results were similar to or different from other studies.

Response 3: 

We discussed the empirical results that were similar to or different from other studies in the session 6.1 Discussion of empirical results (in green).

6.1 Discussion

Under the background of paying equal attention to the “introduction of foreign investment and investment in foreign countries”, based on the analysis framework of Copeland and Taylor, this paper examines effects of China's spatial spillover andtwo-way FDI coordinated development on domestic carbon emissions whichis realized through three channels — the scale, composition, and technique effects. 

Carbon emissions in China have a significant spatial spillover effect and remarkable spatial agglomeration phenomenon among provinces. This result coincides with the existing studies using spatial econometric models^[42,64-66], which indicates that spatial correlation characteristics of environmental pollution are of existence. If researchers and policy makers ignore these linkages, it will likely lead to unreliable results and subsequently incorrect policy guidance.

The two-way FDI coordinated development has a significant effect on emission reduction. Previous studies only focused on effects of IFDI or OFDI on environment^[21,37,38], however this study is the first to conclude that the findings agree with China’s strategy of IFDI and OFDI coordinated development and guidance for China’s emission reduction, and thus sustainable development goal of opening up and environmental protection can be integrated effectively.

We found that the two-way FDI coordinated development has a positive scale effect on carbon effect, which is consistent with existing studies that the scale effect can increase emissions^[13-15]. Composition effect, which has a negative effect on emissions is different from previous studies, which considered that capital-intensive production will increase emissions,^[18,67]but it supports that higher capital/labor ratio implies more advanced technology level^[37], which will help in reducing emissions. The key driver of the emission reduction effect is due to the leading role of the technique effect, which has far exceeded the scale and the composition effect. Therefore, the total effect of the two-way FDI on carbon emissions is consistent with the direction of the technique effect.

Author Response File: Author Response.docx