Influence,of,local,environmental,target,constraints,on,enterprise,OFDI

DING Qi-nan , ZHOU Quan , GAO Hong-gui

1 Zhongnan University of Economics and Law, Wuhan 430073, CHINA

2 Northeastern University at Qinhuangdao, Qinhuangdao 066004, CHINA

Abstract: By collecting the environmental target constraints in the annual government work report of prefecture-level cities, this paper studies its impact on local enterprises’ OFDI. The main conclusions are as follows: (1) Both direct and indirect constraints have a significant positive impact on enterprises’ OFDI, and the degree of direct constraints is stronger than that of indirect constraints. (2) Environmental target constraints of local governments will aff ect corporate OFDI behavior by aff ecting production costs, local economic development level and local openness, and (3) There are regional heterogeneity and investment type heterogeneity in the impact of environmental goal constraints on corporate OFDI. The suggestions are as follows: government departments can appropriately strengthen the constraint intensity of environmental targets and formulate specific restraint rules for industries with diff erent levels of pollution, so as to eff ectively make use of the technology spillover eff ects brought about by OFDI to promote the upgrading of domestic industrial structure. When making OFDI, enterprises should clarify the investment motivation, strengthen corporate social responsibility, and make use of technology spillover eff ect to promote the upgrading and development of home country industry while being beneficial to their own development.

Key words: local environmental target constraint; environmental regulation; outward foreign direct investment

Since the reform and opening up, with the rapid growth of the economy, severe environmental problems have also come along. The Chinese government has issued many laws on environmental protection in order to further enhance the intensity of environmental protection and constantly strengthen the ability of environmental protection work. In the face of increasingly serious environmental problems, in order to achieve the task of environmental control, the government has implemented restrictions on the discharge of environmental pollutants. At the national level,the Ninth Five-year Planputs forward the requirements for the control of the total discharge of pollutants for the first time, and the binding targets for the discharge of national key pollutants are clearly stipulated inthe Eleventh Five-year Plan. In addition, local governments have included key pollution emission targets in government work reports since 2003, reflecting the confidence and eff orts of local governments in environmental control, and at the same time enabling local governments to effectively improve local environmental regulation, thus promoting the improvement of local environmental quality.

As one of the most typical global economic and trade activities, OFDI creates huge social and economic benefits for Chinese individuals and small and medium-sized enterprises, and is of great strategic significance to China’s economic and social development. Since the reform and opening up, China’s market economy has developed rapidly. By 2020, China’s OFDI flow ranks first among the countries (regions) in the world. In this process, foreign investment has made full use of overseas markets and resources, eff ectively promoted the development of local industry, and realized the upgrading and transformation of industrial institutions, which has led to a high level of environmental protection while high-quality economic development.

At present, the theories about constraint and OFDI mainly include “Porter Hypothesis” and “Pollution Haven”. But at present, most research results focus on the inflow of FDI (foreign direct investment) caused by the environmental regulation of the host country, but seldom study the relationship between OFDI and the home country environmental regulation, because OFDI can produce reverse technology spillover effect. Therefore, it will not only help domestic enterprises to overcome the shortcomings of environmental pollution emissions, but also help to enhance China’s global influence. It is thought-provoking to study the driving factors of Chinese enterprises’ OFDI. This paper will use enterprise-level data to test the actual impact of environmental target constraints on the OFDI activities of local listed enterprises in various prefecture-level cities. It is of certain practical significance to accurately predict its effect in terms of environmental targets in government work reports for the coming year, which is helpful for local governments to balance the relationship between environmental protection and OFDI.

At present, the research on China’s OFDI mainly discusses the motivation and influence of China’s OFDI. Among them, the common motivation theories of China’s foreign investment at home and abroad are market-oriented theory, resource-oriented theory and institutional promotion theory (Lu et al., 2016). Three of them are mainly oriented to the consolidation, expansion and development of foreign markets (Taylor, 2002; Deng, 2002), scarce natural resources or important strategic resources in the home country (Kolstadetal, 2009; Wang et al., 2014), and the acquisition of foreign low-cost factors of production. In the theory of institutional promotion, some scholars believe that China’s political and economic system and the superiority of OFDI policy have caused the overseas capital activities of domestic enterprises to a large extent (Pei et al., 2011). In addition, some scholars believe that the overseas capital activities of some Chinese enterprises are to a large extent an escape response to China’s domestic institutional constraints (Witt et al., 2007).

In addition to the study of motivation, some scholars believe that the technology spillover effect brought by OFDI is conducive to promoting the upgrading of industrial structure (Zhao et al., 2006; Zhou et al., 2013; Song et al., 2021). What’s more, Liu and Zhao believe that environmental regulations promote the reverse green technology spillover of OFDI from a national point of view, and there is regional heterogeneity.

Definitely, some scholars have discussed the impact of home country environmental constraints on domestic OFDI outflow. Most of these studies believe that the former plays a positive role in promoting the latter. Mariana (2007) studies the relationship between the intensity of environmental regulation in 25 European countries and the establishment of overseas subsidiaries of their enterprises, and concludes that with the improvement of the intensity of environmental regulation in their home countries, domestic enterprises, especially those in high-pollution industries, will carry out more OFDI activities. Based on the inter-city panel data of South Korea, Xu (2017) concluded that the increase of the intensity of environmental regulation will lead to an increase in the OFDI of heavy chemical industries with high energy consumption and high pollution, while the impact of environmental regulation on high-tech industries with low energy consumption and low pollution is very small.

Besides, Zhou and Pang (2013) believe that from the perspective of investment location and industry, it is not realistic for polluting industries to transfer abroad through OFDI activities, and there may be no “Pollution Haven” effect here. Yin (2017) divides environmental regulation into formal and informal according to the two ways of government enforcement and government guidance. For formal environmental regulation, he believes that the higher the intensity of such environmental regulation in the central region, the more OFDI activities occur, while in the eastern and western regions, it is a significant inhibitory eff ect. Relatively speaking, informal environmental regulation plays a positive role in promoting OFDI in the central and western regions, while the impact of the eastern region is not significant. Borghesiet al.(2015) believes that it depends on the strictness of China’s environmental regulations whether China’s overseas investment will change the mentality and practice of Chinese companies engaging in green business.

At present, the research on environmental issues and FDI and OFDI has been rich, but on the basis of the existing theory and literature, there are still several aspects worth discussing: first, compared with the host country environmental constraints on Chinese enterprises’ OFDI, the related impact of home country environmental constraints may be more direct, so the relevant research from this direction will be carried in this paper as to further explore the motivation of China’s OFDI. Second, most of the current studies on environmental constraints are based on the intensity of environmental regulation, but there are many methods for the indicators of the intensity of environmental regulation. Therefore, selecting a more stable and systematic index to measure environmental constraints is also helpful to enhance the accuracy of the research. Third, few literatures focus on policies and OFDI data at national and provincial level, but lack of research on OFDI at prefectural and municipal level and enterprise level, and there is little research on the depth and breadth of foreign investment.

3.1 Analysis of mechanism at the enterprise level

First, from the perspective of prefecturelevel cities, when the local economy is relatively backward, people pay more attention to economic issues and ignore environmental problems. With the increase of income, environmental problems will be given highlight. The environmental target constraint of each prefecture-level city will aff ect the local economic growth to a certain extent (Yu, 2020). The existence of environmental pollutant restraint policy can improve the virtuous circle between national economic development and environmental protection. The absence of goal constraints may make the local to pay less attention to environmental protection, resulting in economic growth at the cost of environmental pollution, thus the conflict become more obvious. What’s more, consumers’ environmental awareness is conducive to local economic growth, and economic growth

From the perspective of micro-enterprises, if environmental targets are put forward for the government, enterprises will bear additional production costs at this time, which may have a relevant impact on enterprises’ OFDI decisions.

First, enterprises choose to turn to foreign areas with relatively loose environmental constraints to carry out production activities, so as to avoid the additional environmental costs caused by environmental constraints. At this time, environmental goal constraints promote enterprises to carry out OFDI activities. Second, enterprises choose to bear additional environmental costs and continue to produce locally, which does not promote OFDI activities, and enterprises may not be able to make sufficient investment in R&D and innovation activities because of environmental costs. Meantime, foreign environmental constraints may not be conducive to enterprises to carry out OFDI activities. Third, in order to make up for the loss of profits caused by environmental constraints imposed by the domestic government, enterprises may actively open up overseas markets, including exports, cross-border mergers and acquisitions, etc., as the process of enterprise globalization is a change from simple to complex, export enterprises will increasingly prefer to carry out OFDI (Ji, 2015). Therefore, it will increase the possibility of carrying out OFDI activities (Zhang, 2008; Wu, 2021).

3.2 Analysis of the mechanism at the prefecture level

is also flexibly related to the local pollution control tax rate (Yang, 2019). The higher the level of economic and social development of the investing country, the larger will be the scale of its OFDI. In addition, the economic situation of the home country can promote OFDI growth through macroeconomic growth and industrial structure transformation (Li, 2018). Therefore, by the same token, when the economic growth level of prefecturelevel cities increases, it will also play a role in promoting overseas investment for local enterprises.

Second, the amount of foreign investment attracted by various regions will also affect the foreign investment decisions of local enterprises. The high degree of opening to the outside world has played a role in promoting local OFDI (Zheng, 2012). On the one hand, the increase of foreign investment will bring a certain competitive eff ect to local enterprises, and the entry of foreign capital will promote the development of local economy, expand the industrial demand base, generate economies of scale, and thus enhance the competitive advantage of host enterprises, and eventually it can promote the scale of foreign investment. However, when the effect of agglomeration competition is too large, it may also lead to the “crowding out” eff ect of local enterprises, which is not conducive to the expansion of OFDI. On the other hand, the increase of foreign investment will also bring certain technology spillover effects. The embedding of foreign R&D brings advanced technology and management experience to the enterprises of the host country, and promotes the improvement of their factor endowment level and the internal structure of the company. It can also cultivate and enhance the international competitive advantage of enterprises in the host country, so as to promote enterprises to go out smoothly. At the same time, the external information and the way of international profit also enhance the behavior of host country enterprises to make OFDI (Ji, 2020).

Based on the above theoretical basis and mechanism analysis, this paper puts forward the following hypotheses:

Hypothesis 1: environmental goal constraints have a certain impact on enterprise OFDI.

Hypothesis 2: environmental goal constraints will have an impact on enterprise OFDI by aff ecting enterprise production costs.

Hypothesis 3: environmental goal constraints will aff ect the OFDI of local enterprises by aff ecting the level of local economic development and openness to the outside world respectively.

4.1 Model setting

4.1.1 Basic regression model

In order to investigate the impact of local government environmental target constraints on OFDI, the econometric model is constructed as follows:

Where: lnofdiijtrepresent the OFDI flow,iandjrepresent enterprises and cities respectively, andtdenotes the year;Djtrepresents the environmental target constraint and is measured using direct constraint (D1jt), indirect constraint (D2jt) and no constraint (D3jt) respectively;Controlrepresents the control variable;μijtis the enterprise fixed effect;δijtis the city fixed eff ect;θijtis the year fixed eff ect; andεijtis the random disturbance term.

4.1.2 Intermediary Eff ect Model

In order to study the intermediate mechanism of the influence of environmental goal constraints on OFDI outflow, this paper mainly refers to the method of Liet al. (2020), and constructs the regression model as follows:

Where:Xijtrepresents the mechanism variables, and the production cost of enterprises (lncost), the level of economic development (lnRgdp) and the degree of openness to the outside world (Open) are selected in turn as proxy variables for testing respectively.

4.2 Variable description

4.2.1 Explained variables

OFDI flow (lnofdi): this paper mainly takes the annual OFDI flow data of enterprises as the explained variable. The data are mainly from the overseas direct investment database of listed companies in the CSMAR database, which has been screened out for three years of OFDI discontinuous enterprises, ST enterprises and tax havens registered in the whole island of Cayman. In addition, this paper uses three different indicators: stock of OFDI (lnstock), depth of OFDI (shendu) and breadth of OFDI (guangdu) in the robustness test. The depth of OFDI (shendu) is mainly measured by the number of overseas subsidiaries owned by the enterprise in a given year with reference to the research of Marano (2016), and the breadth of OFDI (guangdu) is measured by the number of countries / regions where the enterprise’s overseas subsidiaries are located in a given year with reference to the research of Vermeulen (2002).

4.2.2 Core explanatory variables

Environmental target constraint: this paper manually extracts and collates the specific environmental target information set in the “Government Work Reports” of various cities and the specific completion of the environmental targets of the public report. On the basis of Li’s (2020) classification of environmental goal constraints, specific values such as industrial wastewater emission reduction and comprehensive utilization rate of solid waste mentioned in the government work report are also included in the direct constraint; otherwise, it will be included in the unrestrained. Finally, the environmental target constraint of the enterprise is obtained by matching the target constraint of the prefecture-level city and the city itself.

The improved environmental target constraint is still divided into three types: direct constraint, indirect constraint and unconstraint. Among them, if there are clear constraint indexes, such as sulfur dioxide, nitrogen oxide, industrial wastewater and solid waste comprehensive utilization rate, and a detailed established reduction or percentage of pollutants, it is a direct constraint; if there is no clear sulfur dioxide and nitrogen oxide index, but having days of good air quality, PM2.5 or PM10 concentration indicators, it is an indirect constraint; if there is no such index, there is no restriction.

Therefore, the environmental constraint objectives in different cities can be divided into three cases: direct constraint (D1), indirect constraint (D2) and unconstrained (D3). Among them, if a city has direct constraints, D1 will be assigned to 1; if there are indirect constraints, D2 will be assigned to 1; if there are no clear environmental constraints in the city government work report, D3 will be assigned to 1. It should be noted that this paper diff ers from Li (2020) in that D1 and D2 are both assigned a value of 1 when both direct and indirect constraints are proposed in the government work report of a region.

4.2.3 Mechanism variables

Enterprise production cost (lncost): this paper selects the average main cost of the enterprise as the agent variable of the enterprise production cost (lncost), because the main cost is the most important part of the enterprise production cost, which can reflect the production cost of the enterprise to a certain extent.

Level of economic development (lnRgdp): the level of national economic development usually refers to the growth of total output or per capita output of a country or region, in order to reflect the change of a country’s economic strength or the improvement of people’s living conditions and social and economic development level. This paper mainly selects the logarithm of the annual growth rate of the total output of each prefecture-level city.

Degree of openness (Open): this paper mainly reflects the degree of openness of the region through the sum of the degree of dependence on foreign trade and the degree of dependence on foreign investment, in which the degree of dependence on foreign trade can show the degree of integration of the region’s economic and trade activities with the global market, which is expressed by the ratio of the total import and export trade of the region to the GDP of the same period in a certain year. The degree of dependence on foreign investment reflects the degree of openness in the field of OFDI, which is expressed by the ratio of the total OFDI in this region to the GDP of the same period in a certain year.

4.2.4 Control variables

Other control variables used in this paper include: level of scientific development (Keji): representative of science and technology expenditure of prefecture-level municipal governments in current years; investment rate (Invest) is the ratio of investment in fixed assets to GDP in that year; population density (Midu) is divided by the total population at the end of the year by the land area of the administrative area. Period expense rate (qjfyl) is the ratio of period expenses to operating income, in which period expenses include management, sales and finance expenses; enterprise size (qygm) is using the logarithm of the total assets of the enterprise to measure; proportion of independent directors (dsbl) is using the proportion of independent directors on the board of directors of listed companies to express; corporate assetliability ratio (zcfzl) is using the total liabilities / total assets of the enterprise to measure.

4.3 Data sources and descriptions

The data in this paper are obtained from the China City Statistical Yearbook, the CSMAR database and the government work reports of each city. The article mainly collected the government work reports of 284 prefecture-level cities nationwide from 2011-2016 by hand, and mainly consulted the descriptions of environmental target constraints in them and sorted them out relatedly, and the relevant results are shown in Table 1; the descriptive statistics of other main variables are shown in Table 2.

5.1 The impact of environmental goal constraints on enterprise OFDI

This part is mainly about the empirical analysis of the impact of environmental target constraints on OFDI. Table 3 reports the results of the benchmark regression. Among them, (1) (2) are the research results of the direct constraints of prefecture-level cities on the OFDI of local enterprises when controlling the urban fixed effect and firm fixed eff ect and then adding the annual fixed eff ect, respectively, and (3) (4) and (5) (6) are indirect constraints and unconstrained eff ects on enterprise OFDI respectively.

Table 1 Environmental target constraints in 284 prefecture-level city government work reports

First of all, according to the two-column results of (1) (2), we can see that the influence coefficient of direct constraint (D1) on enterprise OFDI flow increases significantly after increasing the fixed effect of the year, and the coefficient also increases, and finally it is 0.262, and through the significance test of 1%, it shows that the formulation of direct constraint by the municipal government promotes the OFDI of local enterprises. The reason is that the direct constraint mainly focuses on the reduction of specific pollutants, which urges local governments to choose imperative environmental regulation tools in order to achieve their goals, and enterprises will face the upper limit of punishment. At this time, relative to the choice to pay the environmental cost to continue production and turn to other parts of the country. Enterprises are more inclined to transfer abroad, actively open up new markets, seek technological innovation compensation effects and reverse technology spillover effects to continue production, thus increasing their OFDI.

Table 2 Descriptive statistics

Table 3 Results of the impact of environmental target constraints on corporate OFDI flows

Secondly, according to the results of (3) (4), we can see that after adding the year fixed eff ect, the influence coefficient of indirect constraint (D2) on enterprise OFDI flow decreases, and the significance also decreases, and the final coefficient is 0.096, which is relatively smaller than the direct constraint coefficient, and only passes the significance test of 10%, indicating that the formulation of indirect constraints by municipal governments also promotes the OFDI of local enterprises. However, compared with the direct constraint, the indirect constraint does not clearly specify the industrial pollutant emission limit of the enterprise, so the regulation ability of the enterprise is not as good as the direct constraint, so its influence on the enterprise OFDI behavior is not as good as the direct constraint.

Finally, according to the two-column results of (5) (6), the influence coefficient of unconstrained (D3) on enterprise OFDI flow also decreases after adding the year fixed effect, and finally it is-0.128, and passes the significance test of 5%. Its coefficient is negative, which to a certain extent means that the unrestrained implementation of the municipal government reduces the possibility of local enterprises’ outbound investment. The reason of restriction absence means that the government does not carry out strict supervision or low supervision over high pollution emission enterprises. When they are under weak regulatory pressure, the cost of local production decreases or does not increase, they choose to obtain technological innovation through OFDI and may face other costs, thus reduce OFDI and continue to choose local production.

5.2 Robustness test

This part mainly considers the robustness of the estimation results from two aspects: the selection of indicators and the possible endogenous problems. On the basis of literature review, this paper measures enterprise OFDI through three diff erent indicators: OFDI stock, OFDI depth and OFDI breadth (in which OFDI depth and breadth are enterprise-level data) and regression, the final results are basically consistent with the benchmark regression results. In addition, because of the strict exogenous relationship between the total number of prefecture-level cities in the province where each prefecture-level city is located and the OFDI behavior of enterprises, this paper takes it as a tool variable of environmental target constraint and makes a 2SLS regression between environmental target constraint and enterprise OFDI. The regression results show that the weak tool variables and unidentifiable original assumptions are rejected, and the influence results are consistent with the previous ones, so as a whole. Different types of environmental goal constraints have a relatively robust eff ect on OFDI outflow, and the detailed results are shown in the appendix.

5.3 Mechanism test

This part mainly takes the enterprise production cost, economic growth and the degree of openness to the outside world as the intermediate mechanism. Because each mechanism variable may have different effects in different types of environmental target constraints, it is divided into three parts here.

First of all, Table 4 shows the results of the mechanism test and analysis when the enterprise is faced with the direct constraints of the prefecturelevel municipal government. Among them, column (1) is listed as the benchmark regression result, column (2) and (3) are listed as the analysis of production cost mechanism, in which column (2) can see that the production cost of enterprises will rise when facing the direct constraints of prefecturelevel city environment. Column (3) shows that after adding the production cost variable, the coefficient is significantly positive, and the direct constraint coefficient decreases. This also reflects from the side that the direct constraint of the diversion of enterprise production cost plays an important role in promoting OFDI. Columns (4) and (5) are the analysis of the local economic development mechanism. We can see that direct constraints will promote the economic growth of prefecturelevel cities; the level of economic development of prefecture-level cities will also promote local enterprises to carry out foreign investment activities when enterprises are faced with direct constraints. At the same time, the influence coefficient of direct constraints also decreases, which effectively tests the effect of this mechanism. Columns (6) and (7) are the analysis of the mechanism of the level of opening to the outside world, which shows that direct constraints will restrain the local level of opening to the outside world, and the level of opening to the outside world has a significant positive impact on their foreign investment activities when enterprises are faced with direct constraints.

Table 4 Mechanism test-direct constraint group

Secondly, Table 5 shows the results of mechanism test and analysis when enterprises are faced with indirect constraints of prefecture-level municipal governments. The meaning of the contents shown in each column is consistent with that in Table 4, in which column (2) shows that the production cost will also increase when facing the indirect constraints of prefecture-level city environment, and column (3) shows that after adding the production cost variable, the coefficient is significantly positive, and the indirect constraint coefficient and its significance are reduced, at this time, the diversion eff ect of enterprise production cost on indirect constraints is more obvious. As can be seen in column (4), indirect constraints will promote the economic growth of prefecture-level cities, and when enterprises are faced with indirect constraints, the influence coefficient of the level of economic development of prefecture-level cities on their outbound investment activities is significantly positive; column (6) shows that indirect constraints will inhibit the level of local opening to the outside world, and column (7) shows that when enterprises are faced with indirect constraints, the level of openness has a positive impact on their outbound investment activities, but it is not significant.

Finally, Table 6 shows the results of mechanism test and analysis when enterprises are faced with unrestrained municipal governments at the prefecture level. The meaning of the contents shown in each column is consistent with that in Tables 4 and 5, where column (2) shows that the production cost of enterprises will decrease when they are not faced with the environ-mental targets of prefecture-level cities, while column (3) shows that after the addition of production cost variables, the coefficient is significantly positive, and the eff ect of unconstrained influence is significantly reduced, at this time, the shunt eff ect of enterprise production cost on the impact of indirect constraints is more obvious. As can be seen in column (4), when a place

does not set environmental target constraints, the economic development of the prefecture-level city will show a downward trend, and when enterprises are faced with no constraints, the influence coefficient of the economic development level of prefecture-level cities on their foreign investment activities is significantly positive. Here the inhibitory effect of unrestrained on enterprises’ OFDI is also relatively weakened. Column (6) shows that the influence coefficient of unrestraint on the level of local opening to the outside world is positive, but not significant. Column (7) shows that when enterprises do not face constraints, the influence coefficient of the level of opening to the outside world on their outbound investment activities is positive, but not significant, and the logic consequence is that the inhibitory effect of unrestrained on enterprises’ OFDI is weakened.

Table 5 Mechanism test-indirect constraint group

Table 6 Mechanism test- Unconstrained group

On the whole, whether it is the production cost mechanism at the enterprise level, the economic development level mechanism at the prefecture level and the opening degree mechanism at the prefectural level, the effect is more obvious, that is, hypothesis 2 and hypothesis 3 have been verified.

Table 7 Coastal area impact results

5.4 Heterogeneity test

The above empirical methods have confirmed the impact of diff erent types of environmental goal constraints on corporate OFDI outflow, and tested its mechanism. Furthermore, the following will analyze whether there are differences in different regions and different types of foreign investment enterprises.

5.4.1 Regional heterogeneity test

This part is mainly the heterogeneity test of the influence of environmental goal constraints on OFDI outflow in diff erent regions. The article is mainly divided into two groups: coastal areas and inland areas. Table 7 shows that both direct constraints and indirect constraints have a positive effect on enterprise OFDI outflow in coastal areas, and their coefficients are 0.252 and 0.159 respectively, in which the promoting effect of direct constraints is stronger than that of indirect constraints. Constraint absence has a significant negative effect on OFDI outflow. Table 8 shows that, in inland areas, except for the significant positive eff ect of direct constraint on enterprise OFDI outflow, the other two kinds of constraint coefficients are not significant. It may be due to the better ecological environment in the central and western regions where heavy industry has lower development than coastal areas, so the local environmental policy is relatively loose, and the impact on enterprise OFDI outflow is not deep. Therefore, it can reflect that environmental goal constraints have diff erent effects on OFDI outflow in the two groups of regions. Tests show that there is regional heterogeneity in the impact of environmental goal constraints on OFDI outflow.

Table 8 Inland area impact results

5.4.2 Test of heterogeneity of types of outward investment

To a certain extent, the industry distribution of foreign investment can also reflect the motivation of Chinese enterprises to invest abroad. For a long time, the distribution of Chinese enterprises in the industry of OFDI is relatively concentrated. By the end of 2020, the top five stocks of OFDI in China’s non-financial industries are leasing and business services, wholesale and retail, information transmission, software and information technology services, manufacturing and mining. The stock of OFDI in these five major industries has reached 80% of the total. According to the information content of “business scope of overseas enterprises” in the OFDI list, this paper divides Chinese enterprises’ foreign investment into four investment types of trade service investment, manufacturing investment, resource development investment and diversified investment. We analyzes whether the impact of environmental target constraints on OFDI will be diff erent in diff erent investment types of enterprises. The results are shown in Tables 9, 10 and 11.

Table 9 shows that when using direct constraints to separate regression for enterprises of diff erent investment types, it is found that the coefficients of column (1), (2) and (4) are significantly positive, while the significant degree and size of column (3) and (5) are decreased. This shows that the direct constraints on environmental targets put forward by prefecture-level cities can improve the OFDI probability of business service-oriented, manufacturing and diversified investment enterprises, while the impact on resource development and other types is relatively weak. In Tables 10 and 11, it can be found that whether indirect or unconstrained, the impact on all types of outward investment is consistent with the benchmark regression results, compared with their coefficients, the impact on resource development and other types of enterprises is significantly lighter than that of the other three types of enterprises.

The possible reasons for the above diff erences are as follows: first, the behavior of investment in resource development depends on the resource wealth of the host country, and the main purpose of the investment is to make use of the cheap natural resources of the host country. This type of investment is generally based on the requirements of national strategy or industry development. Other types of enterprises are mainly securities investment, consulting and other businesses, so they have little to do with the direct constraints of the domestic environment, indirect constraints and non-constraints still have a certain impact. Second, when the local government puts forward relevant environmental constraints, enterprises may transfer their production bases and go to other regions or countries with more relaxed environmental constraints to carry out production activities, or they may choose to set up subsidiaries or cooperation centers overseas. The types of investment in manufacturing, trade services and diversified investment are the latter, so the impact on these three types of investment is negative and significant when there are environmental constraints. When there is no obvious constraint, it shows a positive eff ect.

This paper studies the impact of environmental target constraints on corporate OFDI outflow from the perspective of home country. On the basis of current situation analysis, theoretical basis and mechanism analysis, this paper takes listed companies as the research object and selects the OFDI data of listed companies in CSMAR database to construct a multiple fixed effect model. Through empirical analysis of full samples and sub-samples of diff erent categories, benchmark regression and heterogeneity test are carried out. In addition, the relevant mechanisms are tested. The following conclusions are drawn:

Table 9 Investment type heterogeneity - direct constraint group

Table 10 Investment type heterogeneity - indirect constraint group

Table 11 Investment type heterogeneity - unconstrained group

First, both direct and indirect constraints in environmental goal constraints have a positive eff ect on enterprise OFDI. Among them, the influence of direct constraint on enterprise OFDI is deeper than that of indirect constraint, which may be caused by its direct restriction on industrial pollutants, and the results of both are significant. Constraint absence has a negative eff ect on enterprise OFDI outflow, and the result is significant, which indicates that enterprises can choose to reduce their OFDI behavior because they do not have to bear additional environmental governance costs. The effect of environmental target constraints on enterprise OFDI outflow shown in the above benchmark regression is roughly the same as that in the robustness test, and the eff ect of direct constraints on enterprise OFDI outflow is the most robust.

Second, the effect of environmental target constraints on enterprise OFDI outflow is mainly realized through the following enterprise-level mechanisms and prefecture-level city-level mechanisms. One is the mechanism at the enterprise level, in which environmental target constraints affect the OFDI behavior of enterprises by aff ecting their production costs, and both direct and indirect constraints will lead to an increase in production costs. Under the circumstance, enterprises choose to carry out cross-border production transfer activities in order to avoid the additional environmental compliance costs brought by domestic environmental constraints, and the influence of direct constraints is more intense. When enterprises face constraints absence, their production costs will be reduced, and their OFDI behavior will be reduced. The second is the mechanism at the prefecturelevel city level, which includes the level of economic development of prefecture-level cities and the degree of openness to the outside world. The test results show that constraints will promote the level of local economic development and restrain the degree of local openness to the outside world. The former will promote the OFDI outflow of local enterprises, while the latter diverts the impact of environmental goal constraints on corporate OFDI outflow.

Third, the impact of environmental goal constraints on corporate OFDI has a certain regional heterogeneity and investment type heterogeneity. On the one hand, though the impact coefficients of the three types of constraints on enterprise OFDI are all in the same direction in different regions, the local government’s environmental management is relatively loose compared with the coastal areas and the impact of environmental target constraints on OFDI outflow is not significant, which contribute to the well ecological environment in the central inland area and the relatively backward development of heavy industry. While, in the coastal areas, both direct and indirect constraints have a significant positive eff ect on OFDI outflow, and the negative eff ect of constraint absence is not significant. On the other hand, prefecture-level cities propose that environmental goal constraints can improve the OFDI probability of business serviceoriented, manufacturing and diversified investment enterprises; nevertheless, the impact on resource development and other types is relatively weak.

How to achieve win-win progress between ecological and environmental protection construction and economic and social development has always been an important issue in our country. How to achieve environmental constraints and OFDI to give full play to the role of a bridge between highquality economic development and high-level environmental protection, this paper puts forward the following policy recommendations. The first is to strengthen the intensity of environmental regulation of government departments and implement differentiated environmental constraints. Intensity of environmental constraints on inland areas should be deepened continuously. Governments at all levels should implement diff erential constraints on industries with diff erent levels of pollution by strengthening the adjustment of local industrial structure, speeding up the elimination of backward production capacity, and promoting the upgrading of foreign investment structure. The second is to increase the rational allocation of resources by enterprises to promote the green development and transformation of enterprises. From a longterm point of view, when enterprises are faced with strong environmental goal constraints, they can improve their green production efficiency by increasing domestic R & D investment, for another they can also choose to carry out OFDI activities to introduce advanced green environmental protection technology from abroad to achieve the positive effect of protecting the environment. Therefore, enterprises must establish a new concept of development, strengthen corporate social responsibility, transform the production line green, adjust the industrial structure, and improve the green total factor productivity of enterprises.