The Moderating Role of Host Investment Environments on the Relationship between Enterprises’ OFDI and Green Innovation: Evidence from China

Abstract

This study investigates whether and how outward foreign direct investment (OFDI) affects green innovation using a sample of Chinese A-share listed companies from 2006 to 2020. We find that OFDI significantly promotes green innovation, not only for gaining legitimacy, but also to mitigate liabilities of foreignness and cultivate competitiveness. The result is robust for endogeneity and among other robustness tests. Moreover, host economic development, environmental regulation, and national governance positively moderate the relationship between OFDI and green innovation, indicating that the heterogeneity of host investment environments effectively affects OFDI enterprises’ motivation to make green innovation. Further analysis shows that OFDI promotes green innovation more significantly in state-owned and less-overseas-experienced enterprises, heavy pollution industries, and areas with strict environmental regulations in the home country. This study introduces the heterogenous host investment environments into the relationship between OFDI and green innovation, and examines its moderating role, which reveals the mechanism of OFDI affecting green innovation. This study also provides a valuable reference for guiding enterprises’ internationalization and resisting overseas investment risks.

1. Introduction

In response to the global climate problem, countries around the world are accelerating the formation of the global climate governance pattern and building the global environmental and climate governance system. The Paris Agreement specifies the “hard targets” of environmental protection, ensuring sustainable development, and making unified arrangements for global action against climate change. China has proposed the carbon peaking and carbon neutrality goals of “reaching the peak of carbon by 2030” and “achieving carbon neutrality by 2060” to deal with the climate problem.

Since the reform and opening up, China’s policy of opening up to the outside world has undergone a change from “bringing in” to “going out”, and the total amount of foreign direct investment has been increasing steadily. In 2020, China’s outward foreign direct investment (OFDI) flow reached $153.71 billion, surpassing the United States for the first time and ranking first in the world. OFDI has become an efficient economic driver in China’s economic growth. By the end of 2020, China had 28,000 domestic investors with 45,000 OFDI enterprises in 189 countries (regions) around the world.

Chinese enterprises are widely engaged in OFDI globally. The international community’s growing focus on environmental performance makes Chinese OFDI enterprises receive more pressure and face a sharp increase in external uncertainty and environmental risks. OFDI faces not only liabilities of foreignness in host countries but also faces severe international challenges and stakeholder pressure [1,2,3]. Then, the core of this paper is how Chinese OFDI enterprises deal with the increasing environmental concerns and whether the heterogeneity of host investment environments influences enterprises’ performance.

Strategic CSR could be an effective way that leads enterprises to be different from competitors in a way that lowers costs or better serves a particular set of customer needs [4]. Better CSR performance is helpful to gain legitimacy [5,6,7,8,9], win the recognition of consumers and the public [10], as well as overcome the liability of foreignness [11,12]. Therefore, firms could take green innovation as a strategic social responsibility tool to mitigate host concerns about enterprises’ environmental performance and alleviate the increasing uncertainty and risk in host countries. Moreover, the existing literature does not form a unified conclusion about the relationship between OFDI and innovation [13,14,15,16]. It is still insufficient in the study of heterogeneity of the host investment environment on the relationship between OFDI and green innovation.

Using OFDI and host investment environments data of Chinese A-share listed companies in SSE and SZSE from 2006 to 2020, this paper empirically examines the effect of OFDI on green innovation and explores the moderating role of host investment environments on the relationship between OFDI and green innovation. The study finds that OFDI significantly promotes green innovation, not only for gaining legitimacy, but also to mitigate liabilities of foreignness and cultivate competitive advantages. The host economic development, environmental regulations, and national governance significantly positively moderate the promotion effect of OFDI on green innovation. It implies that the heterogeneity of the host investment environment effectively affects OFDI enterprises’ motivation to make green innovation and thus has a differential effect. Further analysis shows that OFDI promotes green innovation more significantly in state-owned and less-overseas-experienced enterprises, heavily polluting industries, and areas with strict environmental regulations in the home country.

Compared with the existing studies, there are several contributions of our study to the literature. First, this paper makes an in-depth study of OFDI, which is one method of internationalization and enriches the economic consequences related to corporations’ OFDI from the perspective of green innovation. Previous studies mainly based on the macro level, such as the national or regional level, to explore the influencing factors of OFDI [17,18,19] and its economic consequences related to technological innovation and green performance [20,21,22]. Distinctly, we focus on the micro firm level and examine the effect of enterprises’ OFDI on green innovation. On the one hand, it is of great practical significance to explore the green innovation effect of OFDI for improving the global environmental governance system. On the other hand, green innovation could serve as an emerging strategic CSR behavior and provide a new theoretical logic to explain the behavioral decisions of enterprises’ OFDI.

Second, it expands the research related to the influencing factors of green innovation from the perspective of OFDI. While previous studies have mainly examined the impact of domestic or developed countries’ environmental regulations, and different environmental regulatory instruments on green innovation [22,23,24], this paper extends it to an international perspective and explores OFDI as an important factor influencing green innovation performance. We extend the influencing factor of green innovation from national environmental regulations and market mechanism policies to global factors and country differences. In the context of global countries joining hands in environmental governance, it provides a new perspective for studying firms’ innovation behavior decisions and green innovation motivations.

Third, the perspective of the host investment environment is introduced to explore the mechanism of the role of enterprises’ OFDI in influencing green innovation. Previously, no consistent conclusions have been obtained on the effects of OFDI on economic consequences, such as firm value, firm performance, and investment and financing behaviors [13,14,15,16,25], probably due to the neglect of the differential effects of host investment environments on research findings. This paper further analyzes the mechanism of the effect of enterprises’ OFDI on green innovation from the perspective of host investment environments. We find that host economic development and institutional environments have a moderating effect, implying that firms adopt different green innovation strategies in host countries based on different intrinsic motivations to cope with overseas investment risks. The findings of this paper enrich the research on the mechanism of enterprises’ OFDI affecting green innovation.

The rest of the paper is organized as follows: The second part presents a review of the literature; the third part is the theoretical analysis and hypothesis; the fourth part is the research design; the fifth part is the empirical results and analysis; the sixth part is the further analysis; and finally, the last part is the discussion and implications.

2. Literature Review

2.1. Economic Consequences of OFDI-Related Studies

A large number of research studies have studied path choice [18], entry modes [19], and market selection [17,26,27] of Chinese enterprises’ internationalization statically or dynamically [28,29] and their impact on innovation and economic development. Some studies found that the reverse technology spillover effect of foreign direct investment not only improves the total factor productivity (TFP) of the home country but also promotes the scientific and technological progress and economic growth of the home country in the long run [20,21]. Using macro, middle, and micro-level data, studies verify the positive effect of OFDI on TFP and innovation development of the home country [13,15,16,30,31]. Based on Chinese manufacturing enterprises, Bai [22] found that OFDI in developed countries can generate reverse green technology transfer and promote environmental innovation of the parent enterprise. However, Bitzer and Gorg [14] concluded that, for an aggregate of 17 OECD countries, OFDI had a negative effect on domestic TFP. The early study based on the Chinese data before 2008 did not find a positive reverse spillover effect by OFDI as well [32,33].

In this regard, the heterogeneity of the host country is considered as a reason for whether OFDI can produce a reverse spillover effect [34,35,36]. Global R&D innovation exchange capacity is highly concentrated in developed economies, while emerging and developing economies have limited innovation capacity and lack investment in cutting-edge science and technology [37]. Investing in developed economies will create more opportunities to acquire cutting-edge technologies [38]. Therefore, the impact of OFDI on the green innovation of enterprises may also vary with the heterogeneity of the host investment environment.

2.2. Influencing Factors of Green Innovation

Green innovation is an important way to achieve low-carbon economic growth and fulfill the commitment to the Paris Agreement. Green innovation consists of new or modified processes, techniques, practices, systems, and products to minimize the use of natural resources per unit of output, the release of toxic substances, and enhance product quality. Many green innovations combine environmental benefits with a benefit for the establishment or user [39,40,41]. Based on strategic CSR [4], for any company, the strategy must go beyond best practices. It is about choosing a unique position—doing things differently from competitors in a way that lowers costs or better serves a particular set of customer needs. For enterprises, green innovation is conducive to forming a unique competitive advantage and reducing environmental costs. Green innovations help enterprises to gain more stakeholder support, strengthen their ties with customers, and ultimately realize the mutual benefits between enterprises and society.

As one of the most important drivers of green innovation, based on the hot issue of whether environmental regulation can induce green technology innovation, relevant domestic and foreign studies mainly have the following views: First, environmental regulations crowd out green innovation inputs due to increasing enterprises’ pollution control costs, which in turn hinders green technology innovation [42,43]. Second, environmental regulations generate innovation compensation effects and promote the innovation and diffusion of green production by enterprises [23]. According to the Porter hypothesis, reasonable environmental regulations can stimulate innovative production technologies and production processes and optimize resource allocation, thus stimulating the “innovation compensation” effect of enterprises and achieving the “win-win” effect of improving environmental quality and enhancing enterprise competitiveness [44].

Social norms and social environmental awareness are also among the elements that drive the development of green innovation in enterprises. Based on the power change theory and organizational information processing theory, both suppliers’ and customers’ green pressure have significant positive effects on green innovation behavior in the context of environmental uncertainty [25,45].

This paper enriches the existing literature in the following aspects: first, this paper explores the behavior and environmental performance of OFDI enterprises and the factors influencing green innovation based on the strategic CSR view and liabilities of foreignness. Second, the effect of OFDI on green innovation may vary with the heterogeneity of host countries, for example, economic development, environmental regulations, and national governance. Third, this paper explores the different effects of OFDI on green innovation from the perspective of enterprise, industry, and regional characteristics, and provides practical implications for the improvement of outward foreign direct investment policies by the government to promote the overseas management and sustainable development of enterprises under different scenarios.

3. Hypothesis Development

3.1. OFDI and Green Innovation

Enterprises’ outward foreign direct investment inevitably faces a high degree of uncertainty in the external environment, exposed to host investments and operational risks such as political, economic, trade, and socio-cultural barriers, and the expectations of a wider range of stakeholders such as communities, customers, and investors. Based on the strategic CSR view, green innovation not only helps to reduce environmental costs and strengthen the connection with customers but also enables enterprises to form a unique competitive advantage. Ultimately, it can realize the mutual benefits between enterprises and society. OFDI may enhance enterprises’ environmental responsibility performance and contribute to green innovation through the following mechanisms.

The first mechanism is gaining legitimacy and conducting risk management. Legitimacy is a critical resource for the survival and development of organizations. According to legitimacy theory [46], business organizations need society to grant them the right to exist, among other rights. Enterprises making overseas investments need to consider the influence of non-market factors such as host governments, communities, media, and the public on their legitimacy. Chinese companies making OFDI may face higher litigation risks due to violations of the host institutional regulatory or social norms. Unfamiliar host environmental regulations may raise the production costs of enterprises and exacerbate their risk of environmental violations. On this condition, green innovations are an excellent environmental protection behavior in front-end pollution control. It can more effectively promote pollution reduction, reduce firms’ violation risks, and improve their ecological performance, which effectively improves their green competitiveness. Positive ecological responsibility performance can help reduce the perceived risk to the outside and help companies gain the legitimacy to survive [23,47]. Thus, based on multiple institutional regulatory pressures from host governments and societies, especially in the face of growing environmental violation risks, OFDI firms are more likely to behave with a better green innovation performance to conduct risk management and seek legitimacy.

The second mechanism is alleviating the liability of foreignness and cultivating competitive advantages. The global focus on firms’ environmental performance could exacerbate the disadvantage of Chinese firms. As an outsider in host countries, OFDI firms have a natural disadvantage compared to local firms and usually face higher production and trade costs [1,48]. For this reason, OFDI firms try to improve their reputation to attract more overseas investors and improve their competitiveness. Under stakeholder theory, CSR activities provide several benefits to a firm’s stakeholders, including enhancing external stakeholders’ perception and recognition of the firm, mitigating the adverse effects of communication and psychological distance, and contributing to firm value. We posit that as a strategic tool for firms to distinguish themselves from other competitors, social responsibility has a positive effect on multinational enterprises to enhance their overseas image under the new normal. Based on reputation theory, firms respond to the demands of stakeholders by undertaking environmentally friendly activities to build a good firm image and improve firm reputation. Firms’ performance in green innovation transfers a signal to the outside that the firm is actively undertaking environmental responsibility. Therefore, green innovation, as a concrete manifestation of firms’ environmental responsibility fulfillment, helps firms to overcome liabilities of foreignness and cultivate competitiveness. Accordingly, we hypothesize that:

Hypothesis 1. 

Outward foreign direct investment positively promotes enterprises’ green innovation.

3.2. Heterogeneity of Host Investment Environments

Because there is little concern about the impact of host investment environments in the existing literature, there are inconsistent findings on the relationship between OFDI and innovation. In this paper, we will explore the heterogeneity of host investment environments in the relationship between OFDI and green innovation from three dimensions: economic development, environmental regulation, and national governance.

From the perspective of host economic development, the host economic development may drive its environmental awareness. According to the environmental Kuznets curve hypothesis [49], the relationship between environmental quality and per capita income shows an inverted “U” shaped relationship, i.e., at the early stage of economic growth, pollution emissions will increase with income growth. When income exceeds a certain level, environmental quality gradually improves, for example, urban pollution in developed countries declined after the 1980s compared to the previous decade. In addition, consumers in countries with higher levels of economic development prefer green products and are willing to pay a premium price for them. Consequently, host countries with higher levels of economic development are more environmentally conscious. In contrast, countries with lower levels of economic development have weaker social, environmental, or sustainability drivers for green innovation due to weaker government capacity and a lack of physical and technological capital. Thus, the promotion effect of enterprises’ OFDI on green innovation would be enhanced in host countries with a higher level of economic development.

From the perspective of the host environmental regulation, it directly affects the intensity of environmental governance. Environmental regulation is an important factor influencing green innovation, which may have a compensatory effect on innovation and thus promote the innovation and diffusion of green production. According to Porter’s hypothesis, reasonable environmental regulations can stimulate innovative production technologies and production processes and optimize resource allocation, thus stimulating the “innovation compensation” effect of enterprises and achieving the “win-win” effect of improving environmental quality and enhancing enterprise competitiveness [44]. In countries with stricter environmental regulations, OFDI enterprises have greater legitimacy pressure. And firms will be subject to serious legal or administrative penalties in case of non-compliance, facing higher risks. Stricter environmental regulations positively influence firms’ propensity to innovate environmentally. Thus, the promotion effect of enterprises’ OFDI on green innovation would be enhanced in host countries with stricter environmental regulations.

From a host national governance perspective, a country with stronger governance implies stricter government regulation and more efficient environmental enforcement, with more constraints on firms’ polluting behavior. High-quality state governance may also have a positive impact on corporate governance, promoting firms to do better in green innovation. Scott [50] suggests that regulatory pressures are mainly government administrative directives, binding requirements, or the coercive force of laws and regulations that drive firms to implement green innovation through coercive power. The institutional theory explains firms’ green innovation behavior driven by external pressures. In the face of higher governance efficiency and regulatory capacity in the host country, highly polluting and inefficient firms may bear higher environmental governance costs and be harder to gain institutional legitimacy. Thus, the promotion effect of enterprises’ OFDI on green innovation would be enhanced in host countries with stricter governance. Enterprises may be obliged by the stricter host governance to engage in green innovation to avoid the penalty associated with their illegitimate behavior and to seek institutional legitimacy. Based on the above analysis, we propose:

Hypothesis 2a. 

Host economic development positively moderates the relationship between OFDI and green innovation.

Hypothesis 2b. 

Host environmental regulation positively moderates the relationship between OFDI and green innovation.

Hypothesis 2c. 

Host national governance positively moderates the relationship between OFDI and green innovation.

4. Data and Research Design

4.1. Sample and Data Source

The initial research sample consists of A-share listed companies in SSE and SZSE from 2006 to 2020. We exclude financial industries and the specially treated companies, excluding the sample of overseas affiliates established in Hong Kong, Macau, and Taiwan, and overseas affiliates in tax havens such as Bermuda, the Cayman Islands, and the British Virgin Islands. Additionally, samples with missing variables are excluded. To eliminate the effect of extreme values, all continuous variables are Winsorized at the top and bottom 1%.

The data sources of this paper are as follows: ① OFDI and other financial statement information are obtained from China Stock Market & Accounting Research Database (CSMAR), ② data of green innovation are obtained from China Research Data Service Platform (CNRDS), ③ host countries’ information is taken from World Bank.

4.2. Regression Models and Variables

To analyze the effect of OFDI on green innovation, we regress variables using the following model:

$$greinv\_tgreinv=\beta 0+\beta 1 ISOFDI \left(OFDI\right)+\beta j Controls+Year FE+FIRM FE+\epsilon$$

(1)

Patents are a robust indicator of green innovations [51], and the literature generally holds that invention patents are more innovative than utility model patents. Following Cornaggia et al. [52] and Amore and Bennedsen [53], we use the total number of green invention patents and green utility model patents as the measurement index of the firms’ innovation quantity and the application of green invention patents as the measurement index of the firms’ innovation quality, where greinv_t is the logarithm of the total green patent application number, and greinv is the logarithm of the green invention patent application number.

Based on Bai et al. [22] and Dunning [54], we use two variables to measure firms’ outward foreign direct investment. ISOFDI equals 1 if the enterprise has overseas affiliates in the current year and 0 otherwise. OFDI is the number of overseas affiliates in the current year.

$$\begin{array}{c}greinv\_t\left(greinv\right)=\beta 0+\beta 1 OFDI+\beta 2HEC\left(HER,HSQ\right)+\beta 3 OFDI*HEC\left(HER,HSQ\right)\\ +\beta j Controls+Year FE+FIRM FE+\epsilon \end{array}$$

(2)

The measurement of moderating variables is developed based on Hernndez and Nieto [55] and Marano and Kostova [56]. Because each firm may invest in several host countries, we measure the host investment environments faced by firms as follows: The host investment environment includes the host economy (HEC), the host environmental regulation (HER), and the host national governance (HSQ). HEC takes 1 when all of an enterprise’s overseas affiliates are in host countries with higher GDP per capita than China and 0 otherwise. HER takes 1 when all of an enterprise’s overseas affiliates are in host countries with an environmental performance index greater than 60 and 0 otherwise. The country-level environmental performance is measured by Environmental Performance Index (EPI) [57]. HSQ takes 1 when all of an enterprise’s overseas affiliates are in host countries with a higher national governance index (WGI) than China and 0 otherwise.

Following the existing literature [22,24], variables that may affect firms’ green innovation are controlled as below: firm size (size), financial leverage (lev), profitability (Lroa), cash flow (cfo), Tobin’s q (tobinq), growth rate (growth), fixed assets (fixed), shareholding of institutional investors (IIholder), equity concentration (first), firm age (lnage), and nature of property rights (soe). The specific variables are defined as shown in

Table 1. Variable definition table.

Variable	Definition
greinv_t	Logarithm of number of green patent applications and 1
greinv	Logarithm of number of green invention patent applications and 1
ISOFDI	1 if the listed company has overseas affiliates, 0 otherwise
OFDI	Number of overseas affiliates
HEC	HEC takes 1 when all of an enterprise’s overseas affiliates are in host countries with higher GDP per capita than China and 0 otherwise.
HER	HER takes 1 when all of an enterprise’s overseas affiliates are in host countries with an environmental performance index greater than 60 and 0 otherwise.
HSQ	HSQ takes 1 when all of an enterprise’s overseas affiliates are in host countries with a higher national governance index (WGI) than China and 0 otherwise.
size	Logarithm of total assets
lev	Total liabilities/Total assets
Lroa	Previous year’s net profit/Previous year’s total assets
cfo	Net cash flow from operating activities/Total assets
tobinq	Market capitalization/Total assets
growth	(Current period operating revenue–Prior period operating revenue)/Prior period operating revenue
fixed	Fixed assets/Total assets
IIholder	Shareholding of institutional investors
first	Percentage of shareholding of the largest shareholder
lnage	Logarithm of firm age
soe	1 if it is a state-owned enterprise, 0 otherwise

.

5. Empirical Evidence

5.1. Descriptive Statistics and Correlation Analysis

Table 2. Descriptive statistics.

Panel A: Descriptive Statistics for the Full Sample
Variable	N	Mean	sd	Min	p50	Max
greinv_t	34,248	0.7896	1.1346	0.0000	0.0000	4.6151
greinv	34,248	0.5406	0.9289	0.0000	0.0000	4.1431
ISOFDI	34,248	0.2929	0.4551	0.0000	0.0000	1.0000
OFDI	34,248	1.1151	3.2438	0.0000	0.0000	31.0000
size	34,248	22.0318	1.3031	19.5041	21.8501	26.0706
lev	34,248	0.4336	0.2128	0.0503	0.4271	0.9526
Lroa	34,248	0.0421	0.0613	−0.2496	0.0412	0.2069
cfo	34,248	0.0478	0.0724	−0.1788	0.0474	0.2506
tobinq	34,248	2.0282	1.3257	0.8743	1.6011	8.8742
growth	34,248	0.1851	0.4711	−0.6428	0.1123	3.1335
fixed	34,248	0.2236	0.1685	0.0020	0.1887	0.7240
IIholder	34,248	45.4560	24.3191	0.3508	47.7300	90.9323
first	34,248	35.1359	14.9841	8.7700	33.1450	74.8200
lnage	34,248	2.7844	0.3805	1.6094	2.8332	3.4657
soe	34,248	0.4073	0.4913	0.0000	0.0000	1.0000
Panel B: Descriptive Statistics for the Subsample of OFDI (ISOFDI = 1)
Variable	N	Mean	sd	Min	p50	Max
greinv_t	10,030	1.2822	1.3626	0.0000	1.0986	4.6151
greinv	10,030	0.9327	1.1719	0.0000	0.6931	4.1431
OFDI	10,030	3.8076	5.0674	1.0000	2.0000	31.0000
size	10,030	22.5886	1.3780	19.5041	22.3757	26.0706
Lev	10,030	0.4511	0.1993	0.0503	0.4494	0.9526
Lroa	10,030	0.0444	0.0595	−0.2496	0.0417	0.2069
cfo	10,030	0.0529	0.0661	−0.1788	0.0509	0.2506
tobinq	10,030	1.9166	1.2085	0.8743	1.5305	8.8742
growth	10,030	0.1810	0.4059	−0.6428	0.1227	3.1335
fixed	10,030	0.2030	0.1452	0.0020	0.1764	0.7240
IIholder	10,030	45.9191	25.4410	0.3508	48.4555	90.9323
first	10,030	34.4547	15.1010	8.7700	32.5200	74.8200
lnage	10,030	2.8288	0.3650	1.6094	2.8904	3.4657
soe	10,030	0.3189	0.4661	0.0000	0.0000	1.0000

presents the descriptive statistics for the full sample (panel A) and the subsample of OFDI (Panel B). Panel A shows that 29.29% of the firm-year sample made OFDI, and the maximum number of overseas subsidiaries was 31, indicating a large variation in the firms’ OFDI. The mean value of total green innovation is 0.7896, and the maximum value is 4.6151. The mean value of green invention patents is 0.5406, and the maximum value is 4.1431, indicating that the overall level of enterprises’ green innovation is low and there is a gap among enterprises.

The mean value of OFDI in Panel B shows that, on average, each OFDI sample has 3.8076 subsidiaries, which is slightly larger than the median two, indicating that the level of OFDI varies widely. The mean value of total green innovation of OFDI enterprises is 1.2822, and the mean value of green invention patents is 0.9327. The rest of the control variables are largely similar to the existing literature.

Table 3. Spearman and Pearson correlation coefficients among firm-level variables.

	Greinv_t	Greinv	ISOFDI	OFDI	Size	Lev	Lroa	Cfo	Tobinq	Growth	Fixed	IIholder	First	Lnage	Soe
greinv_t	1	0.8905 ***	0.2647 ***	0.2791 ***	0.3424 ***	0.0970 ***	0.0147 ***	−0.0042	−0.1214 ***	0.0449 ***	−0.0460 ***	0.0353 ***	−0.0095 *	0.0870 ***	0.0105 *
greinv	0.9307 ***	1	0.2600 ***	0.2758 ***	0.3297 ***	0.0915 ***	0.0162 ***	−0.0014	−0.1033 ***	0.0396 ***	−0.0684 ***	0.0474 ***	−0.0127 **	0.0845 ***	0.0228 ***
ISOFDI	0.2794 ***	0.2717 ***	1	0.9818 ***	0.2655 ***	0.0593 ***	0.0202 ***	0.0432 ***	−0.0648 ***	0.0261 ***	−0.0546 ***	0.0153 ***	−0.0308 ***	0.0741 ***	−0.1157 ***
OFDI	0.2407 ***	0.2454 ***	0.5342 ***	1	0.2902 ***	0.0729 ***	0.0177 ***	0.0453 ***	−0.0754 ***	0.0263 ***	−0.0583 ***	0.0231 ***	−0.0302 ***	0.0805 ***	−0.1137 ***
size	0.4160 ***	0.4063 ***	0.2750 ***	0.2992 ***	1	0.4385 ***	−0.0794 ***	0.0639 ***	−0.4992 ***	0.0304 ***	0.0312 ***	0.3675 ***	0.1548 ***	0.2265 ***	0.2932 ***
lev	0.1269 ***	0.1167 ***	0.0529 ***	0.0960 ***	0.4210 ***	1	−0.4546 ***	−0.1492 ***	−0.3320 ***	0.0044	0.0815 ***	0.1981 ***	0.0385 ***	0.1214 ***	0.3065 ***
Lroa	0.0230 ***	0.0223 ***	0.0238 ***	0.0066	0.0057	−0.3833 ***	1	0.2783 ***	0.1933 ***	0.0964 ***	−0.1268 ***	0.0478 ***	0.1183 ***	−0.1362 ***	−0.2108 ***
cfo	0.0041	0.0086	0.0460 ***	0.0302 ***	0.0635 ***	−0.1574 ***	0.2366 ***	1	0.0919 ***	0.0595 ***	0.2519 ***	0.1304 ***	0.0899 ***	−0.0023	0.0095 *
tobinq	−0.1213 ***	−0.0993 ***	−0.0542 ***	−0.0621 ***	−0.3823 ***	−0.1930 ***	0.0590 ***	0.0762 ***	1	0.0613 ***	−0.0837 ***	−0.1554 ***	−0.1507 ***	−0.0524 ***	−0.2038 ***
growth	0.0108 **	0.0097 *	−0.0056	0.0081	0.0264 ***	0.0426 ***	−0.0132 **	0.0225 ***	0.0283 ***	1	−0.0638 ***	0.0342 ***	0.0156 ***	−0.1331 ***	−0.0687 ***
fixed	−0.0605 ***	−0.0758 ***	−0.0785 ***	−0.0530 ***	0.0855 ***	0.1217 ***	−0.1215 ***	0.2331 ***	−0.0974 ***	−0.0622 ***	1	0.1240 ***	0.0726 ***	−0.0636 ***	0.1899 ***
IIholder	0.0694 ***	0.0809 ***	0.0123 **	0.0531 ***	0.3937 ***	0.2065 ***	0.0535 ***	0.1150 ***	−0.0719 ***	0.0497 ***	0.1549 ***	1	0.5199 ***	−0.0201 ***	0.3992 ***
first	0.0142 ***	0.0109**	−0.0293 ***	−0.0105*	0.1998 ***	0.0407 ***	0.1229 ***	0.0839 ***	−0.1303 ***	0.0256 ***	0.0810 ***	0.5112 ***	1	−0.1308 ***	0.2158 ***
lnage	0.0829 ***	0.0829 ***	0.0751 ***	0.0676 ***	0.2000 ***	0.1383 ***	−0.1150 ***	0.0031	0.0325 ***	−0.0604 ***	−0.0347 ***	−0.0135 **	−0.1367 ***	1	0.0878 ***
soe	0.0358 ***	0.0484 ***	−0.1157 ***	−0.0484 ***	0.3033 ***	0.3025 ***	−0.1414 ***	0.0098*	−0.1325 ***	−0.0419 ***	0.2332 ***	0.4131 ***	0.2180 ***	0.0905 ***	1

Note: Spearman correlation coefficient is at the top right and Pearson correlation coefficient is at the bottom left; * p < 0.1, ** p < 0.05, *** p < 0.01.

presents that the correlation coefficients between OFDI and green innovation are all significantly positive at the 1% level. It shows that OFDI is conducive to promoting green innovation, which tentatively verifies Hypothesis 1.

5.2. Analysis of OFDI and Green Innovation

Table 4. Basic regression results.

	Greinv_t				Greinv
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
ISOFDI	0.2520 ***	0.1479 ***			0.2033 ***	0.1188 ***
	(16.14)	(9.48)			(15.01)	(8.80)
OFDI			0.0312 ***	0.0182 ***			0.0288 ***	0.0183 ***
			(14.12)	(8.33)			(14.03)	(9.08)
size		0.2785 ***		0.2794 ***		0.2270 ***		0.2243 ***
		(25.72)		(25.70)		(24.14)		(23.75)
lev		0.0311		0.0297		0.0138		0.0096
		(0.83)		(0.79)		(0.44)		(0.30)
Lroa		0.1841 **		0.1793 **		0.0744		0.0740
		(2.39)		(2.33)		(1.14)		(1.14)
cfo		−0.1138 *		−0.1045 *		−0.0908 *		−0.0841 *
		(−1.92)		(−1.76)		(−1.82)		(−1.69)
tobinq		0.0085 **		0.0084 **		0.0097 ***		0.0093 ***
		(2.11)		(2.08)		(2.81)		(2.72)
growth		−0.0065		−0.0069		−0.0093		−0.0092
		(−0.75)		(−0.80)		(−1.23)		(−1.23)
fixed		0.1660 ***		0.1770 ***		0.1085 ***		0.1176 ***
		(3.42)		(3.65)		(2.67)		(2.90)
IIholder		0.0003		0.0002		0.0006		0.0005
		(0.58)		(0.34)		(1.50)		(1.27)
first		−0.0021 ***		−0.0021 ***		−0.0020 ***		−0.0020 ***
		(−2.88)		(−2.91)		(−3.22)		(−3.20)
lnage		0.2835 ***		0.3017 ***		0.2365 ***		0.2504 ***
		(5.01)		(5.35)		(4.74)		(5.05)
soe		0.0777 ***		0.0798 ***		0.0735 ***		0.0767 ***
		(2.93)		(2.98)		(3.16)		(3.28)
_cons	0.7158 ***	−6.2188 ***	0.7548 ***	−6.2646 ***	0.4810 ***	−5.1871 ***	0.5085 ***	−5.1513 ***
	(122.99)	(−23.07)	(173.11)	(−23.28)	(96.59)	(−21.94)	(133.75)	(−21.87)
N	34,248	34,248	34,248	34,248	34,248	34,248	34,248	34,248
year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
firm	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
r2_a	0.6627	0.6764	0.6624	0.6763	0.6336	0.6471	0.6343	0.6476

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses.

shows the regression results for model (1). Coefficients of both independent variables are significantly positive at the 1% level, regardless of whether control variables are included or not. The results indicate that corporations’ OFDI is conducive to promoting green innovation, validating Hypothesis 1. The results for the other control variables indicate that the higher the firm’s previous earnings, firm’s size, fixed assets, growth, and firm age are, the stronger the green innovation capability is. However, the higher the firm’s equity concentration is, the lower the level of green innovation. The negative relationship between the cash flow and green innovation of the firm may be due to the fact that the firm does not invest enough in green innovation and does not actively engage in green innovation despite having a high level of cash flow.

5.3. Robustness Test

(1)

Propensity Score Matching Approach

To mitigate the concern that whether a firm makes OFDI may be influenced by other characteristics, we adopt the propensity score matching method (PSM) using firm size (size), the gearing ratio (lev), return on assets (Lroa), cash flow level (cfo), Tobin’s q (tobinq), firm growth (growth), fixed assets percentage (fixed), institutional shareholding (IIholder), equity concentration (first), firm age (age), and nature of ownership (soe) as the influence factors for propensity score matching, and the sample with OFDI and the sample without OFDI were matched by one-to-one repeated sampling of caliper nearest neighbors, and the differences between matched groups of samples are presented in

Table 5. Descriptive statistics of PSM sample.

	ISOFDI = 1	Mean	ISOFDI = 0	Mean	MeanDiff
greinv_t	6144	1.000	5973	0.885	0.115 ***
greinv	6144	0.695	5973	0.599	0.096 ***
size	6144	22.211	5973	22.248	0.037
lev	6144	0.428	5973	0.433	0.005
Lroa	6144	0.045	5973	0.045	0.001
cfo	6144	0.050	5973	0.050	0.000
tobinq	6144	1.985	5973	1.983	−0.002
growth	6144	0.181	5973	0.168	−0.013
fixed	6144	0.207	5973	0.205	−0.001
Iiholder	6144	44.397	5973	44.703	0.306
first	6144	34.744	5973	34.731	−0.013
lnage	6144	2.808	5973	2.808	0.000
soe	6144	0.336	5973	0.344	0.008

Note: *** p < 0.01.

. Except for corporations’ green innovation, none of the group differences in the influence factors are significant. The matched samples were tested again using model (1). The results in

Table 6. Results of propensity score matching approach.

	Greinv_t		Greinv
	(1)	(2)	(3)	(4)
ISOFDI	0.1206 ***		0.0921 ***
	(4.13)		(3.71)
OFDI		0.0125 ***		0.0116 ***
		(3.11)		(3.25)
size	0.3013 ***	0.2961 ***	0.2282 ***	0.2221 ***
	(11.61)	(11.27)	(10.03)	(9.65)
lev	0.0587	0.0512	0.0384	0.0294
	(0.67)	(0.58)	(0.50)	(0.38)
Lroa	0.2147	0.2266	0.1647	0.1763
	(1.22)	(1.29)	(1.05)	(1.13)
cfo	−0.0983	−0.1005	−0.0783	−0.0791
	(−0.76)	(−0.78)	(−0.70)	(−0.71)
tobinq	−0.0044	−0.0049	−0.0083	−0.0087
	(−0.48)	(−0.53)	(−1.05)	(−1.10)
growth	0.0197	0.0189	0.0080	0.0074
	(0.94)	(0.90)	(0.42)	(0.39)
fixed	0.2466 **	0.2680 **	0.1214	0.1346
	(2.03)	(2.21)	(1.12)	(1.25)
Iiholder	−0.0013	−0.0013	0.0003	0.0003
	(−1.24)	(−1.24)	(0.28)	(0.29)
first	−0.0005	−0.0005	−0.0011	−0.0011
	(−0.27)	(−0.32)	(−0.71)	(−0.74)
lnage	0.0485	0.0754	0.0239	0.0454
	(0.39)	(0.60)	(0.21)	(0.40)
soe	0.1143 *	0.1208 *	0.1171 *	0.1222 **
	(1.72)	(1.81)	(1.93)	(2.01)
_cons	−5.9943 ***	−5.9129 ***	−4.5826 ***	−4.4782 ***
	(−9.39)	(−9.21)	(−8.05)	(−7.84)
N	12,117	12,117	12,117	12,117
year	Yes	Yes	Yes	Yes
firm	Yes	Yes	Yes	Yes
r2_a	0.6696	0.6694	0.6309	0.6309

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses.

indicate that our inference remains unchanged with the control of the endogeneity problem.

(2)

Other Robustness Tests

To access the robustness of the baseline results, we make the following robustness tests: (1) we use an alternative independent variable (subcountry) to capture firms’ OFDI, which is measured by the number of host countries in which enterprises made OFDI in the current year; other variables in the model (1) remain unchanged. And the results are shown in

Table 7. Alternative Independent Variables.

	Greinv_t		Greinv
	(1)	(2)	(3)	(4)
subcountry	0.0753 ***	0.0493 ***	0.0697 ***	0.0490 ***
	(17.43)	(11.60)	(17.44)	(12.43)
size		0.2715 ***		0.2167 ***
		(25.04)		(23.06)
lev		0.0256		0.0057
		(0.69)		(0.18)
Lroa		0.1921 **		0.0865
		(2.50)		(1.34)
cfo		−0.1092 *		−0.0886 *
		(−1.85)		(−1.79)
tobinq		0.0081 **		0.0090 ***
		(2.01)		(2.64)
growth		−0.0062		−0.0085
		(−0.71)		(−1.14)
fixed		0.1763 ***		0.1169 ***
		(3.64)		(2.89)
Iiholder		0.0002		0.0005
		(0.41)		(1.36)
first		−0.0021 ***		−0.0020 ***
		(−2.89)		(−3.19)
lnage		0.2858 ***		0.2347 ***
		(5.09)		(4.76)
soe		0.0839 ***		0.0807 ***
		(3.15)		(3.46)
_cons	0.7334 ***	−6.0642 ***	0.4885 ***	−4.9569 ***
	(151.44)	(−22.57)	(114.64)	(−21.10)
N	34,248	34,248	34,248	34,248
year	Yes	Yes	Yes	Yes
firm	Yes	Yes	Yes	Yes
r2_a	0.6643	0.6773	0.6367	0.6490

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses.

. (2) We use alternative dependent variables, which are measured by the logarithm of enterprises’ total granted green patents (gragreinv_t) and granted green invention patents (gragreinv). gragreinv_t is an alternative measurement index of the firms’ innovation quantity using the total granted green patents in a year, and gragreinv is an alternative measurement index of the firms’ innovation quality using the granted green invention patents in a year. The results are shown in

Table 8. Alternative Dependent Variables.

	Gragreinv_t		Gragreinv
	(1)	(2)	(3)	(4)
ISOFDI	0.1439 ***		0.0511 ***
	(10.42)		(5.32)
OFDI		0.0174 ***		0.0118 ***
		(8.76)		(7.30)
size	0.2121 ***	0.2134 ***	0.1003 ***	0.0955 ***
	(21.84)	(21.86)	(15.33)	(14.55)
lev	0.0821 **	0.0810 **	−0.0049	−0.0100
	(2.46)	(2.42)	(−0.23)	(−0.47)
Lroa	−0.0555	−0.0605	−0.2158 ***	−0.2124 ***
	(−0.81)	(−0.88)	(−4.65)	(−4.60)
cfo	0.0326	0.0417	−0.0394	−0.0373
	(0.62)	(0.79)	(−1.14)	(−1.08)
tobinq	0.0053	0.0052	0.0009	0.0005
	(1.45)	(1.43)	(0.36)	(0.20)
growth	−0.0226 ***	−0.0230 ***	−0.0189 ***	−0.0185 ***
	(−2.99)	(−3.04)	(−3.82)	(−3.76)
fixed	0.1995 ***	0.2102 ***	0.0926 ***	0.0967 ***
	(4.72)	(4.96)	(3.37)	(3.53)
Iiholder	0.0001	0.0000	0.0002	0.0001
	(0.30)	(0.03)	(0.69)	(0.54)
first	−0.0011 *	−0.0011 *	−0.0008 *	−0.0007
	(−1.72)	(−1.75)	(−1.73)	(−1.63)
lnage	0.3394 ***	0.3573 ***	0.2472 ***	0.2525 ***
	(6.54)	(6.90)	(6.55)	(6.70)
soe	0.0259	0.0278	0.0019	0.0049
	(1.11)	(1.18)	(0.12)	(0.31)
_cons	−5.0821 ***	−5.1339 ***	−2.6442 ***	−2.5520 ***
	(−21.07)	(−21.29)	(−15.90)	(−15.31)
N	34,248	34,248	34,248	34,248
year	Yes	Yes	Yes	Yes
firm	Yes	Yes	Yes	Yes
r2_a	0.6747	0.6745	0.5804	0.5816

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses.

. (3) To exclude the potential bias of the sample without green patents, we make a sub-sample (green innovation > 0) test, variables in the model (1) remain unchanged, and the results are shown in

Table 9. Subsample (Green Innovation > 0).

	Greinv_t		Greinv
	(1)	(2)	(3)	(4)
ISOFDI	0.0720 ***		0.0693 ***
	(3.20)		(2.92)
OFDI		0.0086 ***		0.0083 ***
		(3.02)		(2.80)
size	0.3446 ***	0.3442 ***	0.3239 ***	0.3235 ***
	(16.64)	(16.52)	(14.90)	(14.77)
lev	0.0519	0.0515	0.0072	0.0069
	(0.68)	(0.67)	(0.09)	(0.09)
Lroa	0.2878 *	0.2928 *	0.1088	0.1135
	(1.86)	(1.89)	(0.66)	(0.69)
cfo	−0.1488	−0.1454	−0.0600	−0.0566
	(−1.31)	(−1.28)	(−0.50)	(−0.47)
tobinq	0.0074	0.0071	0.0057	0.0054
	(0.91)	(0.87)	(0.68)	(0.64)
growth	−0.0244	−0.0250	−0.0245	−0.0251
	(−1.59)	(−1.64)	(−1.47)	(−1.50)
fixed	0.2663 ***	0.2781 ***	0.1713 *	0.1826 *
	(2.73)	(2.86)	(1.69)	(1.80)
Iiholder	0.0008	0.0007	0.0014	0.0014
	(0.93)	(0.81)	(1.59)	(1.48)
first	−0.0011	−0.0010	−0.0013	−0.0012
	(−0.85)	(−0.79)	(−0.94)	(−0.88)
lnage	−0.2759 ***	−0.2702 ***	−0.1961 *	−0.1906 *
	(−2.76)	(−2.71)	(−1.86)	(−1.82)
soe	−0.0220	−0.0176	0.0172	0.0214
	(−0.39)	(−0.31)	(0.29)	(0.36)
_cons	−5.2595 ***	−5.2529 ***	−5.5846 ***	−5.5794 ***
	(−10.19)	(−10.16)	(−10.36)	(−10.34)
N	14,839	14,839	14,839	14,839
year	Yes	Yes	Yes	Yes
firm	Yes	Yes	Yes	Yes
r2_a	0.6513	0.6513	0.6196	0.6196

Note: * p < 0.1, *** p < 0.01; t-values for robust standard error are in parentheses.

. (4) We provide a replacement to the tobit model. The variables in the tobit model remain unchanged as well. And the results are shown in

Table 10. The Tobit Model.

	Greinv_t		Greinv
	(1)	(2)	(3)	(4)
ISOFDI	0.6047 ***		0.6005 ***
	(23.49)		(22.94)
OFDI		0.0395 ***		0.0406 ***
		(11.20)		(12.03)
size	0.6381 ***	0.6757 ***	0.6485 ***	0.6841 ***
	(48.99)	(50.84)	(48.43)	(50.15)
lev	−0.1249 *	−0.1218 *	−0.3207 ***	−0.3182 ***
	(−1.73)	(−1.68)	(−4.31)	(−4.26)
Lroa	0.9640 ***	0.9148 ***	0.6150 ***	0.5705 **
	(4.25)	(4.02)	(2.64)	(2.45)
cfo	−0.5920 ***	−0.4623 **	−0.4588 **	−0.3299 *
	(−3.23)	(−2.51)	(−2.44)	(−1.75)
tobinq	0.0261 **	0.0313 ***	0.0616 ***	0.0667 ***
	(2.44)	(2.92)	(5.70)	(6.16)
growth	0.0569 **	0.0461 *	0.0533 **	0.0424
	(2.22)	(1.79)	(2.02)	(1.60)
fixed	−0.5356 ***	−0.5933 ***	−0.9094 ***	−0.9629 ***
	(−7.08)	(−7.78)	(−11.71)	(−12.33)
Iiholder	−0.0052 ***	−0.0052 ***	−0.0041 ***	−0.0042 ***
	(−8.26)	(−8.26)	(−6.33)	(−6.36)
first	−0.0054 ***	−0.0058 ***	−0.0066 ***	−0.0070 ***
	(−5.80)	(−6.27)	(−6.88)	(−7.30)
lnage	−0.7036 ***	−0.7330 ***	−0.6591 ***	−0.6877 ***
	(−18.89)	(−19.55)	(−17.22)	(−17.88)
soe	0.1440 ***	0.0803 ***	0.2084 ***	0.1451 ***
	(4.87)	(2.70)	(6.81)	(4.74)
_cons	−13.5385 ***	−14.1566 ***	−14.1886 ***	−14.7705 ***
	(−47.03)	(−48.11)	(−47.41)	(−48.32)
N	34,248	34,248	34,248	34,248
year	Yes	Yes	Yes	Yes
firm	No	No	No	No
Pseudo R2	0.0975	0.0930	0.1052	0.1004

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses.

. We find the coefficients of firms’ OFDI are significantly positive, verifying that our results are robust.

5.4. Moderating Role of Host Investment Environments

Based on the fact that OFDI firms face different host economies, environmental regulations, and national governance, we further explore the mechanisms through which OFDI affects green innovation in the above three dimensions using the subsample of enterprises taking OFDI.

(1)

Host Economic Development

The effect of OFDI on corporate green innovation is initially affected by the level of host economic development.

Table 11. Moderating Role of Host Economic Development.

	Greinv_t	Greinv
	(1)	(2)
OFDI	0.0039	0.0048
	(1.22)	(1.63)
OFDI*HEC	0.0131 ***	0.0100 **
	(2.80)	(2.16)
HEC	−0.0741 **	−0.0643 *
	(−1.97)	(−1.84)
size	0.3522 ***	0.2705 ***
	(12.56)	(10.58)
lev	0.0761	0.0264
	(0.78)	(0.31)
Lroa	0.4905 ***	0.4069 **
	(2.67)	(2.45)
cfo	−0.2067	−0.1985
	(−1.46)	(−1.58)
tobinq	0.0022	−0.0065
	(0.22)	(−0.74)
growth	−0.0226	−0.0347 *
	(−1.11)	(−1.87)
fixed	0.1060	−0.0045
	(0.69)	(−0.03)
Iiholder	0.0008	0.0027 ***
	(0.74)	(2.68)
first	0.0027	0.0007
	(1.49)	(0.42)
lnage	0.0083	0.0427
	(0.06)	(0.36)
soe	−0.0445	−0.0154
	(−0.61)	(−0.22)
_cons	−6.8686 ***	−5.4337 ***
	(−9.62)	(−8.22)
N	10,030	10,030
year	Yes	Yes
firm	Yes	Yes
r2_a	0.7687	0.7468

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses.

presents the moderating role of host economic development. Columns (1) and (2) show that host economic development significantly positively moderates the relationship between OFDI and green innovation. The results indicate that when Chinese OFDI firms are subjected to greater environmental protection pressure from higher economy countries, firms will adopt more positive green strategies and make more social commitments to actively assume environmental responsibility. It helps firms reduce their outside uncertainty and risks and is beneficial for mitigating the liability of foreignness and cultivating competitiveness, validating Hypothesis 2a.

(2)

Host Environmental Regulations

Host environmental regulations have a direct effect on the relationship between OFDI and green innovation.

Table 12. Moderating Role of Host Environmental Regulations.

	Greinv_t	Greinv
	(1)	(2)
OFDI	0.0046	0.0046
	(1.45)	(1.58)
OFDI*HER	0.0095 **	0.0112 ***
	(2.17)	(2.68)
HER	−0.0472	−0.0556 **
	(−1.60)	(−2.08)
size	0.3533 ***	0.2714 ***
	(12.59)	(10.62)
lev	0.0862	0.0323
	(0.89)	(0.37)
Lroa	0.4820 ***	0.4002 **
	(2.62)	(2.42)
cfo	−0.2085	−0.2006
	(−1.47)	(−1.60)
tobinq	0.0020	−0.0066
	(0.21)	(−0.75)
growth	−0.0225	−0.0346 *
	(−1.10)	(−1.86)
fixed	0.1083	0.0009
	(0.71)	(0.01)
Iiholder	0.0008	0.0027 ***
	(0.72)	(2.66)
first	0.0027	0.0007
	(1.48)	(0.40)
lnage	0.0075	0.0390
	(0.06)	(0.32)
soe	−0.0447	−0.0147
	(−0.62)	(−0.22)
_cons	−6.9100 ***	−5.4546 ***
	(−9.66)	(−8.25)
N	10,030	10,030
year	Yes	Yes
firm	Yes	Yes
r2_a	0.7686	0.7469

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses.

presents the moderating role of host environmental regulations. And columns (1) and (2) show that the coefficient of OFDI*HER is positively significant at the 5% (1%) level, implying that host environmental regulations have a positive moderating effect on the relationship between OFDI and green innovation. The results indicate that the host environmental regulation has heavy pressure on the green innovation of OFDI firms. When the host environmental regulation is stricter, Chinese OFDI firms will face higher pressure and environmental management costs, which would affect the propensity of firms to engage in green innovation and force them to actively optimize cleaner production technologies to reduce the compliance costs of production and operation in the host country. Therefore, based on the risk management motive and legitimacy motive, OFDI firms will engage more in green innovation under strict host environmental regulations, validating Hypothesis 2b.

(3)

Host National Governance

Host national governance affects enterprises’ social and environmental performance a lot. Enterprises’ OFDI in countries with higher governance scores not only implies better monitoring but also stricter laws and increased exposure to lawsuits, which have stricter restraining and governing effects on firms’ pollution behaviors.

Table 13. Moderating Role of Host National Governance.

	Greinv_t	Greinv
	(1)	(2)
OFDI	0.0036	0.0039
	(1.10)	(1.29)
OFDI*HSQ	0.0113 **	0.0099 **
	(2.24)	(2.08)
HSQ	−0.0906 **	−0.0989 ***
	(−2.34)	(−2.76)
size	0.3513 ***	0.2692 ***
	(12.53)	(10.55)
lev	0.0753	0.0235
	(0.77)	(0.27)
Lroa	0.4884 ***	0.4073 **
	(2.65)	(2.46)
cfo	−0.2102	−0.2024
	(−1.49)	(−1.61)
tobinq	0.0021	−0.0065
	(0.22)	(−0.74)
growth	−0.0220	−0.0340 *
	(−1.08)	(−1.83)
fixed	0.1084	−0.0003
	(0.71)	(−0.00)
Iiholder	0.0008	0.0027 ***
	(0.73)	(2.68)
first	0.0028	0.0008
	(1.52)	(0.47)
lnage	0.0111	0.0443
	(0.09)	(0.37)
soe	−0.0432	−0.0142
	(−0.60)	(−0.21)
_cons	−6.8447 ***	−5.3878 ***
	(−9.60)	(−8.17)
N	10,030	10,030
year	Yes	Yes
firm	Yes	Yes
r2_a	0.7686	0.7469

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses.

presents the moderating role of host governance. As shown in columns (1) and (2), the coefficient of OFDI*HSQ is positively significant at the 5% level, indicating that host national governance has a positive moderating effect on the relationship between OFDI and green innovation. The results suggest that OFDI promotes more green innovation under stricter governance, implying that to meet the host institutional regulatory requirements, OFDI firms behave a better green innovation performance to seek legitimacy, validating Hypothesis 2c.

6. Further Analysis

We have demonstrated how the host investment environments act on the relationship between OFDI and green innovation. In this section, we will further examine whether firms’, industries’, and domestic regions’ characteristics influence the relationship between OFDI and green innovation.

6.1. Firm-Level Characteristics

(1)

Nature of Property Rights Perspective

Property rights make state-owned enterprises have a natural difference from private enterprises in their perception of environmental and social responsibility, which makes state-owned enterprises more responsible [58] for environmental protection. Thus, state-owned enterprises making OFDI engage more in green innovation.

As shown in

Table 14. Nature of Property Rights Perspective.

	Greinv_t		Greinv
	State-Owned Enterprises	Private Enterprises	State-Owned Enterprises	Private Enterprises
	(1)	(2)	(3)	(4)
OFDI	0.0212 ***	0.0024	0.0197 ***	0.0029
	(4.44)	(0.67)	(4.46)	(0.88)
size	0.3674 ***	0.4381 ***	0.2683 ***	0.3743 ***
	(7.38)	(12.39)	(5.68)	(11.90)
lev	0.0120	0.0708	0.1401	−0.0493
	(0.06)	(0.63)	(0.78)	(−0.51)
Lroa	0.0049	0.4640 **	0.1937	0.2931
	(0.01)	(2.20)	(0.55)	(1.62)
cfo	0.1564	−0.3004 *	0.0145	−0.1911
	(0.66)	(−1.74)	(0.06)	(−1.28)
tobinq	−0.0068	0.0020	−0.0130	−0.0047
	(−0.29)	(0.18)	(−0.61)	(−0.49)
growth	−0.0376	−0.0283	−0.0404	−0.0414 *
	(−1.08)	(−1.15)	(−1.28)	(−1.83)
fixed	−0.5283 **	0.2661	−0.5899 **	0.1932
	(−2.03)	(1.49)	(−2.46)	(1.26)
IIholder	0.0013	−0.0000	0.0023	0.0019 *
	(0.56)	(−0.03)	(1.01)	(1.70)
first	−0.0032	0.0048 **	−0.0043	0.0018
	(−1.08)	(2.06)	(−1.52)	(0.89)
lnage	0.0194	0.1492	0.1820	0.1516
	(0.10)	(0.83)	(0.99)	(0.99)
_cons	−6.9134 ***	−9.2864 ***	−5.5081 ***	−8.1068 ***
	(−5.22)	(−10.43)	(−4.38)	(−10.01)
N	3199	6831	3199	6831
year	Yes	Yes	Yes	Yes
firm	Yes	Yes	Yes	Yes
r2_a	0.8203	0.7289	0.7985	0.7063
Coefficient difference	0.019(0.000) ***		0.017(0.000) ***

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses. Differences in coefficients between groups are presented as coefficient difference values; p-values are in parentheses.

, OFDI is significantly positively associated with green innovation only among state-owned enterprises. This result indicates that, compared to private enterprises, state-owned enterprises attach more importance to corporate reputation and green image and have a stronger sense of social responsibility, thus the promotion of green innovation by OFDI is stronger. It further validates the reputational motivation of firms to engage in green innovation.

(2)

Overseas Investment Experience Perspective

Prior research shows a positive effect of prior experience on firms’ likelihood of OFDI [59,60] and suggests that less experienced firms face greater challenges from international expansion and, therefore, benefit largely from CSR reputation in reducing the challenges of internationalization [61].

Table 15. Overseas Investment Experience Perspective.

	Greinv_t		Greinv
	Short Overseas Investment Experience	Long Overseas Investment Experience	Short Overseas Investment Experience	Long Overseas Investment Experience
	(1)	(2)	(3)	(4)
OFDI	0.0176 **	0.0013	0.0168 ***	0.0018
	(2.47)	(0.32)	(2.66)	(0.49)
size	0.2464 ***	0.3489 ***	0.1338 ***	0.2632 ***
	(4.38)	(6.47)	(2.71)	(5.50)
lev	−0.1428	0.1671	−0.1036	0.1569
	(−0.96)	(0.88)	(−0.81)	(0.96)
Lroa	0.3974	0.7212 **	0.4822 **	0.4981 **
	(1.51)	(2.57)	(2.00)	(2.09)
cfo	−0.2134	−0.0015	−0.2551 *	0.1272
	(−1.20)	(−0.01)	(−1.70)	(0.66)
tobinq	−0.0198	0.0203	−0.0144	−0.0004
	(−1.52)	(1.18)	(−1.26)	(−0.03)
growth	0.0111	−0.0025	−0.0065	0.0178
	(0.48)	(−0.06)	(−0.31)	(0.51)
fixed	0.6255 ***	−0.3044	0.2407	−0.2852
	(2.85)	(−1.15)	(1.26)	(−1.24)
IIholder	−0.0017	0.0016	0.0012	0.0035 *
	(−1.05)	(0.74)	(0.85)	(1.77)
first	0.0032	−0.0040	0.0015	−0.0050 *
	(0.94)	(−1.27)	(0.52)	(−1.83)
lnage	−0.4149	−0.3560	−0.4833 *	−0.2655
	(−1.32)	(−1.23)	(−1.78)	(−1.02)
soe	−0.1565	0.1295	−0.0021	0.0968
	(−1.10)	(1.21)	(−0.02)	(1.03)
_cons	−3.3597 **	−5.5056 ***	−1.0497	−4.1879 ***
	(−2.30)	(−3.70)	(−0.82)	(−3.12)
N	5692	4338	5692	4338
year	Yes	Yes	Yes	Yes
firm	Yes	Yes	Yes	Yes
r2_a	0.7855	0.8153	0.7641	0.8096
Coefficient difference	0.016 (0.000) ***		0.015 (0.000) ***

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses. Differences in coefficients between groups are presented as coefficient difference values; p-values are in parentheses.

shows that OFDI is significantly positively related to green innovation among firms with short overseas investment experience, indicating that less experienced firms are more reliant on their green investment in counteracting overseas investment risks and uncertainties, building a green reputation to attract overseas consumers, and overcoming the legitimacy deficit. In contrast, firms with extensive overseas experience may already establish their reputation and gain an understanding of host regulations, values, and norms [60,62,63]; thus, the incentive to manage risk through green innovation is relatively weaker.

6.2. Industry-Level Characteristics

Due to the operational characteristic, heavily polluting industries always face greater challenges of environmental regulations and the risk of environmental violations in the host country. Following Ji and Su [64], based on the Ministry of Environmental Protection’s “List of Listed Companies’ Environmental Verification Industry Classification and Management” (Letter from the Ministry of Environmental Protection [2008] No. 373) and the Ministry of Environmental Protection’s “Guidance on Environmental Information Disclosure for Listed Companies” (Draft for Comments)(2010), we classify enterprises in the following industries as the heavy pollution industries to explore whether the effect of OFDI on green innovation varies in heavy pollution industries or not, e.g., thermal power, iron and steel, cement, electrolytic aluminum, coal, metallurgy, chemical, petrochemical, building materials, paper making, brewing, pharmaceutical, fermentation, textile, tannery, and mining industries.

Columns (1) and (2) in

Table 16. Heavy Pollution Industry Perspective.

	Greinv_t		Greinv
	Heavy Pollution Industry	Non-Heavy Pollution Industry	Heavy Pollution Industry	Non-Heavy Pollution Industry
	(1)	(2)	(3)	(4)
OFDI	0.0089	0.0051	0.0104 *	0.0056 *
	(1.38)	(1.55)	(1.74)	(1.82)
size	0.3349 ***	0.3604 ***	0.2242 ***	0.2875 ***
	(5.51)	(10.76)	(4.12)	(9.38)
lev	−0.1000	0.1398	−0.1264	0.0428
	(−0.54)	(1.19)	(−0.78)	(0.41)
Lroa	1.2367 ***	0.3295	1.1186 ***	0.2777
	(3.25)	(1.56)	(3.26)	(1.45)
cfo	0.0286	−0.2911 *	0.1039	−0.2937 **
	(0.10)	(−1.79)	(0.41)	(−2.01)
tobinq	0.0277	−0.0063	0.0141	−0.0151
	(1.42)	(−0.53)	(0.85)	(−1.42)
growth	0.0176	−0.0214	0.0649 *	−0.0441 **
	(0.44)	(−0.89)	(1.70)	(−2.01)
fixed	0.2791	−0.0343	0.2178	−0.0798
	(1.04)	(−0.18)	(0.92)	(−0.47)
IIholder	−0.0014	0.0005	0.0005	0.0025 **
	(−0.64)	(0.37)	(0.22)	(2.13)
first	0.0019	0.0037 *	0.0000	0.0018
	(0.50)	(1.72)	(0.01)	(0.91)
lnage	−1.0305 ***	0.2350	−0.8652 ***	0.2436 *
	(−3.39)	(1.57)	(−3.06)	(1.78)
soe	−0.1140	−0.0513	−0.1205	−0.0216
	(−0.88)	(−0.60)	(−0.98)	(−0.27)
_cons	−3.7864 **	−7.6152 ***	−2.0275	−6.3232 ***
	(−2.48)	(−9.02)	(−1.46)	(−8.10)
N	2580	7450	2580	7450
year	Yes	Yes	Yes	Yes
firm	Yes	Yes	Yes	Yes
r2_a	0.7420	0.7751	0.7214	0.7527
Coefficient difference	\		0.005 (0.002) ***

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses. Differences in coefficients between groups are presented as coefficient difference values; p-values are in parentheses.

show that the coefficients of OFDI and total green innovation are positive but insignificant in both the heavy pollution and non-heavy pollution industries. While in columns (3) and (4), they are significantly positive at the 10% level, and the coefficients of groups are significantly positively different at the 1% level, indicating that OFDI has a significant promotion effect on green invention innovation in both heavy and non-heavy pollution industry firms, and this promotion effect is stronger in heavy pollution industry firms. Due to the greater pressure of environmental regulations in the host country, enterprises in heavy pollution industries are more motivated to make green innovations to seek institutional legitimacy and send a signal to host countries that they are actively fulfilling their environmental responsibilities.

6.3. Domestic Region-Level Characteristics

The intensity of environmental regulations in the home country also impacts the relationship between corporate OFDI and green innovation. Following Zhang [65], considering the heterogeneity of the industrial structure of each region, we use the ratio of industrial pollution control investment per unit of industrial added value and industrial added value per unit GDP in the registering province to measure domestic local environmental regulation intensity, with larger values indicating stricter environmental regulation in the home country. The environmental regulation of domestic regions is grouped according to the median, and a group not less than the median is defined as strong environmental regulation, while the opposite is weak environmental regulation. The results in

Table 17. Domestic Environmental Regulation Perspective.

	Greinv_t		Greinv
	Strong Environmental Regulation	Weak Environmental Regulation	Strong Environmental Regulation	Weak Environmental Regulation
	(1)	(2)	(3)	(4)
OFDI	0.0132 ***	0.0086 *	0.0125 ***	0.0087 **
	(2.97)	(1.95)	(2.93)	(2.17)
size	0.2909 ***	0.4510 ***	0.2137 ***	0.3534 ***
	(5.97)	(9.93)	(4.92)	(8.64)
lev	0.0317	0.2101	0.0425	0.0538
	(0.21)	(1.34)	(0.32)	(0.39)
Lroa	0.8145 ***	0.3687	0.6147 **	0.4009 *
	(2.67)	(1.37)	(2.19)	(1.74)
cfo	−0.0830	−0.2422	−0.1887	−0.1385
	(−0.40)	(−1.07)	(−1.02)	(−0.69)
tobinq	−0.0163	0.0062	−0.0208	−0.0069
	(−1.05)	(0.40)	(−1.51)	(−0.50)
growth	0.0132	−0.0486	−0.0038	−0.0486 *
	(0.44)	(−1.57)	(−0.14)	(−1.81)
fixed	0.0690	−0.0408	0.0263	−0.2999
	(0.29)	(−0.17)	(0.13)	(−1.47)
IIholder	−0.0005	0.0022	0.0017	0.0034 **
	(−0.30)	(1.18)	(1.12)	(2.13)
first	0.0056 **	−0.0012	0.0033	−0.0028
	(2.05)	(−0.41)	(1.27)	(−1.00)
lnage	0.3939*	−0.0403	0.4611 **	−0.0693
	(1.79)	(−0.22)	(2.26)	(−0.40)
soe	−0.1923	0.0809	−0.1410	0.1402
	(−1.59)	(0.68)	(−1.23)	(1.36)
_cons	−6.6730 ***	−8.8966 ***	−5.4595 ***	−6.8564 ***
	(−5.29)	(−8.08)	(−4.73)	(−6.74)
N	5015	5015	5015	5015
year	Yes	Yes	Yes	Yes
firm	Yes	Yes	Yes	Yes
r2_a	0.7694	0.7778	0.7456	0.7622
Coefficient difference	0.005 (0.080) *		0.004 (0.098) *

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values for robust standard error are in parentheses. Differences in coefficients between groups are presented as coefficient difference values; p-values are in parentheses.

present that the relationship between OFDI and green innovation is significantly positively correlated at the 1% and 10% (1% and 5%) levels. Respectively, the coefficients of differences between groups are positively significant at the 10% level, indicating that environmental regulations in home country regions also strengthen green innovations of OFDI firms.

7. Discussion and Implications

Despite a large literature on internationalization, few studies have explored the mechanisms by which firms’ OFDI affects green innovation. Based on the evidence in China, this paper empirically examines the effect of OFDI on green innovation and the moderating role of host investment environments. The study finds that OFDI significantly promotes green innovation, not only for the sake of gaining legitimacy, but also to mitigate liabilities of foreignness and cultivate competitive advantages. The host economic development, environmental regulations, and national governance significantly positively moderate the promotion effect of OFDI on green innovation. It implies that the heterogeneity of host investment environments effectively affects OFDI enterprises’ motivation to make green innovation and, thus, has a differential effect. Further analysis shows that OFDI promotes green innovation more significantly in state-owned and less-overseas-experienced enterprises, heavy pollution industries, and domestic areas with strict environmental regulations.

In summary, the study makes three important contributions to the literature. First, our findings enrich the study of economic consequences related to enterprises’ OFDI from the perspective of green innovation, which is of great practical significance for improving the global environmental governance system. Second, it deepens the knowledge and understanding of the environmental behavioral decisions of OFDI enterprises and expands the research field of factors influencing corporations’ green innovation. Multinational enterprises are important parts of the global environmental governance system. Exploring the effect of international business practices on green innovation provides a new perspective for studying firms’ innovation behavior and green innovation motives and helps to solve environmental problems in a global context. Third, we analyze firms’ different green innovation strategies to cope with different overseas risks from the perspective of host investment environments, which enriches the relevant studies on the mechanism of the effect of corporate OFDI on green innovation.

There are also important practical implications for governments and enterprises around the world. On the one hand, governments should actively promote enterprises to engage in international exchanges and investments. Governments need to pay more attention to the investment environment of each country and make risk judgments. In addition, governments could attempt to cultivate and nurture some high-quality enterprises with international competitiveness and guide them to upstream host countries that can positively promote firms’ innovation. On the other hand, enterprises should implement strategic corporate social responsibility, including having a global vision, practicing green and sustainable development strategies, and proactively improving environmental performance. Thus, they can cultivate green competitiveness, absorb the advanced clean technology of the host country, and transform the host environmental pressure into a driver to achieve sustainable development, which ultimately leads to common benefits for enterprises and the international community. In addition, enterprises should fully study the market environment of the host country and choose a better investment environment. Enterprises engaging in OFDI around the world, especially those from emerging markets, should take advantage of the host countries’ environment to improve their environmental innovation capabilities and ecological practices for better international competitiveness.