Carbon policy and firms’ OFDI in the digital era: Insights from China

Abstract

Objective

Creating a low-carbon economy has become an essential goal in fostering green development. In the context of achieving sustainability, in 2014, China implemented a carbon emission trading system (ETS) to conserve energy and reduce emissions. This study examined the effect of the carbon ETS on the outward foreign direct investment (OFDI) of Chinese listed firms.

Methods

This study employed a difference-in-differences approach using the data of Chinese listed firms from 2010 to 2017. The data were sourced from the CSMAR database. Firms located in the trial areas of the carbon ETS were considered the treatment group.

Results

The carbon ETS had a significant positive impact on firms’ OFDI, even after several robustness checks. The carbon ETS positively influenced the OFDI of private firms, firms not belonging to polluting industries, and highly profitable firms but not the OFDI of state-owned firms, firms belonging to polluting industries, and less profitable firms. Technological innovation mediated the impact of the carbon ETS on firms’ OFDI but increased production costs did not. Digitalization enhanced the positive effects of the carbon ETS on firms’ OFDI.

Conclusion

The results present vital implications for deepening two-way investments across economies in the era of digitalization. Developing governments should actively promote carbon emission trading policies and digitalization to help enterprises realize green transformation and improve their competitiveness in “going global.”

Keywords

Carbon emission trading policies outward foreign direct investment technological innovation production cost corporate digitalization

Introduction

Greenhouse gas emissions are driving temperatures worldwide and challenging society's sustainable development.^1,2 A low-carbon economy represents a pivotal area in industrial growth,² and creating such an economy is vital for fostering green development.³ China, which was a leading carbon emitter, experienced significant pressure to cut its carbon emissions. As a result, at the 75th United Nations General Assembly, China announced its goal to achieve carbon neutrality by 2060.

Building a low-carbon economy requires shifting from high-carbon to low-carbon economic development, which cannot be achieved without government regulation. Consequently, in 2014, China implemented the carbon emission trading system (ETS) to reduce carbon emissions and conserve energy. The carbon ETS provides momentum to optimize energy consumption and use clean energy. It provides flexibility and incentivizes polluters to choose emission-reducing routes, which eliminates inefficient production methods and discourages overproduction. Furthermore, the system stimulates industrial restructuring and accelerates the shift toward advanced production practices.

Outward foreign direct investment (OFDI) is a key component of China's opening-up strategy and is expanding rapidly. The implementation of carbon emission trading policies are expected to make firms adjust their production methods and improve their competitiveness. Therefore, this study examined whether an improvement in firms’ competitiveness brought about by low-carbon operations can increase their OFDI. It also explored the mechanisms underlying this relationship.

Existing studies have investigated how carbon emission trading policies affect enterprises’ green performance, including their green production efficiency^4,5 and innovations in reducing carbon emissions.^6,7 Liu & Sun⁸ and Feng et al.⁹ found that carbon emission trading policies have geographically heterogenous effects in China. Extant studies have also explored the connection between environmental policies and OFDI. They have found positive,^10,11 negative,¹² as well as an inverted U-shaped relationship.¹³ These variations may be due to different samples different policy backgrounds. This study focused on developing countries, represented by China, and explored a relatively accurate causal relationship between carbon emission trading policies and firms’ OFDI.

This study contributes to the literature in several ways. First, it provides novel empirical insights into the effects of carbon emission trading policies on the OFDI of firms in developing countries. By offering empirical evidence that carbon emission trading policies positively influence firms’ OFDI, this study adds to the limited research on the economic consequences of carbon emission trading policies and provides practical implications for developing economies considering the establishment of carbon ETSs.

Second, this study explores the mechanisms through which carbon policies, such as carbon emission trading policies, affect firms’ OFDI. By examining the roles of technological innovation and production costs, this study reveals that technological innovation plays a pivotal role in facilitating firms’ OFDI, whereas the influence of production costs is relatively less significant. These mechanistic insights offer valuable information for policymakers and firms aiming to increase OFDI.

Third, this study introduces a novel dimension to existing research by exploring the moderation of digitalization on the impact of carbon emission trading policies on firms’ OFDI. It empirically demonstrates that digitalization amplifies the positive effects of carbon emission trading policies, suggesting a synergistic effect of digitalization and low-carbon operations.

The remainder of this article is organized as follows. The hypotheses development section explains the development of hypotheses. The methods section describes the study's methodology. The results and discussion section presents results concerning the relationship between carbon emission trading policies and firms’ OFDI, underlying mechanisms, and heterogenous effects. The moderation analysis section analyzes the moderation of digitalization in the impact of carbon emission trading policies on firms’ OFDI. The conclusion section concludes the article.

Hypotheses development

Impact of carbon emission trading policies on firms’ OFDI

As shown in Figure 1, we believe that carbon emission trading policies affect firms’ OFDI through two mechanisms.

Figure 1.

Mechanisms through which carbon trading policies may affect firms’ OFDI.

The first mechanism is technological innovation. Porter and Van der Linde¹⁴ reported that environmental regulations stimulate corporate innovation. When faced with strict environmental policies, firms may optimize the production process¹⁵ or engage in technological innovation¹⁶ to lower emissions and enhance production. Firms’ technological innovations can be classified as production innovations and emission-reducing innovations. Production innovations increase firms’ productivity.¹⁷ With increased productivity, firms’ benefits of increased production efficiency offset environmental compliance costs.¹⁸ Meanwhile, emission-reducing innovations tend to reduce firms’ environmental compliance costs because firms employ cleaner technologies.

Engaging in production innovation may ameliorate firms’ operational efficiency and enhance their productivity and profitability,¹⁷ which encourages OFDI. Engaging in emission-reducing innovations may align firms’ practices with sustainable development goals.¹⁹ This adoption of cleaner technologies may improve firms’ competitiveness in the global market, enhance their capability for OFDI, and increase the number of potential host economies,²⁰ thus encouraging OFDI.²¹ Therefore, an increase in firms’ technological innovations may boost their OFDI.²²

The second mechanism is production cost. Strict environmental policies may place an additional financial burden on firms to cover pollution control costs in production processes, which may increase overall production costs.²³ Strict environmental policies can drive firms to revamp their production processes. To comply with these policies, firms often need to purchase new equipment. This investment, in turn, leads to an increase in overall production costs.²⁴

An increase in overall production costs may push firms to relocate production to economies with more lenient environmental policies and lower production expenses,²⁵ thus driving OFDI. According to Chung,²⁶ firms can mitigate the adverse effects of stringent domestic environmental policies by producing overseas. Consequently, the imposition of environmental policies may incentivize firms to engage in OFDI. Based on these arguments, we posited the following hypothesis.

Hypothesis 1: The introduction of carbon emission trading policies positively affects firms’ OFDI.

Moderating role of digitalization

Advanced technologies and data science tools provide innovative methods for addressing environmental concerns.²⁷ These digital tools facilitate not only the accurate monitoring of environmental quality but also the optimization of environmental governance. Specifically, digital environmental governance is becoming a driving force in promoting the green, resilient, and inclusive recovery of the global economy.²⁸ Therefore, digitalization may play a vital role in the impact of carbon emission trading policies on firms’ OFDI.

First, digitalization improves firms’ operational efficiency and management.²⁹ The adoption of advanced digital technologies can allow firms to monitor carbon emissions in real time, optimize production processes, and reduce energy consumption and emissions.³⁰ This can help firms reduce their carbon emission costs, which can improve their market competitiveness and encourage OFDI.¹

Second, digitalization promotes technological innovation and the green transformation of firms.^31,32 It enables firms to develop products and services that align with environmental standards.^33,34 In the presence of carbon emission trading policies, firms with a high degree of digitalization are more likely to realize green transformation and develop low-carbon and environmentally friendly products, thus gaining a competitive advantage in the international market.³⁵

Third, digitalization strengthens firms’ ability to internationalize their operations.³⁶ Through digital platforms, firms can easily access information and resources about international markets, optimize their resource allocation, and improve the efficiency and success rate of their OFDI.³⁷ Digitalization also facilitates technology exchange and resource sharing between enterprises and international partners.³⁸ Based on these arguments, we formulated the following hypothesis.

Hypothesis 2: Digitalization positively moderates the effect of carbon emission trading policies on firms’ OFDI.

Methods

Empirical analysis model

The carbon emission trading policies were announced in the end of 2013 and expanded nationwide after 2017. Therefore, this study used data of Chinese listed firms from 2010 to 2017. All data were sourced from the CSMAR database. Considering that the carbon ETS is a significant policy intervention, we employed a difference-in-differences approach to mitigate endogeneity issues. Firms located in the trial areas of the carbon ETS comprised the treatment group. Equation 1 presents the empirical analysis model.

O F D I_{i, t} = β_{0} + β_{1} E T S_{i} \times P o s t_{t} + λ \sum X_{i, t} + μ_{i} + τ_{t} + ε_{i, t}

(1)

where

O F D I_{i, t}

was the dependent variable denoting the OFDI level of firm i in year t.

E T S_{i} \times P o s t_{t}

measured the influence of carbon emission trading policies on firms’ OFDI. Specifically,

E T S_{i}

denoted regulatory status. It equaled 1 if firm i was situation in one of the trial areas; otherwise, it equaled 0.

P o s t_{t}

equaled 1 for the year 2014 and onwards. Otherwise, it equaled 0. If coefficient

β_{1}

was positive, it indicated that carbon emission trading policies promote firms’ OFDI. If

β_{1}

was negative, it signified that the policies inhibit firms’ OFDI.

X_{i, t}

represented control variables.

μ_{i}

denoted firm fixed effects, while

τ_{t}

denoted year fixed effects.

ε_{i, t}

was the error term.

Variables

We measured the dependent variable ( $O F D I_{i, t}$ ) using the number of overseas affiliated subsidiaries of the firms. This was done due to the absence of specific OFDI data of A-share listed Chinese firms. The number of overseas affiliated subsidiaries can more accurately reflect firms’ OFDI activities because the establishment of each overseas affiliated subsidiary indicates a significant OFDI arrangement in overseas markets.

Regarding the independent variable ( $E T S_{i} \times P o s t_{t}$ ), the carbon ETS was initially implemented in seven trial areas: Beijing, Shanghai, Tianjin, Chongqing, Hubei, Guangdong, and Shenzhen. If firm i was situated in one of these trial areas, $E T S_{i}$ equaled 1. Otherwise, it equaled 0. The carbon ETS was proposed in the end of 2013 but officially launched in 2014. Therefore, we identified 2014 as the first year of the system. If the year was 2014 or later, $P o s t_{t}$ equaled 1; otherwise, it equaled 0.

Considering the influence of confounding factors, we controlled for the following firm characteristics based on firm and year fixed effects: return on assets (Roa), total assets (Asset), net profit (Profit), age of the firm (Age), asset–liability ratio (Dta), and Tobin's Q (TobinQ).

Results and discussion

Descriptive statistics

Table 1 presents summary statistics for all variables. First, the mean value of OFDI was 1.715, and the standard deviation was 5.652, revealing a significant difference in firms’ OFDI levels. The mean of Treat was 0.429, indicating that 42.9% of the firms in the sample were treatment firms. These results provide a preliminary understanding of firm characteristics and lay the foundation for further analysis.

Table 1.

Descriptive statistics.

Variable	Number of observations	Mean	Standard deviation	Min	Max
OFDI	15,228	1.715	5.652	0	141
Treat	15,228	0.429	0.495	0	1
Roa	15,228	0.048	0.037	0.001	0.179
Asset	15,228	22.07	1.410	19.59	27.248
Profit	15,228	18.68	1.596	14.88	23.466
Age	15,228	1.957	0.932	0	3.178
Dta	15,228	0.423	0.220	0.044	0.935
TobinQ	15,228	1.084	0.343	0.657	2.312

OFDI: outward foreign direct investment.

Baseline results

Column 1 of Table 2 presents the regression results after accounting for firm-fixed and year-fixed effects. The coefficient of $E T S_{i} \times P o s t_{t}$ was significantly positive at the 1% level, implying a robust positive impact of carbon emission trading policies on firms’ OFDI. That is, the implementation of the carbon ETS increased Chinese firms’ overseas subsidiaries. Columns 2 and 3 included the control variables. The coefficients of $E T S_{i} \times P o s t_{t}$ were still significantly positive, confirming that the carbon ETS positively influences firms’ OFDI activities. Thus, Hypothesis 1 was verified.

Table 2.

Baseline results.

	(1)	(2)	(3)
Variable	OFDI	OFDI	OFDI
$E T S_{i} \times P o s t_{t}$	0.747***	0.693***	0.680***
	(0.1199)	(0.1185)	(0.1197)
Roa		0.810***	0.879***
		(0.2319)	(0.2192)
Asset		0.925***	0.926***
		(0.1051)	(0.1255)
Profit		0.016	0.021
		(0.0388)	(0.0397)
Age			−0.000
			(0.1051)
Dta			0.533***
			(0.1571)
TobinQ			0.089
			(0.1327)
Firm FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
Observations	14,971	14,971	14,971
R-squared	0.810	0.813	0.813

Note: OFDI: outward foreign direct investment; ETS: emission trading system.

Robust standard errors are written in parentheses.

***p < .01, **p < .05, *p < .1.

Robustness checks

Parallel trend test

We used the difference-in-differences method relying on the assumption that firms involved in the carbon ETS and those not involved exhibited parallel trends in their OFDI levels before the ETS policy's implementation. Consequently, we employed an event study to check the parallel-trend assumption.

Figure 2 illustrates the parallel trend diagram with the horizontal axis indicating the year and the vertical axis displaying the estimation coefficients for each year. The insignificance of the coefficients for the years preceding policy implementation empirically supported the parallel-trend assumption.

Figure 2.

Parallel trends test.

Alleviating endogeneity

The baseline results demonstrated a notably positive correlation between the implementation of the carbon ETS and the escalation of firms’ OFDI. To mitigate endogeneity issues arising from reverse causality in the baseline estimates, all independent and control variables were lagged by one period. As shown in Column 1 of Table 3, $E T S_{i} \times P o s t_{t}$ has a significantly positive coefficient. This result underscores the enduring validity of our main finding that implementing a carbon ETS promotes firms’ OFDI.

Table 3.

Robustness tests.

	(1)	(2)	(3)	(4)	(5)
	Alleviating endogeneity	Controlling for concurrent policies	Removing loss-making firms	Excluding outliers	Using PSM-DID
Variable	OFDI	OFDI	OFDI	OFDI	OFDI
$E T S_{i} \times P o s t_{t}$	0.536***	0.673***	0.679***	0.670***	0.561***
	(0.164)	(0.119)	(0.120)	(0.119)	(0.124)
$G l o a n_{i} \times A f t e r_{t}$		−0.105
		(0.111)
Control	Yes	Yes	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes
Observations	10,897	14,971	14,952	14,971	12,147
R-squared	0.821	0.813	0.813	0.813	0.820

Note: OFDI: outward foreign direct investment; ETS: emission trading system policies.

Robust standard errors are written in parentheses.

***p < .01, **p < .05, *p < .1.

Controlling for concurrent policies

The Chinese government implemented the green loan policy in 2012. To eliminate the potential confounding effects of this policy, we set two dummy variables: $G l o a n_{i}$ and $A f t e r_{t}$ . $G l o a n_{i}$ equaled 1 for treatment firms in the green loan policy and 0 for nontreatment firms. $A f t e r_{t}$ equaled 0 if the year was before 2012. Otherwise, it equaled 1. We then added $G l o a n_{i} \times A f t e r_{t}$ to the baseline estimation, and Column 2 of Table 3 presents the results. The coefficient of $E T S_{i} \times P o s t_{t}$ was still significantly positive. That is, the carbon ETS had a significant positive impact on firms’ OFDI even after controlling for the impact of the concurrent green loan policy.

Removing loss-making firms

Loss-making firms were excluded from the sample to minimize their impact on the policy's effects. As shown in Column 3 of Table 3, $E T S_{i} \times P o s t_{t}$ had a significantly positive impact even after excluding loss-making firms, confirming the robustness of the baseline results.

Excluding outliers

Considering the interference of outliers on the estimation results, we winsorized continuous variables at the 1% level and used the winsorized sample to reestimate the baseline model. Column 4 of Table 3 presents the results. The coefficient of $E T S_{i} \times P o s t_{t}$ remained significantly positive, indicating that the carbon ETS positively affects firms’ OFDI even after excluding outliers.

Using PSM-DID

We used PSM-DID and reestimated the impact of the carbon ETS to overcome the systematic differences between the treatment and control groups and reduce bias in the DID estimates. Using kernel matching for weight determination, we matched the samples in a 1:2 ratio, which yielded 12,147 observations. Column 5 of Table 3 reports the results. The coefficient of $E T S_{i} \times P o s t_{t}$ was slightly higher compared to that in the baseline results but still significantly positive. This result reaffirms the positive impact of the carbon ETS on firms’ OFDI.

Changing the dependent variable

Considering the potential impact of using different indicators, we used the logarithm of long-term equity investment as a proxy for OFDI.³⁹ Column 1 of Table 4 presents the results. The coefficient of $E T S_{i} \times P o s t_{t}$ was significantly positive, indicating that the positive impact of the carbon ETS still held.

Table 4.

Additional robustness tests.

	(1)	(2)	(3)	(4)	(5)	(6)
	Changing the dependent variable	Changing the independent variable	Changing key control variables	Extending the study period	Conducting sensitivity checks	Adding more controls
Variable	OFDI	OFDI	OFDI	OFDI	OFDI	OFDI
$E T S_{i} \times P o s t_{t}$	0.158**	0.413***	0.611***	0.818***	0.471***	0.623***
	(0.0743)	(0.1093)	(0.1184)	(0.1156)	(0.1355)	(0.1207)
GDPper						−0.828**
						(0.3547)
Govexp						0.477***
						(0.1327)
Invest						−0.191
						(0.1829)
Loan						0.202
						(0.2581)
Indstru						0.009
						(0.0100)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Observations	16,544	35,472	14,687	46,102	10,987	13,622
R-squared	0.678	0.683	0.805	0.538	0.848	0.817

Note: OFDI: outward foreign direct investment; ETS: emission trading system.

Robust standard errors are written in parentheses.

***p < .01, **p < .05, *p < .1.

Changing the independent variable

We redefined the independent variable, considering the potential impact of different definitions of the treatment group. Specifically, we included all nationwide treatment areas after 2018 in the treatment group. Column 2 of Table 4 presents the results. The coefficient of $E T S_{i} \times P o s t_{t}$ remained significantly positive, indicating that the positive impact of the carbon ETS still existed.

Changing key control variables

We replaced the key control variables to test the robustness of the baseline estimates. Specifically, we substituted profitability with net profit margin andv substituted Roa with Roe and reestimated the baseline model. Column 3 of Table 4 presents the results. The coefficient of $E T S_{i} \times P o s t_{t}$ remained significantly positive at the 1% level.

Extending the study period

Considering the potential impact of the study period on the results, we expanded the study period: from 2000 to 2022. This change meant the inclusion of nationwide pilots after the end of 2017. Column 4 of Table 4 presents the results. The coefficient of $E T S_{i} \times P o s t_{t}$ remained significantly positive, indicating that the positive impact of the carbon ETS policy still held.

Conducting sensitivity checks

We altered the timeframe of the analysis to determine whether the treatment effects were consistent. Specifically, we tested the robustness of the baseline estimates by using only data for the two years before and after the treatment. Column 5 of Table 4 presents the results. The coefficient of $E T S_{i} \times P o s t_{t}$ remained significantly positive, indicating the consistent positive impact of the carbon ETS.

Adding more controls

Considering the potential impact of omitted variables, we added more control variables to test the robustness of the baseline estimates. Specifically, we added the following variables to take into account the macroeconomic and industry characteristics of the prefecture-level cities in which the firms were located: GDPper denoting the logarithm of per capita regional GDP; Govexp denoting the logarithm of government expenditure, Invest denoting the logarithm of the total fixed asset investment; Loan signifying the logarithm of the loan balance of financial institutions; and Indstru signifying the proportion of the secondary industry. Column 6 of Table 4 presents the results. The coefficient of $E T S_{i} \times P o s t_{t}$ remained significantly positive.

Heterogeneity analyses

Ownership of firms

We categorized the firms into state-owned and private firms to assess whether the impact of the carbon ETS varies based on firms’ ownership. As shown in Columns 1 and 2 of Table 5, the carbon ETS significantly affected private firms’ OFDI, but its impact on state-owned firms was non-significant. The potential reason is that state-owned firms’ OFDI decisions are largely driven by government directives and policies, rendering them less influenced by carbon emission trading policies.

Table 5.

Heterogeneity analyses.

	(1)	(2)	(3)	(4)	(5)	(6)
	State-owned firms	Private firms	Firms belonging to polluting industries	Firms not belonging to polluting industries	High-profitability firms	Low-profitability firms
Variable	OFDI	OFDI	OFDI	OFDI	OFDI	OFDI
$E T S_{i} \times P o s t_{t}$	0.748	0.606***	0.138	0.884***	0.976***	0.154
	(0.597)	(0.125)	(0.206)	(0.292)	(0.239)	(0.124)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Observations	2162	12,516	6606	4199	7417	6831
R-squared	0.918	0.777	0.862	0.801	0.819	0.866

Note: OFDI: outward foreign direct investment; ETS: emission trading system.

Columns 1 and 2 present the results of the heterogeneity analysis based on firms’ ownership, Columns 3 and 4 present the results based on firms’ industry, and Columns 5 and 6 present the results based on firms’ profitability.

Robust standard errors are written in parentheses.

***p < .01, **p < .05, *p < .1.

Industry of firms

The effect of environmental policies may differ based on the industry, considering that pollution levels differ across industries. Some scholars have even reported that environmental policies in China cause pollution transfer among Chinese firms.^40,41 To investigate this, we divided firms into those belonging to polluting industries and those not belonging to polluting industries. Columns 3 and 4 of Table 5 showed a significantly positive coefficient of $E T S_{i} \times P o s t_{t}$ among firms belonging to nonpolluting industries. However, the coefficient was not significant among firms belonging to polluting industries. This finding indicates that the impact of the carbon ETS is primarily concentrated in nonpolluting industries, reinforcing the notion that China's market-based carbon ETS has not led to pollution transfers among firms.

Profitability of firms

We tested whether the effect of the carbon ETS differs based on firms’ profitability. Specifically, we classified firms based on the median of the profitability. Firms exceeding the median profitability were considered highly profitable firms, whereas those below the median profitability were considered less profitable firms. As shown in Columns 5 and 6 of Table 5, highly profitable firms demonstrated a significantly positive coefficient of $E T S_{i} \times P o s t_{t}$ , whereas the coefficient was not significant among less profitable firms. This finding indicates that the carbon ETS primarily affects the OFDI of highly profitable firms. This may because highly profitable firms can more effectively utilize their financial capital and technological expertise to adapt to carbon ETS requirements, thereby facilitating OFDI expansion through lower financing costs and advanced technological innovation.

Mechanism analysis

Existing studies have suggested that the carbon ETS may enhance firms’ OFDI through technological innovation and production cost mechanisms. First, we examined the impact of the carbon ETS on firms’ technological innovation. We replaced the dependent variable with Patent, a variable representing innovation, measured using the number of patent applications. Column 2 of Table 6 presents the results, revealing that the carbon ETS positively affects firms’ technological innovation at the 1% level.

Table 6.

Mechanism testing.

	(1)	(2)	(3)	(4)
Variable	OFDI	Patent	Cost	OFDI
$E T S_{i} \times P o s t_{t}$	0.680***	0.383***	0.004	0.166**
	(0.1197)	(0.0949)	(0.0136)	(0.0741)
Patent				0.097***
				(0.0223)
Control	Yes	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Observations	14,971	14,752	12,682	14,752
R-squared	0.813	0.875	0.965	0.714

Note: OFDI: outward foreign direct investment; ETS: emission trading system.

Column 1 presents the baseline results. Column 2 presents the results for the technological innovation mechanism. Column 3 displays the results for the production cost mechanism.

Robust standard errors are written in parentheses.

***p < .01, **p < .05, *p < .1.

Second, we examined the impact of the carbon ETS on firms’ production costs by replacing the dependent variable with Cost measured using the total operating expenses. As shown in Column 3 of Table 6, the coefficient for Cost was not significant, suggesting that production costs do not facilitate the impact of the carbon ETS on firms’ OFDI.

Third, considering that the carbon ETS had a significant impact on firms’ technological innovation, we added the variable Patent to the baseline model. As shown in Column 4 of Table 6, the main effect estimated by the baseline model weakened significantly (from 0.680*** to 0.166**) when the mediating variable Patent was controlled for. This finding indicates that Patent is the main transmission mechanism.

These results demonstrated that technological innovation induced by the carbon ETS prompts firms to engage in OFDI, not increased production costs.¹⁶ The carbon ETS aims to promote firms’ energy conservation and emission reduction through market mechanisms and drive green transformation.²⁷ Our results suggest that under the pressure of carbon emission trading policies, Chinese firms enhance their competitiveness through technological innovation and expand to overseas markets, indicating firms’ positive responses to carbon policies.³⁷

Moderation analysis

To examine the moderating role of digitalization in the impact of the carbon ETS on firms’ OFDI, we created a variable Digital denoting corporate digitalization. Corporate digitalization was determined based on their digital intangible assets. Then, we created an interaction term of the moderating variables and independent variables to analyze how digitalization moderates the effect of the carbon ETS on firms’ OFDI. Columns 1–3 of Table 7 present the results. The coefficients of the interaction term were positive. This finding indicated that in firms with more advanced digital technologies, the carbon ETS has a more significant positive effect on innovation, which prompts firms to engage in OFDI.

Table 7.

Moderation testing.

	(1)	(2)	(3)
Variable	OFDI	OFDI	OFDI
$D i g i t a l \times E T S_{i} \times P o s t_{t}$	0.944***	0.896***	0.610**
	(0.246)	(0.278)	(0.275)
Digital		−0.227	−0.009
		(0.203)	(0.190)
$E T S_{i} \times P o s t_{t}$		0.145	0.131
		(0.117)	(0.115)
Control	No	No	Yes
Firm FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
Observations	9174	9174	9174
R-squared	0.726	0.726	0.737

Note: OFDI: outward foreign direct investment; ETS: emission trading system.

Robust standard errors are written in parentheses.

***p < .01, **p < .05, *p < .1.

The positive moderation of digitalization may be due to the fact that advanced digital technologies can more effectively integrate and analyze data related to the carbon ETS and firm operations.⁴² They can provide real-time feedback and facilitate intelligent decision-making, helping firms adapt to the requirements of the carbon ETS and allocate resources more effectively.⁴³ Furthermore, digital platforms can facilitate the sharing and exchange of innovation resources and experiences among firms,³¹ promoting the diffusion and application of innovation achievements, which can further amplify the positive impact of the carbon ETS on firms’ innovation and OFDI. In the era of globalization and digitalization, companies with robust digital capabilities are better positioned to capitalize on international market opportunities and expand their operations through OFDI.⁴⁴ They use their innovation strengths to gain a competitive advantage in the global market, effectively addressing the challenges and opportunities presented by the carbon ETS and climate change. Thus, Hypothesis 2 was supported.

Conclusion

Our results show that, first, the carbon ETS has a significant positive impact on the OFDI of Chinese listed firms. This result remains valid even after several robustness checks. Second, the impact of the carbon ETS is apparent among private firms, firms not belonging to polluting industries, and highly profitable firms. It is not obvious among state-owned firms, firms belonging to polluting industries, and less profitable firms. Third, technological innovation mediates the impact of the carbon ETS on firms’ OFDI, not increased production costs. Fourth, digitalization enhances the positive effects of the carbon ETS on firms’ OFDI.

These results present salient implications for deepening two-way investments across countries in the era of digitalization. First, developing economies should actively adopt and promote trading-based carbon systems. They should focus on efficient and standardized carbon emission trading systems informed by insights from pilot areas, while ensuring a timely and clear policy direction. Second, considering that the carbon ETS promotes Chinese firms’ OFDI through technological innovation, the Chinese government should provide subsidies to promote innovation and let firms use technological upgradation to offset increased costs. Third, considering that the impact of the carbon ETS varies based on firms’ ownership, industry, and profitability, governments should consider firm heterogeneity and implement flexible environmental regulations. Fourth, considering that corporate digitalization enhances the positive effects of the carbon ETS on firms’ OFDI, governments should actively help firms conduct digital transformation by providing special subsidies, tax incentives, and other incentives.

Footnotes

Acknowledgments

We would like to thank Editage () for English language editing.

ORCID iD

Zhihui Zhao

Author contributions/CRediT

Jing Cheng: conceptualization, data curation, writing—original draft preparation; Zhihui Zhao: investigation, methodology, software, and writing—original draft preparation; Wei Liu: supervision, funding acquisition, and writing—reviewing and editing; Ruzhen Fang: data curation, formal analysis, and investigation; Fengwei Wang: resources, validation, and visualization.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the First Phase of the International Students Course “International Investment” in China, School of International Education, Wuhan University; the 2023 School of Economics and Management, Wuhan University English Course Construction Project for International Students—International Investment; and the Major Project of China National Social Science Fund (Grant Nos. 202507, 1201-413400008 and 19VMG003).

Conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

The datasets generated during and/or analyzed during the current study are available on reasonable request.

References

Liu

Zhao

Wen

, et al. Environmental regulation and OFDI: Evidence from Chinese listed firms. Econ Anal Policy 2022; 75: 191–208. http://doi.org/10.1016/j.eap.2022.04.013

Tan

Gao

Liu

. Can environmental information disclosure improve energy efficiency in manufacturing? Evidence from Chinese enterprises. Energies 2024; 17(10): 2342. http://doi.org/10.3390/en17102342.

Gao

Feng

Cao

. The spatial spillover effect of innovation city policy on carbon efficiency: Evidence from China. Singap Econ Rev 2024. http://doi.org/10.1142/S0217590824500024.

Yang

Liu

. Did carbon trade improve green production performance? Evidence from China. Energ Econ 2021; 96. http://doi.org/10.1016/j.eneco.2021.105185.

Zhang

Cheng

Ren

, et al. Effects of carbon emission trading system on corporate green total factor productivity: Does environmental regulation play a role of green blessing? Environ Res 2024; 248. http://doi.org/10.1016/j.envres.2024.118295.

Bai

. Evaluation of China’s carbon emission trading policy from corporate innovation. Financ Res Lett 2021; 39. http://doi.org/10.1016/j.frl.2020.101565.

Ouyang

. Can carbon emission trading policy promote green innovation? The perspective of corporate operating difficulties. J Clean Prod 2023; 420. http://doi.org/10.1016/j.jclepro.2023.138473.

Liu

Sun

. Assessing the impact of emissions trading scheme on low-carbon technological innovation: Evidence from China. Environ Impact Asse 2021; 89: 106589. http://doi.org/10.1016/j.eiar.2021.106589

Feng

Zhao

Yan

. Does carbon emission trading policy has emission reduction effect? An empirical study based on quasi-natural experiment method. J Environ Manage 2024; 351. http://doi.org/10.1016/j.jenvman.2023.119791.

10.

Gong

Liu

, etal. Environmental regulation, hidden economy, and China’s outward foreign direct investment. Chin J Popul Resour Environ 2020; 18: 35–41. http://doi.org/10.1016/j.cjpre.2021.04.009

11.

Liu

Luo

. Environmental regulation and outward foreign direct investment: Evidence from Chinese firms’ overseas mergers and acquisitions. Appl Econ 2024; 56: 2209–2222. http://doi.org/10.1080/00036846.2023.2277693

12.

Zheng

Shi

. Multiple environmental policies and pollution haven hypothesis: Evidence from China’s polluting industries. J Clean Prod 2017; 141: 295–304. http://doi.org/10.1016/j.jclepro.2016.09.091

13.

Naughton

. To shut down or to shift: Multinationals and environmental regulation. Ecol Econ 2014; 103: 68. http://doi.org/10.1016/j.ecolecon.2014.05.004

14.

Porter

Vanderlinde

. Toward a new conception of the environment-competitiveness relationship. J Econ Perspect 1995; 9: 97–118. http://doi.org/10.1257/jep.9.4.97

15.

Rubashkina

Galeotti

Verdolini

. Environmental regulation and competitiveness: Empirical evidence on the porter hypothesis from European manufacturing sectors. Energ Policy 2015; 83: 288–300. http://doi.org/10.1016/j.enpol.2015.02.014

16.

Yan

. Do green technology innovations contribute to carbon dioxide emission reduction? Empirical evidence from patent data. Technol Forecast Soc 2019; 146: 297–303. http://doi.org/10.1016/j.techfore.2019.06.010

17.

Helpman

Melitz

Yeaple

. Export versus FDI with heterogeneous firms. Am Econ Rev 2004; 94: 300–316. http://doi.org/10.1257/000282804322970814

18.

Elliott

RJR

Zhou

. Environmental regulation induced foreign direct investment. Environ Resour Econ 2013; 55: 141–158. http://doi.org/10.1007/s10640-012-9620-0

19.

Gao

Zhou

Wan

. Unlocking sustainability potential: the impact of green finance reform on corporate ESG performance. Corp Soc Resp Env Ma 2024b; 31: 4211–4226. http://doi.org/10.1002/csr.2801

20.

Song

Zhou

, et al. Does environmental regulation affect energy efficiency in China’s thermal power generation? Empirical evidence from a slacks-based DEA model. Energ Policy 2014; 66: 537–546. http://doi.org/10.1016/j.enpol.2013.10.056

21.

Liao

Shi

. Public appeal, environmental regulation and green investment: Evidence from China. Energy Policy 2018; 119: 554–562. http://doi.org/10.1016/j.enpol.2018.05.020

22.

Brunnermeier

Cohen

. Determinants of environmental innovation in US manufacturing industries. J Environ Econ Manag 2003; 45(2): 278–293. http://doi.org/10.1016/S0095-0696(02)00058-X

23.

Frondel

Horbach

Rennings

. What triggers environmental management and innovation? Empirical evidence for Germany. Ecol Econ 2008; 66: 153–160. http://doi.org/10.1016/j.ecolecon.2007.08.016

24.

Kemp

Pontoglio

. The innovation effects of environmental policy instruments-A typical case of the blind men and the elephant? Ecol Econ 2011; 72: 28–36. http://doi.org/10.1016/j.ecolecon.2011.09.014

25.

Cai

, et al. Does environmental regulation drive away inbound foreign direct investment? Evidence from a quasi-natural experiment in China. Sci Total Environ 2016; 123: 73–85. http://doi.org/10.1016/j.scitotenv.2020.140810

26.

Chung

. Environmental regulation and foreign direct investment: Evidence from South Korea. J Dev Econ 2014; 108: 222–236. http://doi.org/10.1016/j.jdeveco.2014.01.003

27.

Zhao

Liu

Ning

, et al. Can China go green through digitalization? Evidence from China’s listed firms. Int Rev Econ Financ 2024; 94: 103398. https://doi.org/10.1016/j.iref.2024.103398

28.

Agboola

Tunay

. Urban resilience in the digital age: The influence of information-communication technology for sustainability. J Clean Prod 2023; 428: 139304. https://doi.org/10.1016/j.jclepro.2023.139304

29.

Quttainah

Ayadi

. The impact of digital integration on corporate sustainability: Emissions reduction, environmental innovation, and resource efficiency in the European. J Innov Knowl 2024; 9(3): 100525. http://doi.org/10.1016/j.jik.2024.100525

30.

Jin

Gao

. Digital transformation and pollution emission of enterprises: Evidence from China’s microenterprises. Energy Rep 2023; 9: 552–567. http://doi.org/10.1016/j.egyr.2022.11.169

31.

Lee

Xiao

. How does information and communication technology affect renewable energy technology innovation? International evidence. Renew Energ 2022; 200: 546–557. http://doi.org/10.1016/j.renene.2022.10.015

32.

Geng

Zheng

. Digitization, perception of policy uncertainty, and corporate green innovation: A study from China. Econ Anal Policy 2023; 80: 544–557. http://doi.org/10.1016/j.eap.2023.09.006

33.

Cai

Zhu

Zhang

, et al. Can direct environmental regulation promote green technology innovation in heavily polluting industries? Evidence from Chinese listed companies. Sci Total Environ 2020; 746. http://doi.org/10.1016/j.scitotenv.2020.140810.

34.

Luo

Yimamu

, et al.

Digitalization and sustainable development: How could digital economy development improve green innovation in China?

Bus Strateg Environ 2023; 32(4): 1847–1871. http://doi.org/10.1002/bse.3223

35.

Zhang

Lyu

, et al. Digital economy: An innovation driving factor for low-carbon development. Environ Impact Asses 2022; 96: 106281. http://doi.org/10.1016/j.eiar.2022.106821

36.

Feliciano-Cestero

Ameen

Kotabe

, et al. Is digital transformation threatened? A systematic literature review of the factors influencing firms’ digital transformation and internationalization. J Bus Res 2023; 157: 113546. http://doi.org/10.1016/j.jbusres.2022.113546

37.

Liu

Zhao

, et al. Is more digital always better? A nonlinear perspective on corporate digital transformation and OFDI. Int Rev Financ Anal 2024; 96: 103649. http://doi.org/10.1016/j.irfa.2024.103649

38.

Luo

. A general framework of digitization risks in international business. J Int Bus Stud 2022; 53(2): 344–361. http://doi.org/10.1057/s41267-021-00448-9

39.

Zhang

Rong

. Green innovation and firm performance: Evidence from listed companies in China. Resour Conserv Recy 2019; 144: 48–55. http://doi.org/10.1016/j.resconrec.2019.01.023

40.

Zhang

. FDI and environmental regulations in China. J Asia Pac Econ 2008; 13: 332–353. http://doi.org/10.1080/13547860802131326

41.

. Assessing the impact of environmental regulation on pollution abatement and collaborative emissions reduction: Micro-evidence from Chinese industrial enterprises. Environ Impact Asses 2020; 82: 106382. http://doi.org/10.1016/j.eiar.2020.106382

42.

Arfi

Hikkerova

Sahut

. External knowledge sources, green innovation and performance. Technol Forecast Soc 2018; 129: 210–220. http://doi.org/10.1016/j.techfore.2017.09.017

43.

Wang

Gong

Yang

. Unveiling the relation between digital technology and low-carbon innovation: Carbon emission trading policy as an antecedent. Technol Forecast Soc 2024; 205: 123522. http://doi.org/10.1016/j.techfore.2024.123522

44.

Zhu

. The influence of digital transformation on Chinese firms' outward foreign direct investment. Financ Res Lett 2025; 72: 106567. http://doi.org/10.1016/j.frl.2024.106567