Foreign direct investment and the structural transition of energy consumption: impact and mechanisms

Abstract

The development of renewable energy is an important pathway to realize the structural transition of energy consumption. This study investigates the impact and mechanism of foreign direct investment (FDI) on transforming the host countries’ energy consumption structure to renewable energy. We conducted empirical tests using cross-country panel data of 65 economies from 2000 to 2020. The results revealed that FDI significantly negatively impacts the structural transition of energy consumption in host countries. The mechanism analysis established that the industry transfer and technology spillover effects of FDI are not conducive to transforming the host country’s energy consumption structure to renewable energy. Heterogeneity analysis revealed that FDI has a greater negative impact on the renewable energy consumption structure in low-income countries and non-OECD countries, as well as in countries with high levels of agricultural and manufacturing development and low levels of service sector development. Moreover, the Kyoto Protocol is not conducive to the increase in the proportion of renewable energy consumption in host countries. By contrast, the negative impact of FDI has been effectively mitigated after signing the Paris Agreement. Further analysis determined that improving governance capacity effectively inhibits the negative impact of FDI on the renewable energy consumption structure. The findings provide a theoretical basis for accurately identifying the effects of FDI on energy structure, while also providing policy insights for countries to formulate and improve their sustainable development efforts.

Introduction

The Paris Agreement states that all parties should strengthen the overall response to the threat of climate change and that the long-standing crude development model and massive global greenhouse gas (GHG) emissions have had extensive and far-reaching impacts on the natural environment, ecosystems, and human society. Accelerating the energy transition and promoting the application and development of renewable energy is an important part of achieving the low-carbon transition of the global development model, as well as a key path and an inevitable choice for achieving global carbon neutrality. However, global energy transition still faces huge challenges. In terms of international capital flows, foreign direct investment (FDI) has played an important role in promoting economic growth and technological progress, but the environmental problems it brings cannot be ignored. How to accurately identify the energy effects of FDI and effectively utilize foreign capital to improve the quality of economic development has become a key issue for policy authorities globally to consider for the sake of achieving carbon neutrality.

Existing literature mostly focuses on the impacts of financial development, policy-driven incentives, and resource endowments on the consumption structure of renewable energy (Yang et al. 2022; Zhang and Xie, 2023; Guo et al. 2023; Khan et al. 2023). However, the research on the impact of FDI on the renewable energy consumption structure started late and has not derived consistent conclusions (Khan et al. 2021; Li et al. 2022). Moreover, in terms of sample selection, studies focused on geographically adjacent or economically similar economies (Tan and Uprasen, 2022; Akpanke et al. 2023). Therefore, to further delineate the specific impacts and mechanisms of FDI on energy structure transformation, this study seeks to improve the existing body of research in four ways.

First, we used cross-country panel data from 65 countries and regions to examine the impact of FDI on renewable energy, which not only includes developed economies with a high level of renewable energy consumption but also emerging economies where the consumption structure is in its initial stage. Accordingly, we can examine the effect of FDI more comprehensively and compare the characteristics of different types of economies.

Second, in terms of mechanism research, this article empirically tests how foreign direct investment affects the renewable energy use structure of host countries through industrial transfer and technology spillovers, providing empirical evidence for accurately identifying the internal mechanism of foreign direct investment affecting the renewable energy use structure.

Third, from the perspective of governance, we investigate the moderating effect of FDI on the renewable energy consumption structure, which provides a logical basis for economies to fully utilize government policy tools to establish a proactive and efficient energy transition policy system and expand the scope of research on the transition of energy structure.

Fourth, we analyze the heterogeneous impact of FDI on the renewable energy consumption structure based on external characteristics or shocks such as income level, industrial advantages, and climate agreements, which helps different countries or economies formulate targeted solutions in light of their own situation and external environment. The conclusions of this study are both theoretically innovative and of strong practical significance.

The rest of this paper is organized as follows. Section “Literature review” presents the literature review. Section “Theoretical analysis and hypothesis development” presents the theoretical analysis and hypothesis formulation, which theoretically and systematically examines the impact mechanism of FDI affecting the energy structure transition. Section “Research design” details the empirical research design, which lays out the model, variable definition, and sample selection. Section “Empirical results and analysis” presents the empirical analysis results, which include the benchmark regression, robustness test, mechanism test, and heterogeneity analysis. Section “Further analysis: role of governance” further analyzes the influence of governance on the energy structure transition. Finally, concluding remarks are presented in Section “Conclusions and discussion”.

Literature review

The relationship between FDI and environmental pollution has been a highly publicized and controversial proposition, with existing studies focused on the two classic “pollution haven” and “pollution halo” hypotheses of FDI (Copeland and Taylor, 1994; Kirkpatrick and Shimamoto, 2008). The “pollution haven” hypothesis suggests that FDI will exacerbate environmental pollution in the host country since stringent environmental regulations will induce pollution-intensive industries to move to countries with lower environmental standards through FDI, thus exerting enormous pressure on the latter’s ecosystem (Lu et al. 2023; Opoku and Boachie, 2020). To maintain their international competitive advantage, some countries will adopt a bottom-up development strategy to attract a large number of foreign-funded enterprises, which ultimately leads to the widespread use of high-pollution and high-energy-consumption production processes by domestic enterprises, and results in the continuous deterioration of energy efficiency and environmental quality (Cheng et al. 2020; Muhammad et al. 2021). The “pollution halo” hypothesis argues that foreign-funded enterprises will bring advanced production technology, green business practices, and scientific and effective management experience, which is conducive to improving the host country’s environmental performance (Demena and Afesorgbor, 2020). Through the technology spillover and demonstration effects, FDI can improve the production efficiency of host country enterprises, strengthen the investment in environmental governance, and enable green development (Hille et al. 2019), which will promote the green transition of the host country’s economy and its capacity for sustainable development (Phung et al. 2022). Foreign-funded enterprises have cleaner and more efficient production technology (Kim and Adilov, 2012). After entering the host country, they will squeeze out the highly polluting and energy-consuming local enterprises, forcing others in the industry to undergo a green production chain transition, significantly reducing the host country’s environmental pollution.

Renewable energy development is a core element in realizing the United Nations Sustainable Development Goals and the third energy transition. Studies have investigated the transition of the energy consumption structure mainly from three perspectives: financial development, government policy (Yang et al. 2022; Zhang and Xie, 2023), and resource endowment (Guo et al. 2023; Khan et al. 2023). Regarding the financial development perspective, such development promotes the clean transition of energy structure by incentivizing renewable energy research and development (R&D) investment and renewable energy capacity investment (Alsagr, 2023; Tang and Zhou, 2023). With the extensive promotion of the concept of sustainable development, the financial sector has begun to develop innovative supply-side financial products, including green credit, bonds, insurance, asset-backed securities, and financial leasing (Dong et al. 2023; He et al. 2019), which provide adequate financing channels and support for renewable energy R&D and capacity investment (Sun and Chen, 2022). Long-term and stable funding sources can effectively mitigate the financial risks of enterprise innovation investment (Zhang et al. 2023). The reduction of external financing constraints can help improve enterprises’ risk-bearing ability to change the R&D investment decisions of micro-entities (Alharbi et al. 2023; Guo et al. 2023), which in turn affect the total amount of R&D investment by renewable energy enterprises and the technology choices of non-renewable energy firms entering the renewable energy sector (Wang et al. 2022). In a related study on the impact of FDI on renewable energy using sample data from developing countries, Li et al. (2022), argued that the introduction of FDI in host countries can result in industrial restructuring and redistribution of resource factors and enhance the energy efficiency of the host countries by eliminating obsolete technologies and energy-intensive industries, thus promoting a green and low-carbon transition of the energy consumption structure. Based on the sample data of countries along the Belt and Road, Khan et al. (2021) confirmed that FDI significantly negatively impacts renewable energy consumption in these countries and concluded that FDI does not incentivize the use of renewable energy. Akpanke et al. (2023) examined data from 15 countries in West Africa and observed temporal inconsistency in the impact of FDI on renewable energy consumption in host countries, with a significant negative impact on renewable energy consumption in the short term and a positive impact in the long term. Tan and Uprasen (2022) investigated Brazil, Russia, India, China, and South Africa and found that the impact of FDI on renewable energy consumption is affected by environmental regulations. FDI has a significant negative effect on renewable energy consumption when environmental regulations are low, but a positive effect when they exceed a certain threshold, and both formal and informal environmental regulations have a similar threshold effect.

Theoretical analysis and hypothesis development

FDI, industrial transfer, and energy structure transition

FDI brings about transnational industrial transfers, which impacts the host country’s industrial development and energy structure transition. This is known as the industrial transfer effect. First, analyzing from the perspective of energy characteristics. Compared to fossil energy, renewable energy has greater risks and uncertainties in terms of technology development and application. The large-scale utilization of renewable energy requires substantial and long-term financial support. Given the long investment cycle, the absence of an effective cost internalization mechanism results in a lack of incentives for foreign enterprises to adopt renewable energy (Wang and Fan, 2023). Moreover, regarding resource availability and raw material costs, fossil energy has already formed a complete supply chain system in terms of extraction, processing, and sales. Foreign-funded enterprises tend to set up their enterprises in energy-rich and low-priced countries or regions and use a large amount of fossil energy as the main input for their production. Second, analyzing from the perspective of industrial selection, undertaking high pollution and high energy-consuming division of labor is the result of competition and cooperation among countries. Based on the theory of comparative advantage, enterprises will transfer the processing link of low-value-added products to less developed regions to take advantage of cheap resources and labor and only retain the high-value-added links such as product design, R&D, and sales to increase economic efficiency (Arkolakis et al. 2018). However, compared with the general trade link, the processing trade production link requires more energy consumption (Nawaz and Rahman, 2024), which brings serious pollution problems and higher pollution control costs. Moreover, the pressure of environmental regulation forces enterprises in developed regions to make heterogeneous business decisions of transition, upgrading, or transfer, which prompts the FDI to transfer the relevant production links to countries or regions with lower environmental regulations. Developing countries are willing to absorb FDI to achieve rapid capital accumulation and high economic growth by taking advantage of the cost advantages of production factors, such as energy and labor, and lowering the standards and intensity of environmental regulations (Xing and Kolstad, 2002). Therefore, we formulate the following hypothesis.

H1: The industrial transfer effect of FDI has a negative impact on the transition of the host country’s energy consumption structure to renewable energy.

FDI, technology spillovers, and energy structure transition

FDI has significant spillover effects on technological innovation in host countries (Aitken and Harrison, 1999; Ali et al. 2023; Javorcik et al. 2018), which influences the industrial supply chain. The quality requirements imposed by downstream foreign-funded enterprises on intermediate products will compel upstream domestic-funded enterprises to undertake technological innovations. The increasing potential returns on R&D investments will help incentivize such technological innovations (Carluccio and Fally, 2013). The entry of upstream foreign-funded firms brings high-quality intermediate products and supply chain management experience to downstream domestic firms. Furthermore, by learning and imitating the advanced business models and production technologies of the former, downstream domestic firms are prompted to improve their production processes and internal organizational structures, thus enhancing their technological innovation capabilities (Burstein and Monge-Naranjo, 2009). In addition, FDI also affects peer competition. Foreign-funded enterprises have a clear demonstration effect, which is manifested through providing high-quality products to the market supply to squeeze the domestic enterprises’ market space, driving them to improve product quality and reduce production costs through technological innovation to gain a competitive advantage. Foreign-funded enterprises also exhibit a labor reservoir effect. While undergoing localization transitions, these enterprises provide advanced management practices and production technologies as references for the host country’s workforce. In the long term, the enhancement of human capital in the host country helps reduce the sunk costs and information gathering costs associated with R&D activities for domestically funded enterprises (Blonigen et al. 2007), thereby improving the technological innovation capabilities of host country firms.

However, the impact of the technology spillover effect of FDI on the renewable energy consumption structure is polarized. First, environment-related technological innovation helps to realize a cleaner transition of the host country’s energy consumption structure. Such innovation promotes energy efficiency, thereby exerting a ‘throttling effect’ on fossil energy. By reducing the costs of renewable energy production and application, expanding the range of renewable energy options, shortening the construction cycle of renewable energy infrastructure, mitigating the risks associated with renewable energy development, and lowering the threshold for renewable energy consumption, innovation attracts more enterprises to enter the renewable energy sector and facilitates the transition to a cleaner energy consumption structure (Wei et al. 2023). Second, non-environment-related technological innovation may increase fossil energy consumption in the host country, which has a negative impact on its energy consumption structure. The main objective of non-environment-related technological innovations is to expand the production scale and reduce the cost of non-energy factors rather than applying energy-saving technologies or increasing the proportion of renewable energy consumption (Ahmad et al. 2020). The technology spillover effects of FDI, which are mainly motivated by industry transfer and profit-seeking capital, are dominated by non-environment-related technological innovations (Behera and Sethi, 2022). Although such innovations can improve production efficiency and manufacturing processes, they are characterized by high energy consumption. The increase in production capacity under the constraints of the economic development goals will lead to increased investment and disorderly expansion of production scale, exacerbating overcapacity and energy consumption. This will not be conducive to transforming host countries into a cleaner energy structure.

In summary, the technology spillover effect of FDI may promote the cleaner transformation of the host country’s energy consumption structure by improving energy use efficiency. Simultaneously, it may also jeopardize the cleaner transformation of the host country’s energy consumption structure by expanding the scale of production and causing the large-scale use of fossil energy. Accordingly, we proposed the following hypotheses.

H2a: Technology spillovers from FDI negatively affect the transition of the host country’s energy consumption structure to renewable energy.

H2b: Technology spillover effect of FDI has a positive impact on the transition of the host country’s energy consumption structure towards renewable energy.

FDI, governance, and energy structure transition

Host country-specific institutional arrangements and governance capacity have a positive impact on the entry of high-quality foreign-funded firms and the transition to a cleaner energy structure (Huynh and Hoang, 2019), mainly reflected in two aspects. First, the threat of terrorism and extremism, political unrest, and regime change will hinder the normal conduct of production and business activities, leading to a decline in the willingness of high-quality foreign-funded enterprises to enter and a large inflow of speculative foreign-funded enterprises, which in turn will have a serious negative impact on the host country’s economy and environment. Therefore, maintaining political stability and improving government effectiveness will help the host country absorb high-quality FDI. The improvement of government efficiency will affect foreign enterprises’ investment decisions, making them more inclined to enter the local area through joint ventures rather than wholly owned corporations. Simultaneously, it will also increase the likelihood of high-tech enterprises entering the country, enhancing the spillover of environment-related technology and its absorption by local enterprises, ultimately leading to the sustainable improvement of the host country’s renewable energy consumption structure. Second, the rule of law, regulatory quality, and corruption control provide favorable institutional safeguards for implementing host country environmental policies and screening foreign-funded enterprises. Increased government corruption means that local governments are dominated by their own economic interests (Wang et al. 2020), and foreign firms are willing to bribe the government to relax environmental controls and audits. Decreased sewage disposal and treatment costs will effectively incentivize foreign firms to gain economic benefits by using many high-pollution and high-energy-consumption production processes, harming the host country’s energy structure transformation (Yu and Liu, 2024). Moreover, the poor legal system and quality of regulation will lead to over-inflated demand for FDI. The crude development model reduces the entry threshold of FDI, and many foreign-funded enterprises that do not meet the environmental and energy consumption standards gain approval from local governments, which leads to a sharp rise in environmental pollution and fossil energy consumption. Accordingly, we formulate the following hypothesis.

H3: Improvement in governance capacity can help curb the negative impact of FDI on the transition of the host country’s energy consumption structure to renewable energy.

Research design

Benchmark model

Following the method of Su et al. (2024), we conducted a Hausman test and found a p-value of 0.000, indicating that the regression results in this study should use a fixed effects model rather than a random effects model (Hausman and Taylor, 1981). On this basis, to verify the relationship between FDI and the structure of energy consumption, we follow the methodology of Zhang et al. (2023) and conduct the benchmark regression as follows:

1

${RNEW}_{it}={\alpha }_0+{\alpha }_{1}Ln{IFDI}_{it}+{\alpha }_2{CONT}_{it}+{\mu }_i+{\phi }_t+{\epsilon }_{it}$where, i denotes country and t denotes year, ${\mathrm{RNEW}}_{it}$ denotes energy consumption structure,$Ln{\mathrm{IFDI}}_{it}$ denotes foreign direct invest, ${\mathrm{CONT}}_{it}$ denotes control variables, which control other variables that may affect the energy structure of the economy, ${\mu }_i$ denotes country fixed effects, ${\phi }_t$ denotes year-fixed effects, and ${\epsilon }_{it}$ is a random perturbation term. In Eq. (1), the coefficient of interest is ${\alpha }_1$. If the estimated value exceeds 0, it indicates that FDI transforms the energy consumption structure. If the value is less than 0, it has a negative impact on the transformation of the energy consumption structure. We conduct empirical analysis using Stata 16 software.

Mechanism test model

To examine the role mechanism of FDI affecting the energy consumption structure, we refer to Wei et al. (2023) method as follows:

2

$X_{it}={\beta }_0+{\beta }_{1}Ln{IFDI}_{it}+{\beta }_2{CONT}_{it}+{\mu }_i+{\phi }_t+{\epsilon }_{it}$

3

${RNEW}_{it}={\gamma }_0+{\gamma }_1{X}_{it}+{\gamma }_{2}Ln{IFDI}_{it}+{\gamma }_3{CONT}_{it}+{\mu }_i+{\phi }_t+{\epsilon }_{it}$where, $X_{it}$ is the mechanism variable, including industrial transfer ($Ln{\mathrm{IRT}}_{it}$) and technological innovation ($Ln{\mathrm{Pat}}_{it}$), and the rest are defined as above. In Eqs. (2) and (3), we focus on the magnitude and sign of the coefficients ${\beta }_1$ and ${\gamma }_1$, which are expected to be ${\beta }_1>0$ and ${\gamma }_1<0$, respectively, according to the previous analysis. That is, FDI may harm the energy consumption structure of the host country through industrial transfer and technological spillovers.

Definition of variables

Explained variables

Energy consumption structure (RNEW). This study focuses on the structure of energy consumption. Much of the literature already uses the carbon dioxide emissions of a country or economy as a proxy variable for energy consumption. However, measuring energy consumption only using GHG emissions cannot accurately reflect the changes in the energy consumption structure. Therefore, to measure the degree of transformation of energy consumption structure to renewable energy, we selected the proportion of renewable energy consumption to total energy consumption. The larger the value of this indicator, the higher the degree of transition to a renewable energy consumption structure.

Explanatory variables

Foreign direct investment (LnIFDI). Compared with the inventory data, the flow data can more intuitively reflect the dynamic effect of the influence of FDI on the host country’s energy consumption structure. Therefore, the proportion of the host country’s FDI flow to the current year’s gross domestic product (GDP) is chosen as a proxy variable for FDI, with a larger value indicating that the economy receives more FDI.

Mechanism variables. Industrial transfer (LnIRT). This study analyzed the direction and scale of international industrial transfers in terms of value flows, with changes in the value-added of domestic industries caused by foreign demand regarded as industrial transfers from foreign countries to host countries. Referring to Koopman et al. (2010), the trade flows of a country or economy are decomposed, which can be split into:

4

$E_{r*}={\underbrace{V_r{B}_{rr}\sum _{s\ne r}{Y}_{rs}}}_{\left(a\right)}+{\underbrace{V_r{B}_{rr}\sum _{s\ne r}{A}_{rs}{X}_{ss}}}_{\left(b\right)}+{\underbrace{V_r{B}_{rr}\sum _{s\ne r}\sum _{t\ne r,s}{A}_{rs}{X}_{st}}}_{\left(c\right)}+{\underbrace{V_r{B}_{rr}\sum _{s\ne r}{A}_{rs}{X}_{sr}}}_{\left(d\right)}+{\underbrace{F{V}_r}}_{\left(e\right)}$where, r represents the home country and s and t represent the foreign country; (a) represents the domestic value-added embedded in exports of final goods and services absorbed by the direct importer, (b) represents the domestic value-added embedded in exports of intermediate inputs used by the direct importer to produce the needed products for the domestic market, (c) represents the domestic value-added embedded in intermediate exports used by the direct importer to produce goods for a third country (indirect value-added exports), (d) represents the domestic value-added embedded in intermediate exports used by direct importers to produce goods for shipment back to their sources (returned domestic value-added), and (e) represents the foreign value-added embedded in total exports (foreign value-added for export).

According to the decomposition of total trade flows, after excluding foreign value-added and value-added returned to the home country, the net value of industrial transfers from foreign to domestic industries can be expressed as follows:

5

$\begin{array}{c}IR{T}_r=E_{r*}-V_r{B}_{rr}\sum _{s\ne r}{A}_{rs}{X}_{sr}-F{V}_r\\ =\underset{\left(f\right)}{\underbrace{V_r{B}_{rr}\sum _{s\ne r}{Y}_{rs}}}+\underset{\left(g\right)}{\underbrace{V_r{B}_{rr}\sum _{s\ne r}{A}_{rs}{X}_{ss}+V_r{B}_{rr}\sum _{s\ne r}\sum _{t\ne r,s}{A}_{rs}{X}_{st}}}\end{array}$where, $IR{T}_r$ represents the net value of foreign industrial transfers to the home country, and the larger the value, the more the home country receives the net value of foreign industrial transfers. Furthermore, when decomposed according to the demand for different products, (f) represents the demand for final products transferred from foreign countries to home countries, and (g) represents the demand for intermediate products transferred from foreign countries to home countries.

Technological innovation (LnPat). Existing studies mainly measure the technological innovation capacity of an economy in terms of R&D inputs and outputs. Technological innovation inputs are generally measured using the ratio of R&D inputs to GDP, while outputs are generally measured using the number of patent applications. We used the number of patent applications as a proxy variable for technological innovation capacity to test whether FDI affects the energy consumption structure of a host country by increasing or decreasing its technological innovation capacity.

Control variables

This study controls for the following variables: (1) level of economic development (LnPerGDP), which is expressed using the per capita GDP of each economy. (2) Capital intensity (LnPercaptial), which is expressed using the economy’s per capita capital stock. (3) Industrial structure (LnStructure), which is expressed using the ratio of value-added in the tertiary industry to that in the secondary industry. (4) Urbanization rate (LnUrbanization), expressed using the ratio of each economy’s urban population to its total population, with the larger the value indicating the higher the degree of urbanization in each economy. (5) Trade openness (LnOpenness), expressed as the ratio of each economy’s total imports and exports to its GDP.

Data sources

We selected 65 economies from 2000 to 2020 as the research object1, excluded the missing sample data within the research interval, conducted 2.5% and 97.5% percentile shrinkage of each continuous variable to mitigate the effect of outliers2, and finally obtained 1312 observation samples. The data on energy structure indicators are from the International Energy Agency (IEA), FDI data come from the UNCTAD database, and the data on mechanism variables and control variables come from the World Bank (WDI) database and the Organization for Economic Co-operation and Development (OECD) TiVA database3. The descriptive statistics of each variable are presented in Table 1.

Table 1. Descriptive statistics of the key variables.

VarName	Obs	Mean	SD	Median	Min	Max
RNEW	1312	0.2000	0.1774	0.1467	0.0001	0.7596
LnIFDI	1312	1.4335	0.8360	1.3111	−0.0399	3.7440
LnIRT	1312	11.0163	1.5699	11.0985	5.8514	14.5445
LnPat	1213	6.8406	2.3424	6.7935	2.1972	12.5126
Gov	1250	−0.0123	1.9283	0.2414	−3.6054	2.8130
LnPerGDP	1312	9.4932	1.2059	9.6754	6.4472	11.3021
LnPercaptial	1312	8.0092	1.1797	8.1693	5.2067	9.8847
LnStructure	1312	1.1889	0.3051	1.1725	0.5726	2.0680
LnUrbanization	1312	0.5291	0.1085	0.5527	0.2412	0.6931
LnOpenness	1312	0.5318	0.2431	0.4693	0.1724	1.2660

Considering the natural logarithmic transformation mitigates heteroscedasticity and cushions against violent fluctuations due to data scale changes, enhancing accuracy, logarithmic transformations of variables are applied in the model.

Empirical results and analysis

Benchmark regression

This study examined the impact of FDI on the energy mix of this economy based on Eq. (1), and the regression results are reported in Table 2. The results in column (1) reveal that the coefficient of FDI is negative at the 1% significance level without adding any control variables and fixed effects, which indicates that FDI reduces the share of renewable energy consumption in the host country. The results in columns (2) and (3), including country and year-fixed effects and country-level control variables, respectively, document that the pre-FDI coefficient is still significantly negative. The addition of both two-way fixed effects and control variables in column (4) shows that the average effect of FDI on energy structure is −0.015 and significant at the 1% level. The conclusions of this study are consistent with Khan et al. (2021), who suggest that FDI has not effectively promoted the transformation of the host country’s energy consumption structure. In terms of control variables, an increase in the degree of industrial upgrading will be conducive to an increase in the share of renewable energy consumption. However, as the urbanization process continues to advance, the proportion of fossil energy use in the economy is elevated. This may be because countries with higher levels of urbanization may pay more attention to economic and industrial development and not pay sufficient attention to environmental pollution, which in turn has a negative correlation with the structure of energy consumption (Wang et al. 2021; Islam et al. 2022). These results suggest that FDI mainly aims to avoid environmental regulations and seek cheap energy, which is not conducive to transitioning the energy consumption structure of host countries into a cleaner one.

Table 2. Baseline regressions results.

	(1)	(2)	(3)	(4)
VARIABLES	RNEW	RNEW	RNEW	RNEW
LnIFDI	−0.0191***	−0.0171***	−0.0141***	−0.0150***
	(0.0043)	(0.0054)	(0.0038)	(0.0046)
LnPerGDP			0.0266	−0.0392
			(0.0282)	(0.0281)
LnPercaptial			−0.0272	−0.0179
			(0.0227)	(0.0201)
LnStructure			0.1728***	0.0980**
			(0.0341)	(0.0466)
LnUrbanization			−0.6614***	−0.9408***
			(0.2453)	(0.3086)
LnOpenness			0.0800**	0.0290
			(0.0354)	(0.0511)
cons	0.2323***	0.2183***	0.2920*	1.0504***
	(0.0256)	(0.0126)	(0.1688)	(0.2566)
Year FE	No	Yes	No	Yes
Country FE	No	Yes	No	Yes
Adj. R2	0.0540	0.0911	0.2305	0.3390
Observation	1312	1312	1312	1312

* * *, * *, and * respectively represent the significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are in parentheses.

Robustness tests

Endogeneity test

Considering the reverse causality issue that higher fossil energy consumption may attract foreign direct investment, as well as the potential issue of omitted variables, we conduct the following two types of endogeneity tests. Firstly, we extend the lagged period of the dependent variable to the control variable to construct a dynamic model, in order to alleviate potential reverse causality problems. Secondly, we referred to Li et al. (2022) and used the FDI in the lagged period (denoted as IV_LAG) as an instrumental variable. The logic is that this directly affects the FDI in the current period but is not related to the stochastic perturbation term, which satisfies the basic requirements of instrumental variable relevance and exogeneity. The results of the endogeneity test based on the instrumental variable method are reported in Table 3. Columns (1) and (2) show that after adding the lagged period dependent variable, the coefficient before the core explanatory variable is still significantly negative, and the conclusion is consistent regardless of whether the control variable is included. Columns (3) show that the pre-coefficients of the instrumental variables IV_LAG are positive at the 1% level. The regression results of the second stage of the instrumental variable method are reported in columns (4). After considering the endogeneity issue, FDI has a significant negative impact on the renewable energy consumption structure of the host country, which is consistent with the results of the benchmark regression.

Table 3. Endogeneity test.

	(1)	(2)	(3)	(4)
VARIABLES	RNEW	RNEW	LnIFDI	RNEW
LnIFDI	−0.0018*	−0.0018*		−0.0487***
	(0.0010)	(0.0009)		(0.0081)
L.RNEW	0.9691***	0.9479***
	(0.0132)	(0.0175)
IV_LAG			0.3140***
			(0.0499)
cons	0.0063*	0.0324	3.7638***	1.3415***
	(0.0036)	(0.0542)	(1.2637)	(0.1178)
Control	No	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Country FE	Yes	Yes	Yes	Yes
Adj. R2	0.9017	0.9032	0.1860	0.9550
Observation	1228	1228	1228	1228

* * *, * *, and * respectively represent the significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are in parentheses.

Other robustness tests

In addition to considering endogeneity issues, we performed the following robustness tests. First, the explanatory variables were substituted. The benchmark regression reflects the energy structure effect of FDI in terms of the proportion of renewable energy consumption. We further substituted the host country’s energy consumption structure with carbon emissions, assuming that if FDI is conducive to transforming the host country’s energy structure into a cleaner one, its carbon emissions will be reduced, and vice versa. To accurately trace whether the increase in foreign demand causes the increase in the host country’s carbon emission, we adopted the ratio of total implied carbon of export trade to the domestic value-added of export (LnCIE).

Second, the explanatory variables were substituted. The previous section used the share of FDI in GDP as a measure. GDP is a flow concept, so the flow was changed to an inventory. Thus, we use the share of FDI in fixed assets as a substitute for the robustness test to reflect the amount of FDI received by the host country.

Third, anomalous years were excluded. Considering that the financial crisis in 2008 and the COVID-19 pandemic in 2020 had considerable impacts on international financial markets and capital flows (Wen et al. 2023), the impact of major events is circumvented by excluding the samples in 2008 and 2020.

Fourth, the lag period should be considered. Owing to the existence of time costs for using FDI in the host country and the need for a certain construction cycle for energy projects, these factors mean that FDI often cannot play a truly effective role in the current year. That is, there may be a time-lag effect on the impact of FDI on the energy consumption structure of the host country. Simultaneously, to alleviate the interference of other factors in the same period, we lagged all the explanatory variables by one and two periods.

Fifth, high-dimensional fixed effects were added. The impact of FDI on the energy consumption structure of the host country may be disturbed by the region in which the economy is located. To control for the overall trend of a specific region over time, we introduced time-region high-dimensional fixed effects to mitigate the potential impact.

According to the above methodology, the results in columns (1)–(6) of Table 4 are tested for robustness. The results reveal that after substituting the explanatory variables, excluding abnormal years, considering lagged effects, and adding high-dimensional fixed effects, the results are consistent with the conclusions of the baseline regression. There is an obvious lagged effect of FDI, and with the increase of the lagged period, the absolute value of the coefficients of the pre-FDI period gradually increases. These results suggest that the baseline regression results are robustly established.

Table 4. Robustness test.

	(1)	(2)	(3)	(4)	(5)	(6)
VARIABLES	LnCIE	RENW	RENW	F. RENW	F2. RENW	RENW
LnIFDI	0.0407*	−0.0113***	−0.0167***	−0.0160***	−0.0172***	−0.0097***
	(0.0218)	(0.0033)	(0.0047)	(0.0050)	(0.0052)	(0.0035)
cons	8.0582***	1.0190***	1.0015***	0.9278***	0.8066***	0.8881***
	(1.1731)	(0.2565)	(0.2440)	(0.2519)	(0.2461)	(0.2262)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Country FE	Yes	Yes	Yes	Yes	Yes	Yes
Year × area FE	No	No	No	No	No	Yes
Adj. R2	0.7685	0.3306	0.3500	0.3399	0.3431	0.5256
Observation	1195	1291	1194	1228	1162	1312

***, **, and * respectively represent the significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are in parentheses.

Mechanism test

Industry transfer effect

Based on considerations of costs and benefits, under the pressure of environmental regulations, FDI may transfer some industries to host countries to seek cheaper energy and labor, thereby hindering the transformation of the host country’s energy consumption structure. To examine whether FDI can affect the energy consumption structure of the host country through industrial transfer, we conducted a test based on Eqs. (2) and (3), and the regression results are reported in Table 5. Column (1) indicates that FDI can effectively transfer industries to the host country. Furthermore, the pre-coefficient of LnIRT in column (2) is significantly negative, which indicates that FDI will harm the renewable energy consumption structure of the host country by exerting the industry transfer effect. By dividing the industry transfer according to foreign demand, the effects of FDI on the transfer of demand for intermediate and final products are reported in columns (3) and (5), respectively. FDI has a significant promotion effect on both demand products. According to columns (4) and (6), adding the intermediate and final product demand shift has a significant negative effect on the host country’s renewable energy consumption structure when final product demand is shifted through FDI, while shifting the intermediate product demand does not have a significant effect on the energy structure. The test results indicate that FDI can negatively affect renewable energy consumption by transferring industries to the host country. The transfer of final product demand has a more significant negative effect on the host country’s renewable energy consumption structure than the transfer of intermediate product demand. The results are similar to those of Ma et al. (2023), but this study further examines the effects from the perspective of transferred industry types, clearly reflecting the heterogeneous impacts of different transferred industries. Accordingly, H1 is verified.

Table 5. Mechanism test: industrial transfer.

	(1)	(2)	(3)	(4)	(5)	(6)
VARIABLES	LnIRT	RENW	LnIRT_med	RENW	LnIRT_fin	RENW
LnIFDI	0.0382***	−0.0130***	0.0453**	−0.0150***	0.0383***	−0.0130***
	(0.0134)	(0.0039)	(0.0225)	(0.0047)	(0.0134)	(0.0039)
LnIRT		−0.0534*
		(0.0286)
LnIRT_med				0.0010
				(0.0161)
LnIRT_fin						−0.0535*
						(0.0287)
cons	3.3680***	1.2301***	−5.8639***	1.0564***	3.4054***	1.2327***
	(0.5655)	(0.2953)	(1.0099)	(0.2575)	(0.5652)	(0.2961)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Country FE	Yes	Yes	Yes	Yes	Yes	Yes
Adj. R2	0.9446	0.3586	0.8801	0.3385	0.9445	0.3588
Observation	1312	1312	1312	1312	1312	1312

***, **, and * respectively represent the significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are in parentheses.

Technology spillover effect

According to the theoretical analysis, the technology spillover effects of FDI on the host country’s renewable energy consumption structure can have either positive or negative uncertain impacts, which may be related to the different types of technology spillovers. Table 6 reports the mechanism tests based on technology spillovers to verify the impact of FDI’s technology spillover effects on the renewable energy consumption structure. The results in column (1) indicate that the increase in FDI will significantly improve the host country’s technological innovation capacity, manifested in the increase in the number of R&D patents. Similar to Li et al. (2024), this study also confirms that FDI has a significant technology spillover effect. The results of adding the number of patents are reported in column (2). As the number of patents in the host country increases, it will significantly reduce the proportion of renewable energy consumption, which is detrimental to the energy structure transformation of the host country. In addition, with reference to the OECD’s categorization, patents are categorized into environment-related (LnGreen) and non-environment-related technology patents (LnNonGreen) according to the international patent classification table for testing. The results in columns (3) and (5) document that FDI does not significantly promote environment-related green patents but effectively increases the number of applications for non-environment-related production patents. According to columns (4) and (6), after including the number of patent applications of the two types of patents, the pre-coefficient of non-environmentally related production patents is negative at the 1% level. This finding indicates that the technology spillover effect of FDI is mainly manifested in the increasing number of production patents of the host country, which contributes to the expansion of the scale of production and the increase of the consumption of fossil energy, and ultimately results in the host country’s failure to improve the structure of energy consumption effectively4. Accordingly, H2a is verified.

Table 6. Mechanism test: technology spillover.

	(1)	(2)	(3)	(4)	(5)	(6)
VARIABLES	LnPat	RNEW	LnGreen	RNEW	LnNonGreen	RNEW
LnIFDI	0.0810**	−0.0088***	−0.0061	−0.0085***	0.0806**	−0.0065***
	(0.0374)	(0.0027)	(0.0295)	(0.0030)	(0.0369)	(0.0023)
LnPat		−0.0288***
		(0.0084)
LnGreen				−0.0050
				(0.0048)
LnNonGreen						−0.0248***
						(0.0075)
cons	1.7569	1.1562***	−2.4591	0.8284***	2.9242	0.9132***
	(2.7326)	(0.2746)	(3.4406)	(0.2246)	(2.7420)	(0.2129)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Country FE	Yes	Yes	Yes	Yes	Yes	Yes
Adj. R2	0.2473	0.4526	0.5472	0.3925	0.2119	0.4490
Observation	1213	1213	1162	1162	1162	1162

***, **, and * respectively represent the significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are in parentheses.

Heterogeneity test

Heterogeneity test based on income level and OECD countries

Referring to Shahbaz et al. (2022), we analyzed the samples by dividing them into high- and lower-middle-income countries based on income level according to the World Bank’s classification criteria. The results of heterogeneity analysis based on income level are reported in columns (1) and (2) of Table 7. The results show that FDI has a significant negative impact on the renewable energy consumption structure in both high- and lower-middle-income countries. However, the negative impact of FDI on the renewable energy consumption structure in lower-middle-income countries is higher relative to high-income countries; the difference between the two holds at the 1% level. Lower environmental regulations and cheap resource factors lead to lower-middle-income countries becoming the main target of FDI to transfer industries, and the large use of fossil energy will be detrimental to the transformation of the energy consumption structure of these countries. Moreover, owing to the constraints of the stage of economic development, these countries have economic growth as their main goal and are willing to harm the environment to obtain rapid economic growth, which further leads to a decline in the proportion of their renewable energy consumption.

Table 7. Heterogeneity test: income levels and OECD countries.

	(1)	(2)	(3)	(4)
VARIABLES	High income	Low income	OECD	Non-OECD
LnIFDI	−0.0052**	−0.0407***	−0.0040	−0.0311***
	(0.0025)	(0.0110)	(0.0027)	(0.0084)
cons	0.4690	0.7677***	0.9213***	0.8045***
	(0.3134)	(0.2603)	(0.2675)	(0.2467)
Control	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Country FE	Yes	Yes	Yes	Yes
Difference	P-value = 0.000***		P-value = 0.000***
Adj. R2	0.5745	0.4384	0.5605	0.4133
Observation	844	468	744	568

***, **, and * respectively represent the significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are in parentheses.

Compared to non-OECD countries, OECD countries have a more complete energy transition system and a higher level of production technology. They are at the forefront of building environmentally friendly countries and sustainable development. These countries play an important role in global energy conservation, emissions reduction, and energy transition (Wei et al. 2023). We divided the samples into OECD and non-OECD countries, and the regression results are presented in columns (3) and (4) of Table 7. The results reveal that FDI significantly negatively impacts the renewable energy consumption structure in non-OECD countries while failing the significance test for OECD countries. Further examination of the difference in coefficients between groups reveals that FDI has a greater negative impact on non-OECD countries’ renewable energy consumption structure, and the difference between the two passes the 1% significance test. OECD countries not only have better environmental performance in economic development but also mitigate the negative impact of FDI on their energy consumption structure through various forms of environmental protection policies to transition their energy consumption structure to renewable energy.

Heterogeneity test based on industrial dominance

Economies with different industrial advantages are affected differently by FDI. Those dominated by resource- and technology-intensive industries usually need to consume more basic energy to gain development advantages, and their energy consumption structure is more affected by FDI. However, economies dominated by service industries usually have higher technological levels and stronger environmental protection policies. They can recognize the environmental performance of FDI better and resolve its negative impact on their energy consumption structure. Based on this, we divided the samples into two groups of high and low levels of industrial development according to the proportion of GDP accounted for by agriculture, manufacturing, and service industries.

The results of heterogeneity tests based on the level of agricultural development are reported in columns (1) and (2) of Table 8. The results indicate that the negative impact of FDI is stronger in regions with higher agricultural development relative to those with low agricultural development, and their between-group differences are significant at the 1% level. Columns (3) and (4) report the heterogeneity test based on the level of manufacturing development, and the results reveal that the higher the level of manufacturing development, the stronger the negative impact of FDI on its energy consumption structure. Columns (5) and (6) report the test based on the level of development of the service sector. The results show that FDI exhibits a stronger negative impact in regions with a lower level of development of the service sector relative to those with a high level, which significantly reduces the proportion of their renewable energy consumption.

Table 8. Heterogeneity test: different industries.

	(1)	(2)	(3)	(4)	(5)	(6)
	Agriculture industry		Manufacturing industry		Service industry
VARIABLES	High	Low	High	Low	High	Low
LnIFDI	−0.0309***	−0.0073**	−0.0246***	−0.0067*	−0.0084**	−0.0273***
	(0.0091)	(0.0032)	(0.0089)	(0.0038)	(0.0035)	(0.0082)
cons	0.7457***	0.9841**	1.2010***	0.9688***	1.0364***	0.9396***
	(0.2574)	(0.4021)	(0.3573)	(0.3178)	(0.3073)	(0.2630)
Control	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes
Country FE	Yes	Yes	Yes	Yes	Yes	Yes
Difference	P-value = 0.007***		P-value = 0.025**		P-value = 0.024**
Adj. R2	0.3361	0.5667	0.2900	0.4654	0.5632	0.3128
Observation	651	661	669	643	651	661

***, **, and * respectively represent the significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are in parentheses.

Heterogeneity tests based on different periods of time

To test the impact of the Kyoto Protocol in 2005 and its official entry into force in the Paris Agreement in 2016 on the energy consumption structure of host countries, we divided the samples into three segments, namely, before 2005, 2005–2016, and after 2016, for the time-period regression test. The results are presented in Table 9. Columns (1) and (5) report the effects of FDI on the host country’s energy consumption efficiency before 2005 and after 2016, respectively. The results document that the FDI coefficient is positive but fails the significance test. Results for 2005–2016 are reported in column (2), where the FDI coefficient is negative at the 1% level for samples that fall within this period. Further distinguishing between different income countries, the results in columns (3) and (4) demonstrate that FDI has a greater negative impact on the renewable energy consumption structure in low-income countries relative to high-income countries. The empirical P-value is significant at the 1% level. This indicates that the formal entry of the Kyoto Protocol into force has brought greater environmental pressure on developed countries, which are incentivized to use FDI to transfer industries and seek cheap energy and labor factors in developing countries to improve their own environmental performance and maintain international competitiveness; in turn, this negatively affects the structure of renewable energy consumption in developing countries and has a “pollution haven” effect. The Paris Agreement stipulates that all countries must take action to reduce GHG and emphasizes inter-country cooperation and shared responsibility (Chen et al. 2023). The Agreement is more comprehensive and forward-looking than the Kyoto Protocol and can be more effective in realizing sustainable economic development and promoting the transition of the energy consumption structure of all countries to renewable energy.

Table 9. Heterogeneity test: different periods.

	(1)	(2)	(3)	(4)	(5)
VARIABLES	Before 2005	2005–2016	High income	Low income	After 2016
LnIFDI	0.0011	−0.0105***	−0.0049***	−0.0289***	0.0037
	(0.0016)	(0.0033)	(0.0015)	(0.0096)	(0.0035)
cons	0.1881	1.2947***	0.7802**	0.6964***	1.4230*
	(0.3328)	(0.2438)	(0.3549)	(0.2242)	(0.7399)
Control	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes
Country FE	Yes	Yes	Yes	Yes	Yes
Difference			P-value = 0.000***
Adj. R2	0.0263	0.4594	0.6366	0.4492	0.2940
Observation	312	702	453	249	238

***, **, and * respectively represent the significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are in parentheses.

Further analysis: role of governance

The introduction of relevant government policies will affect the development and structural adjustment of the energy industry. To further test whether governance can inhibit the negative impact of FDI on renewable energy consumption, we constructed the following model to test the regulatory effect of governance.

6

$\begin{array}{c}{RNEW}_{it}={\theta }_0+{\theta }_1{M}_{it}\times Ln{IFDI}_{it}+{\theta }_2{M}_{it}\hfill \\ +{\theta }_{3}Ln{IFDI}_{it}+{\theta }_4{CONT}_{it}+{\mu }_i+{\phi }_t+{\epsilon }_{it}\hfill \end{array}$where, $M_{it}$ is the moderating variable, which contains the comprehensive indicator of governance and its six sub-dimensions, and the other variables are as previously defined. The coefficient ${\theta }_1$ in Eq. (6) is the focus in this study. If ${\theta }_1>0$, FDI positively affects the energy consumption structure of the host country as governance improves.

We referred to Wei et al. (2023) and used the World Governance Indicator (WGI) as a proxy variable for governance5, which categorizes the quality of national governance into six dimensions, including control of corruption (CC), government effectiveness (GE), political stability (PV), regulatory quality (RQ), rule of law (RL), and voice and accountability (VA). Governance (Gov) is calculated using PCA technology and the larger the indicator, the stronger the government’s governance capacity.

This study conducted a test based on Eq. (6), and the regression results are presented in columns (1) and (2) of Panel A in Table 10. The results document that the pre-coefficient of the interaction term between Gov and FDI is significantly positive, which indicates that the improvement of the host country’s governance capacity can effectively inhibit the negative impact of FDI on its energy consumption structure and increase the proportion of renewable energy consumption. The results of this study are similar to those of Dossou et al. (2023), finding that the interaction between FDI and governance quality can effectively improve the energy structure. Furthermore, these six dimensions are subjected to interaction tests separately, and the regression results based on the sub-indicators of governance are reported in columns (3) and (4) of Panel A and Panel B in Table 10. The interaction terms between the above six sub-indicators and FDI are all significantly positive. To further reflect the marginal effect of the coefficients before the interaction terms, we plotted the coefficients for each sub-dimension in Fig. 1. It can be observed that the interaction coefficient of RQ is the highest, indicating that regulatory quality plays a dominant role in improving the host country’s energy consumption structure.

Table 10. Moderating effect: governance.

Panel A
	(1)	(2)	(3)	(4)
VARIABLES	M = Gov	M = Gov	M = CC	M = GE
M × LnIFDI	0.0054*	0.0057**	0.0101**	0.0122**
	(0.0029)	(0.0025)	(0.0048)	(0.0053)
M	−0.0457***	−0.0268**	−0.0146	−0.0491**
	(0.0138)	(0.0106)	(0.0170)	(0.0189)
LnIFDI	−0.0191***	−0.0177***	−0.0240***	−0.0265***
	(0.0056)	(0.0052)	(0.0078)	(0.0084)
cons	0.2178***	1.1212***	1.1665***	1.1097***
	(0.0114)	(0.2384)	(0.2518)	(0.2413)
Control	No	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Country FE	Yes	Yes	Yes	Yes
Adj. R2	0.1731	0.3799	0.3667	0.3787
Observation	1250	1250	1250	1250
Panel B
	(1)	(2)	(3)	(4)
VARIABLES	M = PV	M = RL	M = RQ	M = VA
M×LnIFDI	0.0094**	0.0112**	0.0161**	0.0103*
	(0.0039)	(0.0055)	(0.0068)	(0.0058)
M	−0.0376***	−0.0359	−0.0244	−0.0437***
	(0.0141)	(0.0219)	(0.0169)	(0.0147)
LnIFDI	−0.0194***	−0.0253***	−0.0310***	−0.0218***
	(0.0056)	(0.0085)	(0.0101)	(0.0079)
cons	1.1425***	1.1620***	1.1888***	1.2065***
	(0.2537)	(0.2475)	(0.2502)	(0.2498)
Control	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
Country FE	Yes	Yes	Yes	Yes
Adj. R2	0.3747	0.3682	0.3723	0.3766
Observation	1250	1250	1250	1250

***, **, and * respectively represent the significance at the 1%, 5%, and 10% levels, respectively. Robust standard errors are in parentheses.

Fig. 1 [Images not available. See PDF.]

The margin effect of the interaction term.

Blue circles represent the point estimates of the coefficients, while blue lines indicate the corresponding 90% confidence intervals, and the red line shows the position of the zero value.

According to the definition of the sub-indicator, RQ reflects the government’s ability to formulate and implement policies and regulations to promote the development of the private sector, institutionally ensure that the formation of a better business environment can attract high-quality foreign firms, promote the entry of foreign firms into the host country in the form of joint ventures or holdings, help local firms to fully learn and absorb the advanced business concepts and management modes of foreign firms, drive firms toward high-end manufacturing, developing the service industry, undertaking the green transformation of production, and finally transitioning the energy consumption structure to renewable energy. Therefore, H3 is verified.

Conclusions and discussion

Conclusions

This study examined the impact of FDI on the energy consumption structure using a cross-country panel data model for 65 countries and regions from 2000 to 2020 to accurately identify the effect of FDI on energy structure and improve the policy system for renewable energy consumption. First, FDI significantly negatively impacts host countries’ renewable energy consumption structure, and the conclusion still holds after a series of robustness and endogeneity tests. Second, FDI negatively affects the renewable energy consumption structure of host countries through the industry transfer effect and the technology spillover effect. Third, the impact of FDI is affected by the income level, whether the host country is an OECD country, the level of industrial development, and international agreements. Compared to high-income countries, the relatively low level of environmental regulations in lower-middle-income countries contributes to the large use of cheap fossil energy in the host country by FDI, which is not conducive to transitioning its energy consumption structure into a cleaner one. Owing to the difference between OECD and non-OECD countries in terms of energy technology and development, FDI negatively impacts the non-OECD countries’ renewable energy consumption structure more than that of OECD countries. FDI has a greater negative impact on the renewable energy consumption structure of economies with a high level of agricultural development, a high level of manufacturing development, and a low level of service development. Compared with the Paris Agreement, the signing of the Kyoto Protocol has intensified the significant negative impact of FDI on the proportion of renewable energy consumption in host countries through industrial transfer. Fourth, improving governance capacity can effectively mitigate the negative impact of FDI on the renewable energy consumption structure.

Limitations and future directions

This study has some limitations that provide directions for future research: First, this study primarily examines the impact of FDI on energy structure from a macroeconomic perspective, lacking research based on the perspective of microeconomic agents. Future studies could focus on multinational corporations to investigate the impact of foreign direct investment by microeconomic agents on energy consumption structure, further refining the research on the energy transition effects of FDI. Second, this study mainly examines the impact of FDI on the energy consumption structure of host countries, without considering the impact of FDI on the energy consumption structure of home countries. Future research could explore the reverse spillover effects of FDI, examining how FDI influences the energy consumption structure of home countries, thereby further revealing the transmission mechanisms of FDI on energy consumption structure. Third, this study primarily examines the overall impact of FDI on the energy consumption structure of host countries. In fact, FDI can take various forms, such as greenfield investments and cross-border mergers and acquisitions. Whether different forms of FDI have differential impacts on the energy consumption structure of host countries is another key area for future investigation.

Policy implications

This study offers the following policy implications: First, government departments should enhance public governance capabilities, improve administrative efficiency, and create a fair and transparent business environment. The government should actively attract high-quality FDI with green attributes by offering tax incentives, fiscal subsidies, and other policy measures to encourage foreign enterprises that excel in environmental protection and sustainable development. This will promote domestic economic development, spur local enterprises to innovate and advance in environmental protection technologies, and ultimately achieve a win-win situation for the economy and the environment. Second, fully utilize the technology spillover and economic growth effects brought by FDI to promote the flow of energy usage towards high-value-added industries. On this basis, develop stable and reasonable renewable energy plans tailored to the country’s economic development situation, accelerate the development of the renewable energy industry, and reduce dependence on traditional energy sources. Third, countries should strengthen international cooperation, make full use of the platform provided by the Paris Agreement, accelerate the construction of renewable energy infrastructure, cooperate in renewable energy R&D and innovation, actively respond to global climate change issues, and help the global economy to achieve a green transformation while pursuing high-quality development.

Acknowledgements

This research was supported by the National Social Science Fund of China [No. 22AJY021].

Author contributions

Xia Fang: methodology, data curation, formal analysis, writing-reviewing and editing; Zhenyu Yang: conceptualization, data curation, writing-original draft, writing-reviewing and editing; Yun Zhang: supervision, resources, validation, writing-review and editing; Xiao Miao: methodology, formal analysis, writing-reviewing and editing.

Data availability

The data used in this article has been submitted and the data file can be obtained online.

Competing interests

The authors declare no competing interests.

Ethical approval

Ethical approval is not applicable because this article does not contain any studies with human or animal subjects.

Informed consent

Informed consent is not applicable because this article does not contain any studies with human or animal subjects.

1

These 65 sample countries (regions) include: Argentina, Australia, Austria, Belgium, Brazil, Brunei Darussalam, Bulgaria, Cambodia, Canada, Chile, China P.R.: Hong Kong, China P.R.: Mainland, Colombia, Costa Rica, Croatia Rep. of, Cyprus, Czech Rep., Denmark, Estonia Rep. of, Finland, France, Germany, Greece, Hungary, Iceland, India, Indonesia, Ireland, Israel, Italy, Japan, Kazakhstan Rep. of, Korea Rep. of, Lao People’s Dem. Rep., Latvia, Lithuania, Luxembourg, Malaysia, Malta, Mexico, Morocco, Myanmar, Netherlands, New Zealand, Norway, Peru, Philippines, Poland Rep. of, Portugal, Romania, Russian Federation, Saudi Arabia, Singapore, Slovak Rep., Slovenia, Rep. of, South Africa, Spain, Sweden, Switzerland, Thailand, Tunisia, Turkey, United Kingdom, United States, Vietnam. The research sample in this study has a wide geographical distribution, including developed and emerging economies, and the sample period lasts for 21 years, which is representative.

2

Our results are robust when we without winsorization and alternatively winsorize at 1 and 99% levels and at 5 and 95% levels.

3

WDI database: https://data.worldbank.org. UNCTAD database: https://unctad.org/statistics. IEA database: https://www.iea.org/data-and-statistics. OECD database: https://www.oecd.org/en/data.html.

4

Theoretically, middle- and low- income countries tend to have relatively lower levels of production technology and environmental regulations. Thus, FDI may be incentivized to foster non-environmental technological innovation (production process innovation) in these countries to leverage their lower-cost fossil fuels for expanded production. To rigorously examine the robustness of our argument, we conduct a subgroup analysis based on income levels, revealing that this effect holds exclusively in the sample of middle- and low-income countries. We appreciate the valuable comments from the reviewers.

5

WGI come from the World Bank database: https://www.worldbank.org/en/publication/worldwide-governance-indicators.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1057/s41599-024-04280-y.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

Ahmad, M; Khattak, SI; Khan, A; Rahman, ZU. Innovation, foreign direct investment (FDI), and the energy-pollution–growth nexus in OECD region: a simultaneous equation modeling approach. Environ Ecol Stat; 2020; 27, pp. 203-232. [DOI: https://dx.doi.org/10.1007/s10651-020-00442-8] 1185.65191

Aitken, BJ; Harrison, AE. Do domestic firms benefit from direct foreign investment? evidence from Venezuela. Am Econ Rev; 1999; 89, pp. 605-618. [DOI: https://dx.doi.org/10.1257/aer.89.3.605]

Akpanke, TA; Deka, A; Ozdeser, H; Seraj, M. Does foreign direct investment promote renewable energy use? an insight from West African countries. Renew Energy Focus; 2023; 44, pp. 124-131. [DOI: https://dx.doi.org/10.1016/j.ref.2022.11.007] 1537.11007

Alharbi, SS; Al Mamun, M; Boubaker, S; Rizvi, SKA. Green finance and renewable energy: a worldwide evidence. Energy Econ; 2023; 118, [DOI: https://dx.doi.org/10.1016/j.eneco.2022.106499] 106499.

Ali, N; Phoungthong, K; Khan, A; Abbas, S; Dilanchiev, A; Tariq, S; Sadiq, MN. Does FDI foster technological innovations? empirical evidence from BRICS economies. PLOS ONE; 2023; 18, e0282498. [DOI: https://dx.doi.org/10.1371/journal.pone.0282498] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36893169][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9997950]

Alsagr, N. Financial efficiency and its impact on renewable energy investment: empirical evidence from advanced and emerging economies. J Clean Prod; 2023; 401, [DOI: https://dx.doi.org/10.1016/j.jclepro.2023.136738] 1537.91312136738.

Arkolakis, C; Ramondo, N; Rodríguez-Clare, A; Yeaple, S. Innovation and production in the global economy. Am Econ Rev; 2018; 108, pp. 2128-2173. [DOI: https://dx.doi.org/10.1257/aer.20141743] 1409.91170

Behera, P; Sethi, N. Nexus between environment regulation, FDI, and green technology innovation in OECD countries. Environ Sci Pollut Res Int; 2022; 29, pp. 52940-52953. [DOI: https://dx.doi.org/10.1007/s11356-022-19458-7] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35275367]

Blonigen, BA; Davies, RB; Waddell, GR; Naughton, HT. FDI in space: spatial autoregressive relationships in foreign direct investment. Eur Econ Rev; 2007; 51, 5 pp. 1303-1325. [DOI: https://dx.doi.org/10.1016/j.euroecorev.2006.08.006]

Burstein, AT; Monge-Naranjo, A. Foreign know-how, firm control, and the income of developing countries. Q J Econ; 2009; 124, pp. 149-195. [DOI: https://dx.doi.org/10.1162/qjec.2009.124.1.149] 1186.91165

Carluccio, J; Fally, T. Foreign entry and spillovers with technological incompatibilities in the supply chain. J Int Econ; 2013; 90, pp. 123-135. [DOI: https://dx.doi.org/10.1016/j.jinteco.2012.08.004] 1295.65024

Chen, L; Wen, F; Zhang, Y; Miao, X. Oil supply expectations and corporate social responsibility. Int Rev Financ Anal; 2023; 87, [DOI: https://dx.doi.org/10.1016/j.irfa.2023.102638] 1513.92004102638.

Cheng, Z; Li, L; Liu, J. The impact of foreign direct investment on urban PM2.5 pollution in China. J Environ Manag; 2020; 265, [DOI: https://dx.doi.org/10.1016/j.jenvman.2020.110532] 1544.32038110532.

Copeland, BR; Taylor, MS. North-South trade and the environment. Q J Econ; 1994; 109, pp. 755-787. [DOI: https://dx.doi.org/10.2307/2118421] 0813.90012

Demena, BA; Afesorgbor, SK. The effect of FDI on environmental emissions: evidence from a meta-analysis. Energy Policy; 2020; 138, [DOI: https://dx.doi.org/10.1016/j.enpol.2019.111192] 1520.93428111192.

Dong, W; Li, Y; Gao, P; Sun, Y. Role of trade and green bond market in renewable energy deployment in Southeast Asia. Renew Energy; 2023; 204, pp. 313-319. [DOI: https://dx.doi.org/10.1016/j.renene.2023.01.022] 1113.76333

Dossou, TAM; Kambaye, EN; Asongu, SA; Alinsato, AS; Berhe, MW; Dossou, KP. Foreign direct investment and renewable energy development in sub-saharan Africa: does governance quality matter?. Renew Energy; 2023; 219, [DOI: https://dx.doi.org/10.1016/j.renene.2023.119403] 119403.

Guo, W; Yang, B; Ji, J; Liu, X. Green finance development drives renewable energy development: mechanism analysis and empirical research. Renew Energy; 2023; 215, [DOI: https://dx.doi.org/10.1016/j.renene.2023.118982] 1533.93777118982.

Hausman, JA; Taylor, WE. Panel data and unobservable individual effects. Econometrica; 1981; 49, 6 pp. 1377-1398.636158[DOI: https://dx.doi.org/10.2307/1911406] 0464.90013

He, L; Zhang, L; Zhong, Z; Wang, D; Wang, F. Green credit, renewable energy investment and green economy development: empirical analysis based on 150 listed companies of China. J Clean Prod; 2019; 208, pp. 363-372. [DOI: https://dx.doi.org/10.1016/j.jclepro.2018.10.119] 1454.74054

Hille, E; Shahbaz, M; Moosa, I. The impact of FDI on regional air pollution in the Republic of Korea: a way ahead to achieve the green growth strategy?. Energy Econ; 2019; 81, pp. 308-326. [DOI: https://dx.doi.org/10.1016/j.eneco.2019.04.004] 0941.93019

Huynh, CM; Hoang, HH. Foreign direct investment and air pollution in Asian countries: does institutional quality matter?. Appl Econ Lett; 2019; 26, pp. 1388-1392. [DOI: https://dx.doi.org/10.1080/13504851.2018.1563668] 1482.91045

Islam, MM; Irfan, M; Shahbaz, M; Vo, XY. Renewable and non-renewable energy consumption in Bangladesh: the relative influencing profiles of economic factors, urbanization, physical infrastructure and institutional quality. Renew Energ; 2022; 184, pp. 1130-1149. [DOI: https://dx.doi.org/10.1016/j.renene.2021.12.020]

Javorcik, BS; Lo Turco, A; Maggioni, D. New and improved: does FDI boost production complexity in host countries?. Econ J; 2018; 128, pp. 2507-2537. [DOI: https://dx.doi.org/10.1111/ecoj.12530]

Khan, A; Chenggang, Y; Hussain, J; Kui, Z. Impact of technological innovation, financial development and foreign direct investment on renewable energy, non-renewable energy and the environment in Belt & Road initiative countries. Renew Energy; 2021; 171, pp. 479-491. [DOI: https://dx.doi.org/10.1016/j.renene.2021.02.075]

Khan, H; Weili, L; Khan, I; Zhang, J. The nexus between natural resources, renewable energy consumption, economic growth, and carbon dioxide emission in BRI countries. Environ Sci Pollut Res Int; 2023; 30, pp. 36692-36709. [DOI: https://dx.doi.org/10.1007/s11356-022-24193-0] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36562975]07738283

Kim, MH; Adilov, N. The lesser of two evils: an empirical investigation of foreign direct investment-pollution tradeoff. Appl Econ; 2012; 44, pp. 2597-2606. [DOI: https://dx.doi.org/10.1080/00036846.2011.566187] 1347.11051

Kirkpatrick, C; Shimamoto, K. The effect of environmental regulation on the locational choice of Japanese foreign direct investment. Appl Econ; 2008; 40, pp. 1399-1409. [DOI: https://dx.doi.org/10.1080/00036840600794330] 0395.05052

Koopman R, Powers W, Wang Z, Wei S (2010) Give credit where credit is due: tracing value added in global production chains. Natl Bureau Econ Res https://doi.org/10.3386/w16426

Li, F; Zhang, J; Li, X. Research on supporting developing countries to achieve green development transition: based on the perspective of renewable energy and foreign direct investment. J Clean Prod; 2022; 372, [DOI: https://dx.doi.org/10.1016/j.jclepro.2022.133726] 133726.

Li, X; Chen, X; Hou, K. FDI technology spillovers in Chinese supplier-customer networks. Int Rev Financ Anal; 2024; 94, [DOI: https://dx.doi.org/10.1016/j.irfa.2024.103285] 07881960103285.

Lu, W; Ikboljon, K; Hayot, BS. Foreign direct investment and renewable energy: examining the environmental Kuznets curve in resource-rich transition economies. Renew Energy; 2023; 208, pp. 301-310. [DOI: https://dx.doi.org/10.1016/j.renene.2023.03.054]

Ma, N; Sun, W; Wang, Z; Li, H; Ma, X; Sun, H. The effects of different forms of FDI on the carbon emissions of multinational enterprises: a complex network approach. Energy Policy; 2023; 181, [DOI: https://dx.doi.org/10.1016/j.enpol.2023.113731] 1474.05081113731.

Muhammad, B; Khan, MK; Khan, MI; Khan, S. Impact of foreign direct investment, natural resources, renewable energy consumption, and economic growth on environmental degradation: evidence from BRICS, developing, developed and global countries. Environ Sci Pollut Res Int; 2021; 28, pp. 21789-21798. [DOI: https://dx.doi.org/10.1007/s11356-020-12084-1] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33411280]1482.30038

Nawaz, A; Rahman, MM. Global value chains participation and environmental degradation in SAARC economies. Environ Dev Sustain; 2024; 26, 6 pp. 15595-15617. [DOI: https://dx.doi.org/10.1007/s10668-023-03265-2]

Opoku, EEO; Boachie, MK. The environmental impact of industrialization and foreign direct investment. Energy Policy; 2020; 137, [DOI: https://dx.doi.org/10.1016/j.enpol.2019.111178] 1457.92176111178.

Phung, TQ; Rasoulinezhad, E; Thu, HLT. How are FDI and green recovery related in southeast Asian economies?. Econ Change Restruct; 2022; 56, 6 pp. 3735-3755. [DOI: https://dx.doi.org/10.1007/s10644-022-09398-0] 1522.92068

Shahbaz, M; Wang, J; Dong, K; Zhao, J. The impact of digital economy on energy transition across the globe: the mediating role of government governance. Renew Sustain Energ Rev; 2022; 166, [DOI: https://dx.doi.org/10.1016/j.rser.2022.112620] 112620.

Su, Z; Li, J; Wei, S; Wu, Z. Effect of inward foreign direct investment on entrepreneurship productivity in emerging markets. J Int Manag; 2024; 30, 1 [DOI: https://dx.doi.org/10.1016/j.intman.2023.101116] 1374.78180101116.

Sun, H; Chen, F. The impact of green finance on China’s regional energy consumption structure based on system GMM. Resour Policy; 2022; 76, [DOI: https://dx.doi.org/10.1016/j.resourpol.2022.102588] 102588.

Tan, Y; Uprasen, U. The effect of foreign direct investment on renewable energy consumption subject to the moderating effect of environmental regulation: evidence from the BRICS countries. Renew Energy; 2022; 201, pp. 135-149. [DOI: https://dx.doi.org/10.1016/j.renene.2022.11.066]

Tang, X; Zhou, X. Impact of green finance on renewable energy development: a spatiotemporal consistency perspective. Renew Energy; 2023; 204, pp. 320-337. [DOI: https://dx.doi.org/10.1016/j.renene.2023.01.012] 1418.90099

Wang, Q; Fan, Z. Green finance and investment behavior of renewable energy enterprises: a case study of China. Int Rev Financ Anal; 2023; 87, [DOI: https://dx.doi.org/10.1016/j.irfa.2023.102564] 1475.91273102564.

Wang, S; Sun, L; Iqbal, S. Green financing role on renewable energy dependence and energy transition in E7 economies. Renew Energy; 2022; 200, pp. 1561-1572. [DOI: https://dx.doi.org/10.1016/j.renene.2022.10.067] 1275.74018

Wang, S; Wang, H; Sun, Q. The impact of foreign direct investment on environmental pollution in China: corruption matters. Int J Environ Res Public Health; 2020; 17, 6477. [DOI: https://dx.doi.org/10.3390/ijerph17186477] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32899562][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7559999]1451.93245

Wang, WZ; Liu, LC; Liao, H; Wei, YM. Impacts of urbanization on carbon emissions: an empirical analysis from OECD countries. Energ Policy; 2021; 151, [DOI: https://dx.doi.org/10.1016/j.enpol.2021.112171] 1456.51004112171.

Wei, J; Wen, J; Wang, XY; Ma, J; Chang, C. Green innovation, natural extreme events, and energy transition: evidence from Asia-Pacific economies. Energy Econ; 2023; 121, [DOI: https://dx.doi.org/10.1016/j.eneco.2023.106638] 106638.

Wen, F; Chen, M; Zhang, Y; Miao, X. Oil price uncertainty and audit fees: evidence from the energy industry. Energy Econ; 2023; 125, [DOI: https://dx.doi.org/10.1016/j.eneco.2023.106852] 1513.92004106852.

Xing, Y; Kolstad, CD. Do lax environmental regulations attract foreign investment?. Environ Resour Econ; 2002; 21, pp. 1-22. [DOI: https://dx.doi.org/10.1023/A:1014537013353] 0909.90106

Yang, QC; Zheng, M; Chang, CP. Energy policy and green innovation: a quantile investigation into renewable energy. Renew Energy; 2022; 189, pp. 1166-1175. [DOI: https://dx.doi.org/10.1016/j.renene.2022.03.046] 1212.35039

Yu, G; Liu, K. Foreign direct investment, environmental regulation, and urban green development efficiency-an empirical study from China. Appl Econ; 2024; 56, 23 pp. 2738-2751. [DOI: https://dx.doi.org/10.1080/00036846.2023.2200232] 1516.37117

Zhang, S; Xie, G. Promoting green investment for renewable energy sources in China: case study from autoregressive distributed Lagged in error correction approach. Renew Energy; 2023; 214, pp. 359-368. [DOI: https://dx.doi.org/10.1016/j.renene.2023.05.131] 1509.65024

Zhang, Y; Fang, X; Yang, Z; Sun, Y; Wang, Q. Green finance, technological innovation, and energy efficiency improvements: evidence from China’s green finance reform pilot zone. Emerg Mark Financ Trade; 2023; 59, 11 pp. 3531-3549. [DOI: https://dx.doi.org/10.1080/1540496X.2023.2226324] 1519.70009