How Improving the Quality of Foreign Direct Investment Can Promote Sustainable Development: Evidence from China

Abstract

Sustainable development is an inevitable derivative outcome of the advancement of social productive forces and the innovation of science and technology. In the current era, a multitude of global issues are intertwined. Sustainable development provides ideas and approaches of crucial value for resolving these difficult situations. This study constructs a micro-level indicator system to assess the quality of foreign direct investment and measures the quality of FDI in China from 2011 to 2022. Using the two-way fixed effects panel model, this study empirically tests the impact of FDI quality on China’s sustainable development and deeply examines the industry heterogeneity. The findings reveal that (1) micro-level FDI quality indicators avoid aggregation bias and lagged responses inherent in macro-level analyses, enabling precise and timely detection of foreign firms’ reactions to macroeconomic shifts. (2) Enhancing FDI quality exerts a positive and significant effect on China’s sustainable development, with notable variations across industries. (3) Further analysis shows that, first, in eastern coastal provinces, well-functioning market mechanisms amplify the positive externalities of high-quality FDI on resource allocation. Second, the moderating role of intellectual property protection in FDI’s human capital effects exhibits significant heterogeneity across industries.

1. Introduction

Peace and development constitute the shared aspiration of humanity worldwide. In 2015, the United Nations adopted Transforming Our World: The 2030 Agenda for Sustainable Development (the 2030 Agenda). Starting from the five Ps of “People, Planet, Prosperity, Peace, and Partnership”, the 2030 Agenda covers three dimensions: social, economic, and environmental. The concept of sustainable development advocated therein and its strategies for transformation are of great significance. Taking a comprehensive view of the current international situation, the smoke of geopolitical conflicts and the undercurrents of global economic fluctuations converge in a turbulent manner. The international economic and political landscape has become increasingly complex and volatile. The development of human society is confronted with a great deal of instability and uncertainty, and sustainable development is regarded as the key to resolving the impasse and social development. At the same time, considering China’s domestic situation, China is also faced with a series of new challenges and problems. China’s proposed “high-quality development”, as a development requirement that incorporates the concept of sustainable development and reflects the New Development Concepts, is the overall requirement put forward by Chinese policymakers for all aspects of economic and social development based on China’s development stage.

For a long time, foreign direct investment (FDI) has been regarded as an important tool for promoting economic prosperity and development [1]. For Chinese policymakers, in the new development stage, China also needs FDI to continue to play its optimal role, to deeply participate in the process of China’s high-quality development, and to assist China in building a modern, powerful country and fostering new forms of productive forces. More importantly, as China’s economy shifts from a stage of high-speed growth to a stage of high-quality development, China’s key focuses in attracting and utilizing FDI have also changed accordingly. In the early days, China blindly pursued the pace of economic growth. When introducing FDI, it emphasized the scale and volume, which consequently led to a massive influx of some FDI projects with high energy consumption [2]. Nowadays, with the emphasis on implementing the new development concepts featuring innovation, coordination, green development, openness, and shared benefits, and the requirement to accelerate the construction of a new development pattern, China’s policy on attracting FDI is gradually upgrading and transforming towards the direction of “expanding the incremental volume, stabilizing the existing volume, and improving the quality”. Even though the incremental volume and the stock of FDI still remain the fundamental aspects of current policy concerns, the quality of FDI will undoubtedly become increasingly important. Therefore, how to more accurately identify high-quality FDI enterprises, so as to introduce and utilize the capital, knowledge, technology, and concepts they possess to better serve China’s high-quality development goals and further promote China’s practice of the concept of sustainable development on the international stage has increasingly become a key issue that requires attention at present.

The exploration of the quality of FDI and its impact on the economy of the host country can be traced back to Luiz and de Mello Jr at the earliest. When they studied the impact of FDI on economic growth, they pointed out that FDI is a comprehensive combination that includes capital stock, technological progress, and management experience, and believed that its impact on economic growth is multifaceted [3]. Javorcik et al. held the view that because of the differences in the source countries of FDI, the extent of the vertical spillover effects of FDI on the economic growth of the host country would also exhibit variations [4]. In fact, whether distinguishing based on the differences in the inherent resources of FDI or differentiating according to its source countries, defining FDI from the dimension of quality remains a key approach. Over the past few decades, scholars have never stopped exploring how to quantify FDI quality and how it influences economic growth. However, there are still some deficiencies in the existing research: (1) The quantitative index system of FDI quality constructed by predecessors is difficult to meet the specific requirements imposed on FDI by the concept of sustainable development under the current international economic and political landscape. (2) Most of the indexes for quantifying the quality of FDI are only established from the economic dimension and at the macro level, without in-depth analysis of the specific characteristics that FDI enterprises should possess. (3) Although some of the literature has already explored the relationship between the quality of FDI and economic growth, few studies have focused on the impact of the quality of FDI on the sustainable development of the host country’s economy and society.

Therefore, on the basis of necessary theoretical analysis, this study constructs the quality evaluation index system of FDI and discusses the supporting mechanism of foreign capital quality improvement for sustainable development. The marginal contribution of this study is to design the quality index system of FDI from the micro level, combined with the latest theory and practice. Second, taking the Agenda 2030 as the framework of sustainable development, this study discusses the impact mechanism of FDI quality on sustainable development through empirical analysis. The follow-up arrangements of the article are as follows: the second part is literature review, the third part is theoretical analysis and research hypotheses, the fourth part is variable description and model setting, the fifth part is empirical analysis, and the sixth part is conclusions, policy recommendations, and outlook.

2. Literature Review

2.1. From Economic Growth to Sustainable Development

For decades, economic development has remained a central proposition in macroeconomic research. As societal attention increasingly extends to natural, social, and political dimensions encompassing income distribution, healthcare, environmental quality, and beyond, modern economic growth theories have undergone continuous evolution. This progression spans from the early neoclassical growth theory focused on capital accumulation [5], to endogenous growth models that internalize technological variables by incorporating knowledge and human capital, and further to contemporary extended frameworks integrating historical, sociopolitical, and even religious determinants [6,7]. Consequently, economic development has transcended its traditional emphasis on quantitative expansion and scale growth, with “quality” emerging as a transformative paradigm and strategic priority for national economic advancement. This paradigm shift in development philosophy and operational models has thereby institutionalized sustainability as an indispensable agenda in global economic governance.

Arrow et al. systematically categorize sustainability research into two paradigmatic strands [8]. The first strand operationalizes sustainability through the lens of contemporary human well-being, specifically evaluating whether current societal operations preserve future generations’ capacity to achieve welfare levels at least equivalent to the present level. The second strand adopts an intergenerational well-being framework, conceptualizing social welfare at any given period as a composite function encompassing both present welfare and the potential welfare of all future generations. While these approaches diverge in their parameterization of sustainability metrics, they converge axiomatically on the foundational premise that sustainability is fundamentally anchored in the principle of intergenerational equity—ensuring non-declining welfare trajectories across generations. This conceptual alignment notwithstanding, sustainability manifests as an emergent property of complex interplay between societal systems, technological frontiers, and natural capital dynamics [9]. To operationalize intergenerational welfare stability, economic systems must maintain equilibrium in per capita composite wealth, defined as the synergistic portfolio of material capital (productive infrastructure), human capital (knowledge stocks and health endowments), and natural capital (regenerative ecosystems). The maintenance of this tripartite capital stock constitutes both the necessary and sufficient condition for sustaining welfare-generating capacities across temporal horizons.

Compared with the traditional growth model, which blindly attaches importance to the growth of GDP and its growth rate, the sustainable development model takes into account the overall development of human society. Sustainable development takes improving total factor productivity as the core task, emphasizes the “multiplier effect” caused by human capital and scientific and technological innovation, and attaches importance to the creation of employment opportunities, the rise of the mid to high end of the industrial value chain, and the decoupling of pollutant emissions from economic growth. Since the United Nations put forward the Sustainable Development Goals (SDGs), some foreign scholars have created the SDGs composite index to monitor the performance of countries in sustainable development by studying the typical variables in the SDGs subset [10,11,12,13]. In contrast, Chinese scholars focus more on clarifying and explaining the logic of China’s high-quality economic development in order to establish a comprehensive measurement method to evaluate the quality of economic development.

2.2. From the Scale of FDI to the Quality of FDI

With the rapid development of multinational corporations in developed countries and the increasing frequency of international capital flows, theories on FDI have continuously evolved. From Hymer’s monopolistic advantage theory and Vernon’s product life cycle theory to Buckley and Casson’s internalization theory, and culminating in Dunning’s eclectic theory of international production, Western scholars have persistently explored the motivations and logic behind multinational corporations’ outward FDI [14,15,16,17].

The eclectic theory integrates various mainstream theories of multinational corporations, with its core framework being the OLI paradigm: ownership-specific advantages (O advantages), location-specific advantages (L advantages), and internalization-specific advantages (I advantages). Among these, ownership-specific advantages constitute the necessary condition for home-country enterprises to engage in FDI, implying that firms undertaking overseas investments typically possess superior production facilities, technological development, management capabilities, and risk mitigation capabilities compared to host-country enterprises. In essence, FDI inherently embodies tangible or intangible resources scarce in host countries, with the quantity and quality of such scarce resources defining the quality of FDI. However, this theory primarily focuses on developed-country multinational corporations, assuming that, under the historical context, developing-country enterprises inherently lack such monopolistic advantages. Based on this implicit assumption, cross-border capital received by developing countries is automatically presumed to exclude “low-quality” or “homogeneous” investments, leading to the oversimplified quantification of FDI at the numerical level alone.

The marginal industry expansion theory, also centered on developed-country multinationals, posits that outward investments should focus on “marginal industries”—those losing or about to lose comparative advantage in home countries but retaining comparative advantage in host countries [18]. Such industries not only facilitate corporate establishment in host countries through appropriate technological gaps but also better align with host countries’ factor endowments. Although this theory makes the capital, technology, and knowledge embedded in FDI more explicit to some extent, it overlooks the socio-environmental issues arising from the transfer of developed countries’ marginal industries. Consequently, while FDI may introduce new knowledge and technologies to developing countries, it simultaneously risks creating monopolistic pressures and environmental pollution in host economies.

Since the 21st century, developing countries have seized opportunities from technological revolutions to rapidly accumulate various foundational factors of production. As their comprehensive capabilities strengthened, they gradually integrated into the international investment system dominated by developed countries, attempting to consolidate natural, human, and technological resources on a broader scale. This significant transformation in global investment patterns has further led to two practical scenarios diverging from the theoretical assumptions discussed above. First, investment initiators are no longer confined to developed countries; multinational corporations from developing countries may also launch reverse gradient investments driven by expansion or profit-seeking motives. Second, for developing countries traditionally positioned as investment recipients, the so-called “marginal industries” that developed countries aim to transfer may no longer inherently possess technological superiority. In both scenarios, the inherent assumption of FDI’s “high-quality” nature is disrupted, necessitating a shift from oversimplified quantitative metrics to quality differentiation based on varying spillover capacities determined by factors such as origin, investment motives, and targets [19]. Consequently, as a form of internationalized capital participating in host economies, FDI aligns with the dynamic development trajectory of transitioning from “quantity” to “quality” in host countries. Specifically, once host countries address quantitative inadequacies, qualitative concerns become increasingly prominent. As developing economies enter new developmental phases, they no longer settle for FDI merely delivering generic scarce resources. Instead, host governments increasingly expect FDI to generate knowledge and technology spillovers while contributing to local societal and environmental well-being, thereby supporting sustainable development.

2.3. Academic Research and Practical Exploration on the Quality of FDI

The pursuit of sustainable development raises the question: What characteristics must FDI possess to align with sustainable development goals? Alternatively, how can enhancing FDI quality contribute to sustainable development? Although no literature has explicitly explored this aspect so far, there are abundant research findings on the external characteristics of FDI, which can provide insights for answering the above questions. Generally speaking, the externalities of FDI can be summarized as economic externalities and environmental externalities.

Economic externalities of FDI. Firstly, FDI possesses superior externalities in technological innovation. Multinational corporations achieve the transnational transfer of technology through initiating FDI, thereby triggering technology spillover and improves the technological level of the host country [20]. By exerting the leading effect of high-end science and technology [21], it can significantly enhance the local innovation performance level [22]. Secondly, FDI has positive externalities in industrial structure optimization [23]. The entry of FDI enterprises further intensifies competition in the domestic market [24]. While eliminating inefficient enterprises, it forces other domestic enterprises to optimize the allocation of resources within industries, thus promoting the efficiency of the industrial structure. Finally, FDI has certain market-driven externalities. Multinational corporations tend to make production investments in markets with a large scale of domestic demand to fully utilize economies of scale [25]. On this basis, in order to expand sales channels, multinational corporations often develop products and technologies that are more suitable for the market demands of the host country. By meeting the personalized and diversified needs of consumers, they can boost residents’ consumption confidence.

Environmental externalities of FDI. Studies on FDI’s environmental impacts predominantly revolve around two hypotheses: the “Pollution Haven” and “Pollution Halo”. The Pollution Haven Hypothesis posits that FDI, as a conduit for transferring low-end industries from multinational corporations, often exhibits energy-intensive, pollution-heavy, and low-value-added characteristics, thereby exacerbating ecological burdens in host countries [26,27]. Conversely, the Pollution Halo Hypothesis argues that multinational corporations typically possess superior green practices and cleaner technologies, whose knowledge and technology spillovers improve host countries’ environmental conditions [28]. Notably, most empirical findings derive from analyses of FDI’s quantitative dimensions, and the observed contradictions stem from the neglect of FDI’s qualitative attributes.

Building on this, the assumption of FDI homogeneity inherently neglects differences in embedded resources, necessitating quality assessment. Early efforts to eliminate the influence of national economic scales on FDI inflows led the United Nations Conference on Trade and Development (UNCTAD) to propose the FDI Performance Index in its World Investment Report 2001 [29]. This single-dimensional metric categorizes regions with scores ≥ 1 as “FDI-successful” and those below as “FDI-unsuccessful”. Seventeen years later, amid a global shift toward sustainable development, the Organization for Economic Co-operation and Development (OECD) launched the FDI Qualities Initiative in 2018 to clarify the relationship between FDI quality and sustainability across nations and policy contexts. In 2022, the OECD published FDI Qualities Indicators 2022 [30], which for the first time established micro-level requirements for foreign enterprises as FDI initiators, addressing a longstanding gap in the field. Concurrently, the OECD introduced the FDI Qualities Policy Toolkit to assist member states, particularly developing countries, in aligning investment policies with sustainability guidelines. At the academic level, Assanie and Singleton proposed that high-quality FDI is associated with more high and new technologies and can generate more technological spillovers [31]. Scholars such as Kumar and Buckley et al. began to shift their research perspectives to developing countries and attempted to construct comprehensive evaluation indices, aiming to assess the actual level of FDI quality from multiple dimensions [32,33]. Alfaro and Charlton employed a new dataset on industrial location and the two-stage least squares method to determine the quality of FDI shaped by the characteristics of the host country [19]. In addition, some scholars in China have also shifted the research focus of FDI to the quality aspect and are dedicated to quantitative research on the quality of FDI.

3. Theoretical Analyses and Research Hypotheses

Based on the above analysis, the external characteristics of FDI have been widely discussed in the academic community. However, to a certain extent, the impact it has on the transformation of the economic and social development models of the host country, due to differences in the quality dimension, has been overlooked. This study argues that as a comprehensive investment behavior encompassing various elements such as physical capital and intangible capital, FDI enables the host country to pursue sustainable development if and only if the various tangible and intangible resources carried by FDI enterprises fully meet the development needs of the host country at its current stage.

Firstly, achieving the effective allocation of resources is one of the core issues in promoting sustainable economic development, which is essentially determined by the limited nature of resources. In the traditional closed economy, resource allocation often focused on a single economic entity [34,35]. With the progress of time and the increasingly rapid process of economic globalization, resources are no longer restricted by the boundaries of a single economic entity. Multinational corporations can establish production and R&D bases in other economies through means such as greenfield investment and cross-border mergers and acquisitions. This measure enables the integration of elements such as capital, technology, and personnel within multinational corporations with local resources like labor and land in the host country. Furthermore, the effective allocation of resources requires that resources flow more precisely to the fields and links with the greatest value-creation potential, in accordance with market demand and the principle of efficiency. Therefore, the highly efficient production and sound internal governance system of FDI enterprises not only improve the efficiency of internal resource allocation within FDI enterprises themselves but also gather knowledge, technology, skills, and experience towards the high-value end, and stimulate local enterprises in the host country to update their production models and reform their internal governance systems. This leads to the transfer of more resources in the direction of low-cost and high-efficiency, thus promoting the flow of various resources towards high-value areas.

Secondly, enhancing the level of human capital is essentially a key means to resolve various development contradictions and achieve the goals of sustainable development. People are both the “beneficiaries” and the “drivers” of sustainable development. The shift of the economic development model from extensive growth to sustainable development requires people to combine existing resources, accumulate knowledge, technology, experience, and good health through education, training, practice, and other means, and ultimately promote the realization of the sustainable development of the entire society. Multinational corporations, by establishing processing plants, offices, research and development institutions, etc., in the host country, generate varying degrees of demand for the labor force of the host country, thereby influencing the level of human capital in the host country. This process is generally referred to as knowledge spillover [36,37]. To continuously boost the level of human capital in the host country, it is required that FDI enterprises continuously improve the quantity and quality of knowledge spillovers. That is to say, FDI enterprises should maintain higher standards in dimensions such as innovation, skills, and work quality. Specifically, on the one hand, in order to better manage the employees in the host country and enable the hired local employees to deeply understand the corporate culture and development strategy, and proficiently operate the production and processing processes of products, FDI enterprises will frequently dispatch experienced managers and technicians to the host country. They will impart relevant management experience and operational skills by carrying out educational activities such as knowledge popularization and skills training within the enterprise. At the same time, high-quality FDI enterprises are more inclined to recruit and cultivate local talents in the host country, so that the enterprise development plan can better adapt to the development stage and environment of the host country. On the other hand, in order to ensure quality control of the final products and expand sales channels, some FDI enterprises will also conduct corresponding knowledge and technology training for the manufacturing and assembly-type and marketing-type enterprises located downstream of their industrial chains.

Finally, practicing low-carbon transformation is key to breaking the deadlock in achieving sustainable development. The slowdown in global economic growth has made the ecological and environmental problems that were masked during periods of prosperity increasingly prominent. Issues such as the shortage of energy supply, the intensification of the greenhouse effect, the frequent occurrence of climate disasters, and the increase in environmental constraints, which are caused by the rise of industrial civilization, have increasingly become key shackles to the realization of the goal of sustainable development for humanity. Currently, given that some cross-border investments are still initiated based on marginal industry transfer, which has an adverse impact on the environment of the host country [38], FDI enterprises that focus on practicing low-carbon principles are becoming increasingly precious to the host country. On the one hand, the higher the degree of environmental emphasis of FDI enterprises, the more likely they are to reengineer production processes or create products with the aim of tackling clean energy and green low-carbon technologies. Therefore, they are more likely to create new knowledge and technologies in these fields and facilitate the diffusion of knowledge and technologies between the upstream and the downstream through the green industrial chain. On the other hand, the active assumption of environmental responsibilities by FDI enterprises helps them overcome the liability of foreignness and adapt to the locational characteristics of the host country. This makes it easier for them to establish contractual relationships with local suppliers, and the likelihood of knowledge and technology spillovers is relatively high. Therefore, FDI enterprises with such characteristics not only embed low-carbon awareness in the industry through knowledge and technology spillovers but also eliminate high-carbon, high-energy-consuming, and high-pollution enterprises within the industry through the market competition mechanism, thus promoting the low-carbon transformation of overall industries in the host country.

In conclusion, FDI enterprises are not only practitioners of the concept of sustainable development but also participants in the economic and social development of the host country. Through initiating international capital flows and maintaining control over the operation of overseas subsidiaries in the long term, multinational corporations provide a practical scenario for the host country to achieve sustainable development.

Based on this analysis, the following hypotheses are proposed:

H1. 

The higher the quality of FDI, the more it can promote the sustainable development of the host country.

H2. 

The higher the quality of FDI, the more efficient the allocation of resources in the host country, the more it promotes human capital accumulation in the host country, and the more it promotes low-carbon transformation in the host country.

Building on the above logical framework, this study identifies three critical pathways through which high-quality FDI drives sustainable development in host countries: enhanced resource allocation efficiency, human capital accumulation, and low-carbon transition. The analysis will focus on these three dimensions:

(1) Government-Market Synergy: Effective collaboration between government and market institutions amplifies the “catalyst effect” of high-quality FDI in optimizing resource allocation. On one hand, the market’s role in resource allocation determines the extent to which resources concentrate in efficient sectors and the speed of phasing out underperforming enterprises following the entry of high-quality foreign firms. A well-functioning market mechanism enables high-quality FDI to integrate and utilize host-country resources effectively, while intensifying competition to push domestic firms beyond their “comfort zones,” spurring technological and managerial upgrades. On the other hand, governments transitioning toward a “service-oriented” model—by rationalizing administrative scale and decentralizing authority—can reduce administrative monopolies and rent-seeking, preserving resource allocation efficiency.

(2) Human Capital and Intellectual Property (IP) Protection: The impact of high-quality FDI on human capital is contingent on IP protection levels. Strong IP safeguards are critical for multinational corporations to maintain core competitiveness and serve as prerequisites for technology collaboration or transfer with domestic firms [39]. Inadequate IP protection may deter knowledge and technology sharing by high-quality foreign enterprises.

(3) Environmental Regulation Intensity: The effectiveness of high-quality FDI as a low-carbon “engine” hinges on host governments’ environmental governance. Proactive measures, such as tax incentives for energy-saving firms, incentivize high-quality FDI to pursue low-carbon development and amplify green technology spillovers. Conversely, stringent environmental enforcement raises pollution costs, compelling all firms to offset expenses through green innovation and productivity enhancements [40].

Based on this analysis, the following hypotheses are proposed.

H3. 

The more coordinated the relationship between the government and the market, the stronger the role of high-quality FDI in improving the efficiency of resource allocation.

H4. 

The higher the level of intellectual property protection, the stronger the role of high-quality FDI in promoting human capital accumulation.

H5. 

The higher the level of environmental regulation, the stronger the role of high-quality FDI in promoting low-carbon transformation.

4. Variable Description and Model Setting

4.1. Data Sources

At the micro level, the sample comprises FDI enterprises listed on China’s A-share market from 2011 to 2022. Financial institutions, ST/ST* firms, and FDI enterprises with severe data deficiencies were excluded. At the macro level, the sample covers 30 provincial-level administrative regions in China (excluding Macao, Hong Kong, Tibet, and Taiwan) over the same period. The selection of 2011 as the starting year is justified by two considerations: firstly, mitigating the impact of the 2008 global financial crisis on both micro-level enterprises and macroeconomic stability. Secondly, 2011 marked a pivotal year for China’s renewed revision of the Catalogue for the Guidance of Foreign Investment Industries (last revised in 2007), reflecting efforts to further liberalize foreign investment. Notably, macro-level data exhibit inherent lagging characteristics. For instance, the fixed-asset investment price index for calculating Green Total Factor Productivity (GTFP) is only updated to 2019, and human capital data from the Central University of Finance and Economics extend only to 2022. To ensure temporal consistency between micro- and macro-level datasets and enhance the accuracy of empirical analysis on FDI quality’s role in sustainable development, 2022 is selected as the endpoint. Missing data points are supplemented using linear interpolation.

4.2. Description of Variables

4.2.1. Core Dependent Variable: Green Total Factor Productivity (GTFP)

Centered on the 5Ps—People, Planet, Prosperity, Peace, and Partnership—the SDGs encompass 17 global objectives, including climate action, quality education, and industrial innovation, with 169 targets and 232 indicators. Synthesizing these, sustainable development emphasizes the sustainability of human well-being, ecosystems, economic systems, geopolitical relations, and global diversity. This study adopts a host-country perspective to examine how international capital inflows influence local sustainable development practices. Following existing literature, GTFP is selected as the core dependent variable [41]. Unlike Total Factor Productivity (TFP), GTFP incorporates undesirable outputs during economic development, aligning more closely with sustainability principles. Building on TFP, environmental factors are integrated, and the Super Slacks-Based Measurement (Super SBM) model is employed to measure GTFP across China’s provinces from 2011 to 2022 [42]. This study uses Max DEA Lite (v12.2) software to measure GTFP.

4.2.2. Secondary Dependent Variables

As outlined in the theoretical and mechanistic analysis above, rational resource allocation, robust human capital, and green low-carbon production practices constitute foundational pillars for maintaining holistic sustainability in an economy. Thus, analyzing economic sustainability requires examining key sub-indicators under GTFP. Accordingly:

Resource Allocation Efficiency (EC₁): Measured using the technical efficiency change index to evaluate the efficiency of resource allocation within the economy.

Human Capital (HC₁): Represented by the logarithm of the China JF real human capital stock, an annual metric published by the Human Capital and Labor Economic Research Center at the Central University of Finance and Economics, to assess the economy’s human capital capacity.

Carbon Emission Intensity (CEI): Defined as carbon emissions per unit of GDP, reflecting the economy’s progress in low-carbon transition.

4.2.3. Core Explanatory Variable: FDI Quality Indicators (fdiq)

This study constructs a micro-level FDI quality index comprising 12 indicators, grounded in China’s requirements for foreign capital characteristics in its new developmental phase and aligned with the OECD’s FDI Quality Indicators 2022 framework. The selection adheres to principles of scientific rigor, comprehensiveness, comparability, data availability, and the integration of absolute and relative metrics.

Productivity and Innovation Cluster: It is characterized by three first-level indicators, namely production efficiency, innovation input, and innovation output. Labor productivity is selected as the secondary indicator of production efficiency because it can reflect the operating revenue created by a unit labor force. R&D investment serves as the material basis for innovation input, and the proportion of R&D personnel reflects the input of innovative knowledge. These two aspects are used comprehensively to measure innovation input. Due to the problems such as data gaps in the number of granted patents and other statistics of some foreign-invested enterprises, in order to ensure the quality of the data, the number of patent applications is chosen as the secondary indicator of innovation output, as it can directly demonstrate the innovation achievements.

Employment and Work Quality Cluster: Most of the employee treatment indicators in the CNRDS database are virtual indicators, which are of no practical significance and are not suitable for synthesis. This article selects the wage level to characterize this aspect. This is because the wage level can not only reflect the work value of employees but also determine their material quality of life, and it is a core factor influencing employees’ work enthusiasm and satisfaction.

Skill Cluster: It is characterized by two first-level indicators: knowledge level and technical level. The knowledge level is measured by the proportion of employees with a postgraduate degree or above. This is because such employees have received systematic education and training and possess rich professional knowledge reserves. A larger number of them indicates a higher level of professional knowledge reserve in the enterprise in this field. The technical level is measured by the proportion of technical staff. Technical staff members master professional technical capabilities. An increase in their proportion shows an expansion of the FDI enterprise’s technical team scale, which reflects the FDI enterprise’s technical strength.

Gender Equality Cluster: The degree of emphasis on women’s rights is used as the primary indicator, and three secondary indicators, namely the proportion of women in the supervision layer, the proportion of female directors, and the proportion of female senior managers, are set up to comprehensively characterize it. These indicators can comprehensively and intuitively reflect the status and role of women in the decision-making of FDI enterprises, demonstrate the efforts made by FDI enterprises to break gender barriers, and are an important manifestation of the internal governance system of FDI enterprises.

Carbon Footprint Cluster: The environmental awareness of FDI enterprises is measured by the environmental attention of senior executives, which is defined as the frequency of keywords related to environmental issues in the corporate social responsibility report [43,44]. This is because the emphasis on the environment by senior executives will guide enterprises to incorporate environmental factors into their strategic planning and allocate resources accordingly [45]. At the same time, the score of the Environment (E) sub-item in the Huazheng ESG rating system is used to comprehensively represent the fulfillment of social responsibilities in the environmental aspect by FDI enterprises. A higher score indicates that FDI enterprises are more willing to assume external environmental responsibilities.

Executives’ environmental attention is defined as the keyword frequency of environment-related topics in corporate social responsibility (CSR) reports. Data on postgraduate, technical, and R&D personnel ratios and green patent applications are sourced from the CNRDS database. Remaining data are obtained from the CSMAR database and Mark data network. Specific indicators are detailed in

Table 1. The quality indicator system of foreign direct investment.

Cluster	Primary Indicators	Secondary Indicators	Index Interpretation
Productivity and Innovation	Production Efficiency	X1: Labor productivity	Business income /Number of employees
	Innovation Investment	X2: R&D investment	R&D investment of the enterprise in that year
		X3: Proportion of R&D personnel	Number of R&D personnel /Number of employees
	Innovation Output	X4: Patent applications	Patent applications of enterprises in that year
Employment and Job Quality	Wage Level	X5: Per capita compensation	Payroll payable /Number of employees
Technical Ability	Knowledge Level	X6: Proportion of graduate students and above	Number of graduate students and above/Number of employees
	Technical Level	X7: Proportion of technicians	Number of enterprise technicians /Number of employees
Gender Equality	Women’s Rights	X8: Proportion of female supervisors	Number of female corporate supervisors /Total number of corporate regulators
		X9: Proportion of female directors	Number of female directors /Total number of directors
		X10: Proportion of female senior managers	Number of female senior managers /Total number of senior managers
Carbon Footprint	Environmental Consciousness	X11: Executive environmental attention	Word frequency of environmental keywords in corporate social responsibility report/Total word frequency of Corporate Social Responsibility Report
	Climate Change	X12: Corporate environmental responsibility	Score of environment (E) sub-item in China Securities ESG rating system
	Resource Utilization
	Environmental Pollution
	Environmentally Friendly
	Environmental Management

.

4.2.4. Control Variables

The analysis includes controls at three levels: firm, industry, and province.

(1)

Firm-level controls:

Ownership type (doe): Dummy variable (1 for state-owned enterprises, 0 otherwise).

Firm age (age): Measured as the number of years since establishment.

Equity concentration (ecd): Proportion of shares held by the top 10 shareholders.

Firm size (size): Natural logarithm of total assets.

Return on assets (roa): Net profit divided by average total assets.

(2)

Industry-level control:

Industry competition (hhi): Herfindahl–Hirschman Index (HHI) calculated for each industry.

(3)

Province-level controls:

Economic development (growth): Natural logarithm of GDP per capita.

Industrialization level (il): Ratio of industrial added value to provincial GDP.

Technology market development (tm): Natural logarithm of technology market transaction value.

Tax burden (tax): Ratio of tax revenue to provincial GDP.

R&D intensity (rd): Ratio of internal R&D expenditure to provincial GDP.

Social consumption (consume): Ratio of total retail sales of consumer goods to provincial GDP.

4.2.5. Moderating Variables

As analyzed in the preceding mechanism discussion, the synergistic government-market relationship, intellectual property protection (IPP) level, and environmental regulation intensity may moderate the impact of FDI quality on sustainable development. Accordingly, this study selects:

Government-market synergy (gov): Measured by the government-market relationship sub-index from the Fan Gang Marketization Index.

Intellectual property protection (ipp): Evaluated using IP protection scores from the National Intellectual Property Development Status Report published by the National Intellectual Property Administration.

Environmental regulation intensity (er): Calculated as the ratio of completed investment in industrial pollution control to the value-added of the secondary sector.

The types, symbols, and names of all variables can be found in

Table 2. Main variable names and symbols.

Variable Type		Variable Symbol	Variable Name
Core Dependent Variable		GTFP	Green Total Factor Productivity
Secondary Dependent Variables		EC1	Resource Allocation Efficiency
		HC1	Human Capital
		CEI	Carbon Emission Intensity
Core Explanatory Variable		fdiq	FDI Quality Indicators
Control Variables	Firm-level Controls	doe	Ownership Type
		age	Firm Age
		ecd	Equity Concentration
		size	Firm Size
		roa	Return on Assets
	Industry-level Controls	hhi	Industry Competition
	Province-level Controls	growth	Economic Development
		il	Industrialization Level
		tm	Technology Market Development
		tax	Tax Burden
		rd	R&D Intensity
		consume	Social Consumption
Moderating Variables		gov	Government-market Synergy
		ipp	Intellectual Property Protection
		er	Environmental Regulation Intensity

.

4.3. Model Construction and Estimation Method

4.3.1. Entropy Weight Method

The entropy weight method, an objective weighting approach, is employed to construct the FDI quality index. This method determines the weight of each indicator in multi-criteria decision analysis based on its information entropy: A higher entropy value indicates less informational contribution from the indicator, resulting in a lower assigned weight, and vice versa. By avoiding subjective biases inherent in expert judgment-based weighting methods, the entropy weight method dynamically adjusts weights according to data variability, thereby objectively, timely, and accurately reflecting the relative importance of indicators [46,47]. Consequently, it has been widely adopted by Chinese domestic scholars studying FDI quality for indicator aggregation.

Prior to weighting, raw data undergo dimensionless normalization. Assuming there are $\mathrm{m}$ evaluation units and $\mathrm{n}$ evaluation indicators, the raw data matrix is denoted as ${\mathrm{X}}_{\mathrm{i}\mathrm{j}}$, where $\mathrm{i}=1,2,\cdots ,\mathrm{m};\mathrm{j}=1,2,\cdots ,\mathrm{n}$.

$${\mathrm{X}}_{\mathrm{i}\mathrm{j}}={\displaystyle \frac{{\mathrm{X}}_{\mathrm{i}\mathrm{j}}-{\mathrm{X}}_{\mathrm{m}\mathrm{i}\mathrm{n}}}{{\mathrm{X}}_{\mathrm{m}\mathrm{a}\mathrm{x}}-{\mathrm{X}}_{\mathrm{m}\mathrm{i}\mathrm{n}}}}(\mathrm{B}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{f}\mathrm{i}\mathrm{t}-\mathrm{b}\mathrm{a}\mathrm{s}\mathrm{e}\mathrm{d})$$

(1)

$${\mathrm{X}}_{\mathrm{i}\mathrm{j}}={\displaystyle \frac{{\mathrm{X}}_{\mathrm{m}\mathrm{a}\mathrm{x}}-{\mathrm{X}}_{\mathrm{i}\mathrm{j}}}{{\mathrm{X}}_{\mathrm{m}\mathrm{a}\mathrm{x}}-{\mathrm{X}}_{\mathrm{m}\mathrm{i}\mathrm{n}}}}(\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}-\mathrm{b}\mathrm{a}\mathrm{s}\mathrm{e}\mathrm{d})$$

(2)

The specific formula of entropy weight method is as follows:

First, calculate the probability of the j-th indicator for the i-th evaluation unit.

$${\mathrm{P}}_{\mathrm{i}\mathrm{j}}={\displaystyle \frac{{\mathrm{X}}_{\mathrm{i}\mathrm{j}}}{\sum _{\mathrm{i}=1}^{\mathrm{m}}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}}}$$

(3)

Next, calculate the information entropy of the j-th indicator.

$${\mathrm{e}}_{\mathrm{j}}=-{\displaystyle \frac{1}{\mathrm{l}\mathrm{n}\mathrm{m}}}\sum _{\mathrm{i}=1}^{\mathrm{m}}{\mathrm{P}}_{\mathrm{i}\mathrm{j}}\mathrm{l}\mathrm{n}{\mathrm{P}}_{\mathrm{i}\mathrm{j}}$$

(4)

Finally, calculate the weight of the j-th indicator.

$${\mathsf{\omega }}_1={\displaystyle \frac{1-{\mathrm{e}}_{\mathrm{j}}}{\sum _{\mathrm{j}=1}^{\mathrm{n}}(1-{\mathrm{e}}_{\mathrm{j}})}}$$

(5)

4.3.2. Benchmark Regression Model

Using a fixed effects model can avoid problems such as heterogeneous bias; therefore, referring to the general practice of the current FDI literature [38,48], the fixed effects model is used to analyze the impact of FDI quality on sustainable development. Based on this, model (6) and model (7) are established to examine the impact of FDI quality on China’s sustainable development.

$$GTF{P}_{p,t}={\beta }_0+{\beta }_{1}fdi{q}_{i,t}+\sum ctr{l}_{i,t}+\sum ctr{l}_{j,t}+\sum ctr{l}_{p,t}+\{F\}+{\epsilon }_{i,t}$$

(6)

$$\begin{array}{c}{EC1}_{p,t}={\gamma }_0+{\gamma }_{1}fdi{q}_{i,t}+\sum ctr{l}_{i,t}+\sum ctr{l}_{j,t}+\sum ctr{l}_{p,t}+\{F\}+{\epsilon }_{i,t}\\ {HC1}_{p,t}={\gamma }_0+{\gamma }_{1}fdi{q}_{i,t}+\sum ctr{l}_{i,t}+\sum ctr{l}_{j,t}+\sum ctr{l}_{p,t}+\{F\}+{\epsilon }_{i,t}\\ {CEI}_{p,t}={\gamma }_0+{\gamma }_{1}fdi{q}_{i,t}+\sum ctr{l}_{i,t}+\sum ctr{l}_{j,t}+\sum ctr{l}_{p,t}+\{F\}+{\epsilon }_{i,t}\end{array}$$

(7)

Among them, $GTF{P}_{p,t}$ is the core explanatory variable, green total factor productivity, ${EC1}_{p,t}$, ${HC1}_{p,t}$, and ${CEI}_{p,t}$ are the resource allocation efficiency, human capital level, and carbon emission level; $fdi{q}_{i,t}$ is the core explanatory variable of the quality of this study; $\sum ctr{l}_{i,t}$, $\sum ctr{l}_{j,t}$, and $\sum ctr{l}_{p,t}$ of foreign direct investment represents the control variables; and ${\epsilon }_{i,t}$ serves as the random disturbance item at the enterprise, industry, and province levels, respectively. In addition, the fixed effects of year, enterprise, industry, and province are controlled in the benchmark model: $\{F\}=\left\{{\phi }_t,{\rho }_i,{\rho }_j,\left.{\rho }_p\right\}\right.$, and they are clustered at the enterprise level.

4.3.3. Moderating Effect Model

To rigorously examine how government–market synergy, intellectual property protection, and environmental regulation intensity moderate the impact of FDI quality on sustainable development, we have established the following moderating effect model with reference to Rong et al.’s research [49]:

$$\begin{array}{c}{Y}_{p,t}={\alpha }_0+{\alpha }_{1}fdi{q}_{i,t}+{\alpha }_{2}M+{\alpha }_{3}M\times fdi{q}_{i,t}+\sum ctr{l}_{i,t}\\ +\sum ctr{l}_{j,t}+\sum ctr{l}_{p,t}+\{F\}+{\epsilon }_{i,t}\end{array}$$

(8)

In this model, $Y_{p,t}$ refers to each explained variable, $M$ is a regulatory variable, and the meaning of other variables is consistent with that of model (6) and model (7). This study will focus on the coefficient of the interaction term ($M\times fdi{q}_{i,t}$) to clarify the actual impact of moderating variables on the sustainable development of the quality of FDI.

We use STATA 17 software to apply the entropy weight method and subsequent empirical analysis. Before regression, all variables were subjected to winsorization.

5. Results and Discussion

5.1. Variable Descriptive Statistics

First of all, as presented in

Table 3. Descriptive statistics of the quality indicator system of foreign direct investment.

Cluster	Obs	Mean	Median	Std. Dev
Productivity and Innovation	1400	0.013	0.006	0.037
Employment and Job Quality	1400	0.001	0.000	0.027
Technical Ability	1400	0.097	0.058	0.127
Gender Equality	1400	0.301	0.279	0.191
Carbon Footprint	1400	0.163	0.135	0.098
fdiq	1400	0.021	0.018	0.024

, the descriptive statistics of the indicator system for FDI quality are summarized. Since all indicators were standardized, the analysis focuses on three statistical measures: mean, median, and standard deviation. The results reveal that the mean values of the five clusters and the composite FDI quality index slightly exceed their respective medians, indicating a right-skewed distribution across the dataset. This skewness suggests that a subset of foreign enterprises exhibits relatively superior performance across clusters, elevating the overall FDI quality level. Furthermore, the skill, gender equality, and carbon footprint clusters display larger standard deviations (0.127, 0.191, and 0.098, respectively), reflecting greater variability and dispersion in these dimensions. This heterogeneity underscores significant differences in foreign enterprises’ performance regarding skill development, gender equality practices, and carbon footprint management.

Second, as shown in

Table 4. Descriptive statistics of the main variables.

Variables	Obs	Mean	Std. Dev	Min	Max
GTFP	1400	0.991	0.022	0.917	1.117
EC1	1400	0.998	0.016	0.954	1.165
HC1	1400	10.120	0.476	8.055	10.880
CEI	1400	0.733	0.507	0.021	3.375
fdiq	1400	−4.009	0.531	−5.633	−0.252
doe	1400	0.716	0.451	0.000	1.000
age	1400	2.861	0.348	1.343	3.716
ecd	1400	0.660	0.145	0.196	1.000
size	1400	21.950	1.136	19.160	26.370
roa	1400	0.062	0.071	−0.653	0.384
hhi	1400	0.186	0.193	0.023	1.000
growth	1400	11.400	0.385	10.010	12.150
il	1400	0.348	0.070	0.107	0.461
tm	1400	0.025	0.032	0.000	0.191
tax	1400	0.095	0.036	0.036	0.188
rd	1400	0.028	0.011	0.007	0.070
consume	1400	0.402	0.037	0.221	0.504

Note: In order to show the real situation of the variables, the variables were not condensed in the descriptive statistics.

, for the dependent variables, the standard deviations of GTFP and EC₁ are 0.022 and 0.016, respectively, indicating relatively small interprovincial disparities in China in green total factor productivity and resource allocation efficiency. In contrast, HC₁ and CEI exhibit significantly larger standard deviations (0.476 and 0.507), demonstrating pronounced provincial heterogeneity in human capital levels and carbon emission intensities. Regarding control variables, the variable ‘firm size’ (size) displays a maximum value of 26.37, a minimum of 19.16, and a standard deviation of 1.136. This reflects substantial variation in asset scales among FDI enterprises in China, confirming that the sample encompasses both startups conducting exploratory operations and large multinational corporations with long-established market presence.

Finally, Pearson correlation analysis was conducted to preliminarily examine the relationships between the core and secondary dependent variables. As shown in

Table 5. Correlation analysis among the explained variables.

	EC1	HC1	CEI
GTFP	0.522 ***
GTFP		0.357 ***
GTFP			−0.196 ***

Note: *** p < 0.01.

, the Pearson correlation coefficients reveal that GTFP is significantly positively correlated with resource allocation efficiency (EC₁) and human capital level (HC₁), while exhibiting a significantly negative correlation with carbon emission intensity (CEI). These findings substantiate that enhancing resource allocation efficiency, fostering human capital accumulation, and reducing carbon emissions are critical for advancing sustainable development.

5.2. Benchmark Regression Analysis

First, to determine whether the fixed effects model we use is correct, we perform the Hausman test. In this test, the null hypothesis is that the preferred model is a random effects model, and the alternative hypothesis is that the preferred model is a fixed effects model [50]. The Hausman test rejects the null hypothesis at the 1% significance level, proving that the fixed effects model is reasonable. Second, the variance inflation factor (VIF) diagnostics (mean VIF = 2.410 < 10) indicate no severe multicollinearity issues. Finally, the coefficient performance of fdiq is examined by progressively controlling for firm, industry, province, and time-fixed effects. All baseline regression results are reported in

Table 6. Benchmark regression results of Model (6).

	(1)	(2)	(3)	(4)	(5)
	GTFP	GTFP	GTFP	GTFP	GTFP
fdiq	0.0034 ***	0.0034 **	0.0039 **	0.0041 **	0.0035 **
	(0.0012)	(0.0017)	(0.0017)	(0.0017)	(0.0015)
doe	0.0023 ***	0.0023 **	0.0022 *	0.0009	−0.0000
	(0.0008)	(0.0011)	(0.0011)	(0.0011)	(0.0011)
age	0.0085	0.0085	0.0090	0.0089	−0.0070
	(0.0052)	(0.0101)	(0.0108)	(0.0106)	(0.0074)
ecd	0.0071 *	0.0071	0.0067	0.0053	0.0067
	(0.0039)	(0.0053)	(0.0063)	(0.0063)	(0.0060)
size	0.0032 ***	0.0032 **	0.0035 **	0.0033 **	0.0024 *
	(0.0008)	(0.0014)	(0.0017)	(0.0016)	(0.0013)
roa	0.0018	0.0018	0.0040	0.0033	0.0041
	(0.0059)	(0.0075)	(0.0076)	(0.0076)	(0.0076)
hhi	−0.0025	−0.0025	−0.0050 *	−0.0054 *	−0.0057 **
	(0.0021)	(0.0026)	(0.0030)	(0.0030)	(0.0028)
growth	0.0391 ***	0.0391 ***	0.0369 ***	0.0458 ***	0.0205
	(0.0042)	(0.0075)	(0.0080)	(0.0061)	(0.0211)
il	0.0624 ***	0.0624 *	0.0500	0.0341	0.1423 ***
	(0.0232)	(0.0345)	(0.0352)	(0.0373)	(0.0448)
tm	−1.1478 ***	−1.1478 ***	−1.1567 ***	−1.1381 ***	−1.1853 ***
	(0.0395)	(0.0882)	(0.0904)	(0.0886)	(0.0818)
tax	0.1492 ***	0.1492 **	0.1392 **	0.1286 **	0.4108 ***
	(0.0424)	(0.0635)	(0.0658)	(0.0637)	(0.1026)
rd	−0.2659 *	−0.2659	−0.3334	−1.3166 *	−2.4379 ***
	(0.1458)	(0.4684)	(0.5190)	(0.7009)	(0.7495)
consume	−0.2297 ***	−0.2297 ***	−0.2405 ***	−0.2706 ***	−0.2008 ***
	(0.0181)	(0.0410)	(0.0454)	(0.0542)	(0.0730)
_cons	0.5504 ***	0.5501 ***	0.5825 ***	0.5349 ***	0.8253 ***
	(0.0487)	(0.0851)	(0.0911)	(0.0788)	(0.2831)
Firm FE	NO	YES	YES	YES	YES
Industry FE	NO	NO	YES	YES	YES
Province FE	NO	NO	NO	YES	YES
Year FE	NO	NO	NO	NO	YES
Obs	1400	1345	1343	1343	1343
adj. R2	0.4180	0.8062	0.8053	0.8076	0.8395

Note: Robust standard errors in brackets. * p < 0.1, ** p < 0.05, *** p < 0.01.

.

As shown in columns (1) to (5) of Table 6, after progressively controlling for individual and time fixed effects, the estimated coefficients of the core explanatory variable (fdiq) remain significantly positive at the 1% or 5% significance levels across all specifications, with minimal coefficient magnitude variation. This provides preliminary evidence for the positive impact of FDI quality on sustainable development. In column (5), the coefficient of fdiq is 0.0035, implying that, ceteris paribus, a 1% increase in firm-level FDI quality corresponds to a 0.35% rise in provincial GTFP. Given that this economic impact propagates from micro-level firms to macro-level provinces, the magnitude holds substantial policy relevance. Further calculation of economic significance follows the formula: $E{\displaystyle \frac{v}{s}}=\left|b{s}_x/\overline{Y}\right|$, where b is the regression coefficient (0.0035), $s_x$ is the standard deviation of fdiq (0.5310), and $\overline{Y}$ is the mean GTFP (0.9910). The result shows that a one-standard-deviation improvement in FDI quality drives a 0.1859 percentage-point increase in GTFP (0.0035 × 0.5310), accounting for 0.1876% of the sample mean GTFP (0.1859/0.9910). These findings confirm H1.

Following the confirmation of FDI quality’s positive impact on the core dependent variable (GTFP), this study conducts fixed effects regression analysis on the three remaining secondary dependent variables based on Model (7). The regression results are presented in

Table 7. Benchmark regression results of Model (7).

	(1)	(2)	(3)
	EC1	HC1	CEI
fdiq	0.0029 **	0.0205 *	−0.0007
	(0.0013)	(0.0115)	(0.0175)
doe	−0.0023 **	−0.0033	−0.0267
	(0.0010)	(0.0076)	(0.0184)
age	−0.0074	−0.1097	0.0894
	(0.0068)	(0.1076)	(0.1444)
ecd	−0.0013	0.0239	−0.1871 *
	(0.0053)	(0.0535)	(0.0981)
size	0.0017	0.0042	0.0019
	(0.0013)	(0.0140)	(0.0131)
roa	0.0039	−0.0136	0.0676
	(0.0061)	(0.0640)	(0.0971)
hhi	−0.0039	−0.0368 **	0.0277
	(0.0025)	(0.0147)	(0.0327)
growth	0.0157	0.4899 **	−0.9616 **
	(0.0161)	(0.2417)	(0.3816)
il	0.0986 **	1.6052 ***	−0.5646
	(0.0496)	(0.5295)	(1.3126)
tm	−0.6014 ***	−2.7653 ***	−1.8456
	(0.0468)	(0.4362)	(1.4865)
tax	0.0078	1.1004 *	−7.3280 ***
	(0.0805)	(0.6287)	(2.3605)
rd	−0.7339	−14.6637 ***	28.8016 ***
	(0.5541)	(5.5178)	(11.0293)
consume	−0.0986 **	0.6763 *	−2.9730 ***
	(0.0484)	(0.3952)	(0.7057)
_cons	0.8584 ***	4.3868	12.8541 ***
	(0.2082)	(2.8606)	(4.8515)
Firm FE	YES	YES	YES
Industry FE	YES	YES	YES
Province FE	YES	YES	YES
Year FE	YES	YES	YES
Obs	1343	1343	1343
adj. R2	0.6350	0.9885	0.9373

Note: Robust standard errors in brackets. * p < 0.1, ** p < 0.05, *** p < 0.01.

.

As shown in columns (1) to (2) of Table 7, the coefficients of fdiq on EC₁ and HC₁ are significantly positive at the 5% and 10% levels, respectively. Specifically, a 1% increase in FDI quality corresponds to a 0.29% improvement in resource allocation efficiency and a 2.05% rise in human capital levels at the provincial level. In standardized terms, a one-standard-deviation increase in firm-level FDI quality drives average changes of 0.1543% (0.2900 × 0.5310/0.9980) in resource allocation efficiency and 0.1066% (0.0205 × 0.5310/10.21) in human capital levels. Notably, column (3) reveals that fdiq exhibits a negative yet statistically insignificant effect on CEI. This implies that despite the improving quality of FDI inflows to China in recent years, it has yet to exert substantial influence on the low-carbon transition at the societal level. This result suggests deficiencies in environmental responsibility awareness and commitment among current FDI enterprises in China. Consequently, H2 is only partially validated.

5.3. Robustness Test

5.3.1. Robustness Test of Model (6)

Replace the dependent variable [51]. The Super SBM model is reapplied to calculate TFP instead of the GTFP, excluding undesirable outputs from the measurement. The regression results are shown in column (1) of

Table 8. Results of the robustness test for Model (6).

	(1)	(2)	(3)	(4)	(5)
	TFP	Clustering Adjustment	Clustering Adjustment	Additional Controls	Reduced Controls
fdiq	0.0035 **	0.0035 **	0.0035 **	0.0035 **	0.0031 **
	(0.0015)	(0.0016)	(0.0016)	(0.0015)	(0.0015)
doe	0.0002	−0.0000	−0.0000	−0.0001	−0.0017
	(0.0011)	(0.0011)	(0.0039)	(0.0011)	(0.0012)
age	−0.0068	−0.0070	−0.0070	−0.0062	−0.0054
	(0.0073)	(0.0045)	(0.0066)	(0.0077)	(0.0075)
ecd	0.0067	0.0067	0.0067	0.0069	0.0045
	(0.0060)	(0.0091)	(0.0050)	(0.0057)	(0.0057)
size	0.0025 *	0.0024 *	0.0024	0.0026 *	0.0025 *
	(0.0013)	(0.0014)	(0.0017)	(0.0014)	(0.0014)
roa	0.0042	0.0041	0.0041	0.0030
	(0.0076)	(0.0082)	(0.0075)	(0.0078)
hhi	−0.0056 **	−0.0057 **	−0.0057 *	−0.0055 *	−0.0039
	(0.0028)	(0.0023)	(0.0030)	(0.0028)	(0.0030)
growth	0.0201	0.0205	0.0205	0.0214	0.0161
	(0.0208)	(0.0230)	(0.0267)	(0.0216)	(0.0244)
il	0.1318 ***	0.1423 ***	0.1423	0.1414 ***	0.1866 ***
	(0.0443)	(0.0522)	(0.1307)	(0.0452)	(0.0530)
tm	−1.1893 ***	−1.1853 ***	−1.1853 ***	−1.1823 ***	−1.1381 ***
	(0.0827)	(0.0812)	(0.1892)	(0.0818)	(0.0687)
tax	0.4016 ***	0.4108 ***	0.4108	0.4129 ***	0.5454 ***
	(0.1028)	(0.0968)	(0.3533)	(0.1034)	(0.1024)
rd	−2.4010 ***	−2.4379 ***	−2.4379 **	−2.4419 ***	−2.1776 ***
	(0.7429)	(0.8026)	(1.0634)	(0.7343)	(0.7798)
consume	−0.2112 ***	−0.2008 **	−0.2008	−0.2033 ***
	(0.0742)	(0.0798)	(0.1205)	(0.0775)
dar				−0.0039
				(0.0041)
open				0.0022
				(0.0096)
_cons	0.8349 ***	0.8253 **	0.8253 **	0.8086 ***	0.7520 **
	(0.2796)	(0.3129)	(0.3836)	(0.2901)	(0.3047)
Firm FE	YES	YES	YES	YES	YES
Industry FE	YES	YES	YES	YES	YES
Province FE	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES
Obs	1343	1343	1343	1343	1343
adj. R2	0.8476	0.8422	0.8424	0.8394	0.8289

Note: Robust standard errors in brackets. * p < 0.1, ** p < 0.05, *** p < 0.01.

. The estimated coefficient of fdiq is still significant at the 5% level, and the magnitude of the coefficient remains unchanged. This indicates that while the benchmark regression results of this study are robust to a certain extent, it also reveals that improving fdiq indeed has no significant impact on the undesirable output embedded in GTFP—specifically, carbon emissions.

Change the clustering method [52]. In this study, while controlling for firm, industry, province, and time fixed effects, the clustering method is changed from the firm level to the industry and province levels. The regression results are shown in columns (2) and (3) of Table 8. The estimated coefficient of fdiq on GTFP is significant at the 5% level, and its magnitude remains nearly identical to that of the benchmark regression.

Add control variables [53]. At the micro-level, a moderate asset–liability ratio is the key for an enterprise to achieve sustainable development in the market. At the macro-level, a high level of opening-up is the driving force for China to smooth the domestic and international dual-circulation and thus achieve high-quality development. Based on this, this study introduces the micro-level control variable, the asset-liability ratio (dar), using the proportion of total corporate liabilities to the total corporate assets. We also introduce the macro-level control variable, the opening-up level (open), using the proportion of the total import and export value of each province to its GDP, and then conduct a re-regression of the model. The regression results are shown in column (4) of Table 8. The estimated coefficient of fdiq on GTFP is significant at the 5% level, which proves the robustness of the conclusions of this study.

Reduce control variables. A large number of control variables may lead to overfitting of the model. Therefore, this study successively removes the control variables roa and consume to recheck the robustness of the model. The regression results in column (5) of Table 8 are basically consistent with the benchmark regression results of this study, which again indicates that the conclusions of this study are robust.

5.3.2. Robustness Test of Model (7)

Alternative Measure for EC₁: EC₁ is replaced with the technical change index EC₂ derived from TFP decomposition. As shown in column (1) of

Table 9. Results of the robustness test for Model (7).

	(1)	(2)	(3)
	EC2	HC2	LNCE
fdiq	0.0029 **	0.0002 *	−0.0060
	(0.0013)	(0.0001)	(0.0154)
doe	−0.0023 **	−0.0005 ***	−0.0529 ***
	(0.0010)	(0.0001)	(0.0123)
age	−0.0072	−0.0015	0.0293
	(0.0067)	(0.0010)	(0.1104)
ecd	−0.0014	0.0008	0.0448
	(0.0053)	(0.0007)	(0.0621)
size	0.0017	−0.0000	−0.0060
	(0.0013)	(0.0001)	(0.0163)
roa	0.0039	0.0004	−0.1229
	(0.0060)	(0.0008)	(0.0917)
hhi	−0.0038	−0.0007 **	0.0049
	(0.0025)	(0.0003)	(0.0290)
growth	0.0163	0.0020	0.2591
	(0.0161)	(0.0045)	(0.1857)
il	0.0921 *	−0.0033	3.0885 ***
	(0.0500)	(0.0073)	(0.7377)
tm	−0.5976 ***	−0.0604 ***	−2.5604 ***
	(0.0472)	(0.0080)	(0.7127)
tax	0.0106	0.0267 **	5.7725 ***
	(0.0801)	(0.0124)	(1.3318)
rd	−0.7750	−0.1443	−5.5761
	(0.5584)	(0.0942)	(3.3928)
consume	−0.1004 **	0.0037	−0.6383
	(0.0485)	(0.0069)	(0.5108)
_cons	0.8543 ***	0.0057	6.5121 ***
	(0.2080)	(0.0561)	(2.2608)
Firm FE	YES	YES	YES
Industry FE	YES	YES	YES
Province FE	YES	YES	YES
Year FE	YES	YES	YES
Obs	1343	1343	1343
adj. R2	0.6356	0.9616	0.9655

Note: Robust standard errors in brackets. * p < 0.1, ** p < 0.05, *** p < 0.01.

, fdiq exhibits a significantly positive coefficient at the 5% level, confirming robustness.

Alternative Measure for HC₁: Substituting HC₁ with HC₂, measured as the ratio of college/university enrollment to regional population. Column (2) of Table 9 reveals a positive effect of fdiq on HC₂ at the 10% level, indicating that a 1% increase in FDI quality raises HC₂ by 0.2 percentage points. The smaller coefficient compared to HC₁ may arise from HC₁’s comprehensive incorporation of educational attainment, work experience, and health status, whereas HC₂ primarily reflects educational levels, suggesting FDI quality improvements more strongly benefit skill-based human capital than knowledge-based capital.

Alternative Measure for CEI: CEI is replaced by using the natural logarithm of total carbon emissions (LNCE). Column (3) of Table 9 shows the fdiq’s coefficient remains negative and insignificant, reiterating that foreign enterprises in China have yet to prioritize the low-carbon transition despite recent quality improvements.

5.4. Endogeneity Test

To address potential estimation bias arising from bidirectional causality between dependent and explanatory variables, this study adopts the Two-Stage Least Squares (2SLS) method, following common practices in existing literature, to mitigate endogeneity in Model (6) and Model (7). Instrumental variables are selected based on relevance and exogeneity criteria, with overidentification tests (Sargan–Hansen) confirming validity [54,55]. Notably, no endogeneity tests are conducted for the carbon emission intensity (CEI) model, as the coefficient of the core explanatory variable (fdiq) on CEI remains consistently negative and statistically insignificant across both baseline regressions and robustness checks.

On the one hand, we employ the ratio of total railway and highway mileage to administrative area as a measure of regional transportation network density (iv1) to instrument for fdiq in the GTFP and EC₁ regressions. The transportation capacity and accessibility are crucial determinants for multinational corporations’ investment decisions, while being relatively exogenous to host countries’ green total factor productivity and resource allocation efficiency.

In the regression results, the Kleibergen–Paap rk LM statistic reports the results of the under-identification test. The null hypothesis is that the instrumental variables are under-identified, and the condition for passing the test is that the p-value of the statistic is less than 0.1. The Kleibergen–Paap Wald rk F statistic reports the results of the weak instrumental variables test, and the condition for passing the test is that the value of the statistic is greater than 10. When the number of instrumental variables is greater than the number of endogenous variables, the Hansen J statistic is used for the over-identification test, and the condition for passing the test is that the p-value of the statistic is greater than 0.1. Finally, an endogeneity test is conducted, and the condition for passing the test is that the p-value is less than 0.1 [56].

The regression results are presented in

Table 10. Results of endogeneity tests based on the two-stage least squares method.

	(1)	(2)	(3)	(4)	(5)	(6)
	fdiq	GTFP	fdiq	EC1	fdiq	HC1
Iv1	0.5748 ***		0.5748 ***
	(0.1782)		(0.1782)
iv2					0.0476 *
					(0.0252)
fdiqm					0.4942 ***
					(0.0722)
fdiq		0.1105 **		0.0769 **		0.0714 *
		(0.0451)		(0.0325)		(0.0373)
controls	YES	YES	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES	YES	YES
Industry FE	YES	YES	YES	YES	YES	YES
Province FE	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
Kleibergen–Paap Wald rk F-statistic	10.40		10.40		31.54
Kleibergen–Paap rk LM p-value	0.0480 **		0.0480 **		0.0360 **
Hansen J-Test	—		—		0.6762
Endogeneity test p-value	0.0030 ***		0.0129 **		0.0867 *
Obs	1343		1343		1037

Note: Robust standard errors in brackets. * p < 0.1, ** p < 0.05, *** p < 0.01.

, columns (1) to (4). Columns (1) and (3) show that in the first stage, iv1 is significantly positively correlated with fdiq at the 1% level, indicating that regions with stronger transportation infrastructure tend to attract higher-quality FDI, consistent with our hypothesis. The weak instruments tests in columns (2) and (4) show that the Kleibergen–Paap Wald rk F-statistic exceeds 10 (10.40), rejecting the null hypothesis of weak instruments. The under-identification test yields a Kleibergen–Paap rk LM statistic with a p-value of 0.0480, significant at the 5% level, confirming the validity of our instrument selection. In the second stage, the coefficients of fdiq remain significantly positive at the 5% level. The endogeneity tests in the second stage produce p-values of 0.0030 and 0.0129, significant at the 1% and 5% levels, respectively, confirming the endogenous nature of the explanatory variables.

On the other hand, for HC₁ analysis, transportation accessibility (iv1) was deemed inappropriate due to potential endogeneity through labor mobility effects. We therefore employed industry-level carbon emissions (iv2), measured as the natural logarithm of sectoral CO₂ emissions. Higher industry pollution levels incentivize green technology spillovers from foreign firms while remaining exogenous to regional human capital development. We also adopted the industry median FDI quality (fdiqm), following established ESG research practices, as a higher-level instrumental variable [57].

Results in Table 10 columns (5)–(6) show: First-stage coefficients for iv2 and fdiqm are significant at the 10% and 1% levels, respectively. The Kleibergen–Paap Wald rk F-statistic (31.54) exceeds the critical threshold, and the Kleibergen–Paap rk LM test shows 5% significance. The second-stage fdiq coefficients maintain baseline significance levels, and the endogeneity test p-value indicates 10% significance. Hansen J statistic (p = 0.6762) fails to reject exogeneity of the instruments. This confirms the validity of our alternative instrumentation strategy.

5.5. Heterogeneity Analysis

First, we explore the heterogeneity of the impact of FDI quality on sustainable development in these two types of industries by dividing the industries into manufacturing and modern services, and non-manufacturing and modern services. The regression results are shown in columns (1) and (2) of

Table 11. Results of industry heterogeneity tests.

	(1)	(2)	(3)	(4)	(5)	(6)
	Manufacturing and Modern Services	Non-Manufacturing and Modern Services	Reverse-Engineerable Industries	Non-Reverse-Engineerable Industries	High-Polluting Industries	Non-High-Polluting Industries
fdiq	0.0035 **	−0.0013	0.0020 **	0.0047	0.0123 *	0.0022 *
	(0.0017)	(0.0028)	(0.0006)	(0.0028)	(0.0057)	(0.0011)
controls	YES	YES	YES	YES	YES	YES
Firm FE	YES	YES	YES	YES	YES	YES
Industry FE	YES	YES	YES	YES	YES	YES
Province FE	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
Obs	1237	98	452	889	268	1074
adj. R2	0.8337	0.9009	0.8892	0.8280	0.8644	0.8450
Intergroup Difference Test	0.002 ***		0.180		0.004 ***

Note: The p-value of the intergroup difference test was calculated by Fisher’s combination test (1000 samples) * p < 0.1, ** p < 0.05, *** p < 0.01.

. The regression results show that high-quality FDI into manufacturing and modern service industries significantly contributes to sustainable development at the 5% level; high-quality FDI into non-manufacturing and modern service industries has a negative but insignificant effect on sustainable development. The above empirical results show that manufacturing and modern service industries, as the main force of foreign investment in China, usually have mature production models and a well-established base of industries, technologies, and talents, and are therefore more capable of absorbing and transforming tangible and intangible resources carried by high-quality foreign investment. For other industries such as resource-intensive industries, labor-intensive services, or primary processing industries, due to the lack of sufficient knowledge and technology suitability, the entry of high-quality foreign capital is more likely to produce a crowding-out effect, oppressing the survival of local enterprises.

Second, the general equipment manufacturing industry, special equipment manufacturing industries, computer, communication, and other electronic equipment manufacturing industries, instrumentation manufacturing industry, metal products, machinery, and equipment repair industries are identified as reverse-engineerable industries, while other industries are defined as non-reverse-engineerable industries, so as to further observe the industrial heterogeneity of the effect of foreign direct investment quality on sustainable development. According to columns (3) and (4) of Table 11, in the industry prone to reverse engineering, the estimated coefficient of fdiq is 0.0020, which significantly promotes sustainable development at the 5% level; in the industry that is not easy to reverse-engineer, the estimated coefficient of fdiq is larger than 0.0047, though the effect on sustainable development is statistically insignificant. In addition, the p-value of the intergroup difference test was 0.180, indicating that there was no significant difference in the promotion of sustainable development by fdiq in the two industries. This study believes that the reasons for this phenomenon can be roughly attributed to the following two points. First, even if local enterprises in the industry that is not easy to be reverse engineered cannot fully enjoy the knowledge and technology spillover generated by high-quality multinational companies through normal channels, the “soft information” they can use, which cannot be accurately transmitted by negative contact, has essentially higher value attributes [58]. Theoretically, it has a stronger positive effect on sustainable development, so the estimation coefficient of fdiq is relatively larger. Second, the speed of local enterprises in industries that are not easy to be reverse engineered to expand their scale or technological innovation through their own efforts is obviously slower than that of technological imitation directly using the existing resources carried by high-quality multinational companies, and this speed is further reflected in the significance of the estimation coefficient.

Finally, industries are classified into high-polluting and non-high-polluting industries to further analyze whether there is any heterogeneity in the contribution of FDI quality to sustainable development in these two types of industries. The regression results are shown in columns (5) and (6) of Table 11. According to the regression results, in the high-polluting industry, a 1% increase in fdiq causes an average increase of 1.23 percentage points in the core explanatory variable GTFP, and the positive effect is significant at the 10% level; in the non-high-polluting industry, the positive effect of fdiq on GTFP is also significant at the 10% level, but the estimated coefficient is small at 0.0022, and the test of intergroup difference p-value similarly confirms that the positive impact of fdiq on sustainability is significantly heterogeneous across these two types of industries. The empirical results suggest that high-quality multinationals entering high-polluting industries are more committed to reducing pollution and carbon emissions than other industries to overcome the outsider disadvantage caused by their environmental characteristics [59,60], and thus generate more green technology spillovers to influence local firms in the industry to implement the host country’s low-carbon initiatives.

5.6. Moderating Effect Analysis

5.6.1. Analysis of the Moderating Effect of the Synergy Between the Government and the Market

The degree of government-market synergy (gov) variable is introduced to examine how it moderates the positive effect of the quality of FDI on resource allocation efficiency and distinguishes regional heterogeneity. The specific regression results are shown in columns (1) and (2) of

Table 12. Analysis of the moderating effect of the degree of collaboration between the government and the market.

	(1) EC1	(2) EC1
	Eastern Coastal Provinces	Non-Eastern Coastal Provinces
fdiq × gov	0.0011 **	−0.0029
	(0.0005)	(0.0050)
controls	YES	YES
Firm FE	YES	YES
Industry FE	YES	YES
Province FE	YES	YES
Year FE	YES	YES
Obs	1160	180
adj. R2	0.7599	0.3851

Note: Before generating the interaction term, the core explanatory variables and regulatory variables are centralized; Robust standard errors in brackets. ** p < 0.05.

.

The regression results show that, in eastern coastal provinces, the estimated coefficient of the interaction term between fdiq and gov is significantly positive at the 5% level. This indicates that governments in these regions leverage market mechanisms through decentralization, enabling high-quality multinationals to exert positive externalities and optimize resource allocation by directing regional resources to flow into high-value segments. In non-eastern coastal provinces, the estimated coefficients of the fdiq × gov interaction terms are negative and insignificant. This may be because the market mechanism in China’s central, western, and northeastern regions is not yet sound, and the entry of individual high-quality foreign capital could lead to overly concentrated local factor inputs, exacerbating resource mismatches. These findings indicate that, to a certain extent, China’s central, western, and northeastern regions still require government guidance to unlock the potential positive role of FDI enterprises in resource allocation.

Therefore, H3 of this study is only verified in the eastern coastal provinces.

5.6.2. Analysis of the Moderating Effect of Intellectual Property Protection Level

By introducing the intellectual property protection level (ipp) variable, we examine how it moderates the positive effect of the quality of foreign direct investment on the level of human capital and differentiates industrial heterogeneity based on the difficulty of knowledge and technology spillovers. The specific regression results are shown in columns (1) and (2) of

Table 13. Analysis of the moderating effect of the level of intellectual property protection.

	(1) HC1	(2) HC1
	Reverse-Engineerable Industries	Non-Reverse-Engineerable Industries
fdiq × ipp	0.1746 **	−0.1840 *
	(0.0788)	(0.0939)
controls	YES	YES
Firm FE	YES	YES
Industry FE	YES	YES
Province FE	YES	YES
Year FE	YES	YES
Obs	452	889
adj. R2	0.9925	0.9870

Note: Robust standard errors in brackets. * p < 0.1, ** p < 0.05.

.

According to the regression results, the estimated coefficients of the fdiq × ipp interaction terms are significantly positive at the 5% level in industries susceptible to reverse engineering, i.e., indicating that stronger intellectual property rights (IPR) protection in these industries enhances the positive effect of high-quality FDI on host-country human capital. In contrast, in industries not susceptible to reverse engineering, the interaction term coefficients are significantly negative at the 10% level, indicating that heightened IP protection diminishes the positive impact of high-quality FDI on human capital development.

This dichotomy may stem from two mechanisms. First, in reverse-engineerable industries, local firms can more easily replicate technologies due to open knowledge dissemination channels. Stronger IPR protection safeguards technology transfer and cooperation initiated by multinationals, boosting their willingness to engage in personnel, knowledge, and technology exchanges while protecting their competitive edge. Second, in non-reverse-engineerable industries, local enterprises have limited capacity to leverage multinationals’ demonstration effects for talent attraction and innovation. Stringent IP protection further restricts their ability to absorb and transform knowledge and technology spillovers from high-quality FDI.

Based on this, H4 of this study is validated only in industries that are susceptible to reverse engineering.

5.6.3. Analysis of the Moderating Effect of Environmental Regulation

Third, even though the current negative impact of FDI quality on the level of carbon emissions is not significant, we still explore whether the degree of environmental regulation (er) variable can play a certain degree of moderating effect on this negative impact by introducing it. The regression results are shown in column (1) of

Table 14. Analysis of the moderating effect of the degree of environmental regulation.

	(1) CEI
fdiq × er	−39.9134 **
	(19.5996)
controls	YES
Firm FE	YES
Industry FE	YES
Province FE	YES
Year FE	YES
Obs	1343
adj. R2	0.9394

Note: Robust standard errors in brackets. ** p < 0.05.

.

From the results, the estimated coefficient of fdiq × er is negative and significant at the 5% level, i.e., stricter environmental regulation strengthens the negative relationship between FDI quality and carbon emissions. This indicates that the Chinese government, through building a robust institutional framework in the environmental field, formulating strict regulations, and enhancing oversight of foreign investment’s production processes, can effectively encourage low-carbon practices.

Specifically, by strengthening environmental regulation through institutional construction and rigorous rule-making, the government helps high-quality foreign investors integrate low-carbon development into their strategic planning and operations, promoting the adoption of green practices. This synergy between policy and foreign investment facilitates the simultaneous advancement of economic growth and environmental protection.

Based on the above analysis, H5 of this study is empirically verified.

6. Conclusions, Policy Recommendations, and Outlook

6.1. Summary of the Findings

At present, the world is entering a period of turbulence and change, with old contradictions and new risks intertwined and superimposed, and economic and social development is facing “great changes, great tests and great cooperation”, making sustainable development the “golden key” to solving global problems. As a comprehensive transfer process of capital, technology, experience, and personnel, FDI, with its qualitative characteristics, contains great potential to promote sustainable development and is an important force to help China’s economy and the world economy prosper together. Therefore, this study takes the new development stage of China as the context for the establishment of the indicator system, takes the five clusters proposed by OECD’s FDI Quality Indicators 2022 as the guiding framework for the establishment of the indicators, and selects a total of 12 indicators to construct micro-level FDI quality indicators, and uses the period of 2011–2022 as the period for the establishment of the indicator system. Taking the data of China’s A-share listed foreign enterprises from 2011 to 2022 as the research sample, the empirical impact of FDI quality on sustainable development was examined using a fixed-effects model, and a series of robustness and endogeneity tests were conducted. In addition, corresponding heterogeneity analyses were conducted based on industry characteristics. Further, the moderating performance of the degree of government-market synergy, the level of intellectual property protection, and the degree of environmental regulation in the impact of FDI quality on sustainable development is examined in detail in the extended analysis. Through the above research, the study draws the following conclusions:

First, the empirical analysis shows that the improvement of FDI quality will have a positive impact on China’s realization of sustainable development. Further, with respect to the dependent variables of sustainable development selected in this study, improving the quality of FDI will significantly alleviate China’s resource mismatch, raise human capital levels, and potentially have a positive impact on promoting the low-carbon transition. It is worth noting that most studies on the construction of FDI quality indicators at the macro level show that FDI quality has a significant positive effect on the environment and carbon emissions [47,61], which may indicate that the environmental performance of FDI quality indicators is overestimated at the macro level.

Second, the results of heterogeneity analysis show that the impact of FDI quality on sustainable development is highly correlated with the industry of FDI inflow and quality multinational corporations located in manufacturing and modern service industries and high-pollution industries play a more significant role in promoting sustainable development. It should be noted that there is no significant difference in the promoting effect of high-quality foreign capital on sustainable development between industries that are vulnerable to reverse engineering and those that are not. This is somewhat inconsistent with the research findings of Chinese scholars. This discrepancy may be attributed to the fact that Chinese scholars have mostly constructed the quality index of FDI from a macro perspective, and they have predominantly focused on the impact of FDI quality on corporate innovation [62]. However, corporate innovation is merely a small part of sustainable development.

Third, the results of the extended analysis show that, on the one hand, the full play of market mechanisms is conducive to the optimization of resource allocation by high-quality FDI in China’s eastern coastal provinces, while non-eastern coastal provinces need to rely on the government to prevent high-quality multinational corporations from exacerbating the mismatch of resources. On the other hand, the improved level of IPR protection will promote the accumulation of human capital of local enterprises by high-quality multinational companies in the reverse-engineerable industries but will have a negative impact on non-reverse-engineerable industries.

6.2. Policy Recommendations

The current international situation is complex and volatile, and cross-border investor considerations for risk diversification are obviously on the rise. Competition among countries to attract investment is becoming increasingly fierce, and the new stage of China’s development has also set higher demands for the quality of foreign investment. Focusing on revitalizing the stock of high-quality foreign investment and actively attracting new high-quality foreign investment is particularly important to help China build a new development pattern and activate economic development in the era of great power competition. Therefore, based on the above findings, this study puts forward the following policy recommendations:

Establishing a micro-level foreign investment quality evaluation system to dynamically assess the quality of FDI. Through the formation of a cross-sectoral and cross-field professional team, we will further refine the requirements for FDI enterprises in terms of technological innovation, skills training, job quality assurance, gender equality, and environmental responsibility, and formulate a comprehensive and more geographically specific foreign-funded investment quality evaluation index framework. This ensures that the evaluation dimensions cover not only the economic contribution of FDI enterprises but also their performance in technological innovation and fulfillment of responsibilities in the value chain, enabling a more scientific and effective assessment of the quality of foreign investment stock and incremental investment. In addition, a quality dynamic monitoring platform is built by fully combining big data, cloud computing, and other cutting-edge technologies, realizing real-time access to data on production and operation, marketing, R&D inputs and other aspects of multinational corporations stationed in the country, and carrying out visual and quantitative analyses of their quality based on the established framework, so as to give timely feedback on the dynamic changes in the quality dimensions of foreign investment, with a view to providing a powerful basis for the government’s adjustment of policies on the attraction and utilization of FDI.

Create a new pattern for attracting and utilizing foreign direct investment in the new era and continue to contribute to high-quality development. In view of the positive role of quality enhancement of FDI in sustainable development, we should continue to build the comprehensive advantages of China’s domestic market, shape a market-oriented, rule-of-law, and internationalized business environment and effectively safeguard the expectations of high-quality FDI in terms of market openness, a sound system, and return on profits, so as to stimulate the potential of high-quality foreign investment in empowering high-quality development of the economy. On the one hand, build a high-level domestic opening platform, deeply create an all-field, multi-dimensional, and high-level open highland, further improve the level of opening up of the modern service industry, attract more high-quality foreign investment to technology-intensive services, and continuously increase the service guarantee for key FDI enterprises and projects in the manufacturing industry, especially to ensure the stock and incremental amount of foreign investment in the high-tech manufacturing industry and to accelerate the shaping of new quality productivity. On the other hand, continue promoting the high-quality development of the “Belt and Road” and further implement RCEP, CPTPP, DEPA, and other multi-bilateral cooperation mechanisms. Try to establish institutional dialogues with economies to form a stable and predictable international capital market cooperation mechanism, deepen the docking of strategies, plans, and mechanisms, and strengthen the interoperability of policies, rules, and standards, so as to release positive signals of opening up to high-quality FDI and allow foreign investors to see China’s strength and potential for development, thus increasing their investment in China.

Adhere to a rational distribution based on local conditions and revitalize existing high-quality foreign investment resources. Given that the impact of the quality of foreign direct investment on sustainable development varies according to the basis of attracting capital in different regions or the channels of technological spillover in different industries, it is necessary to “tailor-make policies” according to the characteristics of regions and industries, so as to better utilize its positive role in the efficiency of resource allocation, human capital, and low-carbon development while retaining high-quality foreign capital stock. On the one hand, through in-depth research on the distribution of foreign investment in the region, combined with the characteristics of local industries, resource endowment, and development planning, we draw an accurate blueprint for the layout and utilization of foreign investment. On the other hand, in view of the heterogeneity of the impact of the quality of FDI on sustainable development caused by intellectual property protection based on industry characteristics, we have accelerated the construction of a multi-level system of laws and regulations on intellectual property protection. Specifically, firstly, on the basis of the existing laws, the regulations on the protection of technological innovation achievements should be further sorted out and refined, and the infringement recognition standards and punishment should be clarified, so as to provide solid legal protection for the technology input and technology output of foreign enterprises, and enhance their sense of security, trust, and access to technology exchange and cooperation in China. Secondly, the strategy of intellectual property protection should be moderately optimized, and a differentiated protection mode should be adopted to differentiated protection mode, and under the premise of ensuring the security of core technology secrets of FDI enterprises, guide them to open up part of their non-critical technologies conditionally, and broaden the learning channels of local enterprises’ knowledge, technology, experience, and management modes—for example, through technological licensing and R&D cooperation to achieve moderate sharing of technological resources. This ensures that local enterprises have opportunities to obtain advanced resources and accumulate industry-specific human capital.

6.3. Research Outlook

In response to the changes in the international and domestic economic situation, this article constructs, for the first time, an indicator system of FDI quality at the micro level and explores in depth the deep-rooted links between FDI quality and sustainable development. However, due to the limitations of research capacity, research data, and research time, the article may have the following shortcomings, which need to be further addressed and improved in the future:

First, while this study takes A-share listed FDI enterprises in China from 2011 to 2022 as the research sample and carries out the construction of the indicator system and subsequent empirical analysis, and the conclusions drawn from this have a certain degree of verifiability and inspiration, there is a certain gap compared with the real situation. On the one hand, limited by the availability of data, this study is unable to obtain the data for all FDI enterprises in China, so the analysis of the reality of the quality of FDI may have a certain bias. On the other hand, constrained by the lag in macro statistics, this study cannot extend the sample period, making it difficult to timely reflect the detailed status, developmental trajectory of China’s attraction and utilization of foreign investment, and the progress of sustainable development. In addition, due to the fact that most of the current FDI enterprises listed on China’s A-share market are distributed in the eastern coastal regions of China, there are still certain difficulties in conducting clustering at the provincial level to study the spatial effects of the quality of FDI.

Secondly, even though this study has conducted a more comprehensive mechanistic discussion of the relationship between FDI quality and sustainable development in the theoretical analysis and further explored the coupling mechanism, there are often complex interconnections between economic entities, which makes it impossible to exhaust the impact mechanisms embedded in this study. Therefore, although this study conducts corresponding empirical tests for the theoretically analyzed influencing mechanisms and puts forward policy recommendations based on the empirical results, there may be unaccounted-for omissions in the research conclusions, and it is yet to be verified whether the proposed recommendations can be fully applied to the current context.