More market-based tools, less supply chain vulnerability: China’s rare earths policy reforms and their impact on the global supply chain

Abstract

Rare earth elements (REEs) are critical metals for clean energy technologies. All countries face urgent challenges in mitigating supply chain risks and stabilizing the global REEs supply. As a major supplier within the global REEs supply chain, China’s market-based policies targeting market participants exert substantial impacts on supply chain security. This study adopts the Autoregressive Distributed Lag model to analyze how China’s market-based REEs policies affect global supply chain vulnerability. The empirical results show that such policies can alleviate supply chain vulnerability. A greater number of market-based policy tools slightly lowers vulnerability (coefficient = −0.02, p < 0.1). A higher proportion of market-based tools significantly curbs vulnerability in the short term (coefficient = −4.56, p < 0.01), yet the long-term effect (coefficient = −1.25) lacks statistical significance. This study puts forward policy adjustments to improve policy efficiency and further reduce supply chain vulnerabilities.

1 Introduction

The vulnerability of the rare earth elements (REEs) supply chain is a key measure of REEs supply security. As essential materials for strategic sectors like clean energy and advanced manufacturing, stable REEs supply critically impacts the safety and sustainable development of new energy vehicles, wind turbines and other related industries. Research identifies supply constraints and price volatility as the main factors affecting supply stability (Jowitt, 2022). Global primary REEs resources—minerals extracted directly from primary deposits—are mainly concentrated in a few countries, especially China, Vietnam, Brazil, Russia, and India, which together hold over 90% of the world’s total REEs reserves (United States Geological Survey (USGS, 2024). Currently, only a few countries, including China, can produce and export REEs-related products on a large scale (Guo and Mai, 2024). Conventional mining and smelting technologies are characterized by high energy consumption, significant environmental risks, and low operational efficiency. These issues stem from the depletion of concentrated primary resources and the legacy of unregulated mining activities. Thus, countries engaged in REEs extraction and production, such as China, place great emphasis on the sustainable development of REEs resources, with the goal of mitigating resource depletion and protecting ecosystems (Han et al., 2016; Balaram, 2019; Yan and Li, 2019; Hamed et al., 2024). These countries will not fully exploit their primary REEs reserves in the short term to meet global market demand. The supply of primary REEs resources is currently constrained, with few supplying countries, leading to concentrated sources (McLellan et al., 2013; Pawar and Ewing, 2022). If relevant nations change policies or face external disruptions like armed conflict, the risk of short-term supply interruptions may rise significantly (Schneider et al., 2014). The REEs recycling market faces challenges such as high operational costs and low recovery efficiencies, impeding the development of an efficient secondary resource supply system (Li et al., 2024). Concentrated primary resources and underdeveloped secondary supplies together create persistent supply pressures, increasing the vulnerability of the REEs supply chain. This concentration increases the risk of disruptions, heightening the vulnerability of the REEs supply chain.

Regarding price volatility, demand for REEs is rising in high-tech sectors, electric vehicle manufacturing, and renewable energy industries (Filho et al., 2023). Due to the concentrated supply of REEs, producers struggle to quickly adjust and expand production capacity to meet rising demand, causing supply-demand imbalance and price volatility (Ilankoon et al., 2022). Price volatility hinders new mining project development. While price increases can boost profitability, attract investment, and expand supply, supply disruptions may further elevate prices, complicating demand fulfillment. Conversely, when supply and demand stabilize, prices often drop, potentially rendering mining unprofitable and causing investors to pause projects (Riddle et al., 2015; García et al., 2017). Delayed capacity expansion means that existing output cannot offset demand surges or emergency supply disruptions caused by wars or earthquakes. Consequently, price volatility hinders the development of new projects, limits source diversification, and increases the vulnerability of the REEs supply chain.

Given the increasing vulnerability of the REEs supply chain, all countries involved in the REEs industry prioritize reducing supply risks. Europe is strengthening cooperation with the United States, Japan, and Australia to establish alternative supply chains and increase investment in REEs recycling and efficient utilization (Charalampides et al., 2015; Yan and Li, 2019). Europe is also exploring new primary sources of REEs and advancing recycling technologies to create supplementary resource pathways, reducing the vulnerability of the supply chain. The United States is expanding domestic REEs mining and production (Wang, 2023) to increase resource availability and reduce import dependence, enhancing supply chain resilience. Japan focuses on recycling waste and end-of-life products containing REEs, promoting alternative materials, and maintaining a comprehensive REEs reserve system (Barteková and Kemp, 2016; Eggert et al., 2016). Expanded utilization of secondary resources and mandatory stockpiling policies reduce Japan’s reliance on primary REEs and mitigate supply disruption risks, thereby decreasing the vulnerability of the REEs supply chain.

China, a key supplier and principal participant in the global REEs supply chain (Lee and Dacass, 2022), has implemented policies to reduce supply chain vulnerabilities. These include subsidies and tax incentives supporting domestic REEs recycling and substitute technology development, encouraging firms to engage in related research and application. Yan and Li (2019) emphasize that increased investment in substitutes and recycling helps meet rising demand, extends REEs reserves, and lowers shortage risks. Other scholars note that advancing substitution technologies and recycling systems can alleviate REEs supply pressures, diversify sources, and reduce supply chain vulnerabilities (Keilhacker and Minner, 2017; Lai et al., 2024). This study refers to such China-issued policies targeting market entities as “market-based policies”.

Existing scholarly literature on China’s REEs policies primarily focuses on the impact of China’s initial export restrictions on the vulnerability of the REEs supply chain and the fundamental policy objectives. Due to the strategic importance of REEs and many countries’ reliance on China’s resources, concerns arose over export quotas limiting REEs availability, increasing the risk of supply disruptions and the vulnerability of supply chains (Mancheri, 2015; Zhang et al., 2024). Scholars generally agree that China’s initial export restriction policies were primarily implemented to address challenges such as the overexploitation of REEs resources, severe environmental damage, smuggling, and pricing issues, rather than geopolitical leverage (Hayes-Labruto et al., 2013; Wübbeke, 2013; Zhang et al., 2015; Ma and Li, 2021). Following China’s defeat in the World Trade Organization (WTO) dispute concerning REEs, export restriction policies were subsequently rescinded. Other studies have analyzed the specific content of REEs policies, focusing on their evolution, comparisons between local and central government policies, and the underlying policy objectives. Notable examples include studies on the evolution of REEs policies (Shen et al., 2020), examinations of environmental regulations concerning REEs (Chai et al., 2020; Pan et al., 2021), and analyses related to the consolidation of the REEs industry (Barteková and Kemp, 2016; Han et al., 2016). However, these studies predominantly overlook the role of China’s market-based policies.

China has proactively tackled challenges in the REEs sector by expanding market-based policy tools, thereby effectively reducing the vulnerability of the REEs supply chain. Yan and Li (2019) argue that enhancing innovation capacity in the REEs industry is more crucial than merely regulating production and export quotas. Faced with insufficient incentives for innovation and urgent demands for industrial upgrading, the government has intensified the implementation of market-based policies. These policies are guided by the government but primarily focus on market participants. The proportion of these policies in annual policy tools is shown in Figure 1. Although China has long dominated upstream and midstream REEs activities (e.g., mining, smelting), separation and purification efficiency—especially for samarium (Sm) and other middle lanthanides—remains limited. Additionally, efforts to enhance precision pollution control in mining and processing are ongoing (Zhang et al., 2022; Lei et al., 2025). Targeted R&D funding and subsidies are implemented to drive corporate technological innovation. This strategy stabilizes supply, reduces environmental pollution, and strengthens China’s industrial advantages in technology, industry, and policy, thereby decreasing the vulnerability of the REEs supply chain (see Section 2.1).

FIGURE 1

This study aims to analyze whether China’s market-based REEs policies reduce the vulnerability of the global REEs supply chain. It quantifies these policies by enumerating the types and quantities of market-based tools China employs. Since policy impact depends on specific tools, classifying and quantifying these tools enables assessment of their impact on the vulnerability of the global REEs supply chain. The research framework conceptualizes China’s market-based REEs policies using two variables: Market-based Policy Tools (), which represents the absolute number of these policies, and Market-based Policy Tools Proportion (), which indicates their relative share among all REEs policies. This study employs econometric modeling, specifically adopting the Autoregressive Distributed Lag (ARDL) model for empirical analysis. Based on the empirical results, this study further proposes recommendations for implementing market-based policies and strategies to reduce the vulnerability of the global REEs supply chain.

The study is organized as follows: Section 2 formulates the theoretical hypotheses; Section 3 introduces the data and models adopted; Section 4 conducts a detailed analysis of the empirical results; Section 5 is the discussion; Section 6 summarizes the study and offers recommendations.

2 Theoretical mechanisms and research hypotheses

2.1 Theoretical mechanisms

2.1.1 The vulnerability of the global REEs supply chain

Supply chain vulnerability refers to the likelihood of negative outcomes in a supply chain. This occurs when the combined influence of risk sources—fundamental causes that may trigger risks and lead to losses, categorized into three categories: supply-side, demand-side, and catastrophic—and risk drivers—conditions, behaviors, or variables that convert potential risks from these sources into actual risks—exceeds the effectiveness of risk mitigation measures (Wagner and Bode, 2006; Bode and Wagner, 2015).

This classification offers a universal theoretical framework applicable to all industrial supply chains. To tailor this framework to the REEs industry, the following section aligns risk sources, risk drivers, and mitigation strategies with three distinctive features of the REEs supply chain, thereby operationalizing this theoretical model for REEs research.

Firstly, the uneven global distribution of REEs resources creates significant supply-side risks. REEs mining and smelting capacities are limited by natural reserves and are geographically concentrated in a few regions, which easily leads to supply imbalances. This concentration of mining, mineral processing, and refining capacity is a primary risk factor that increases the vulnerability of the REEs supply chain, as reliance on a single or a few dominant suppliers amplifies the risk of supply disruptions. Additionally, geopolitical regulatory measures, such as export quotas and restrictions on overseas mineral investments, can cause sudden shocks to supply stability. Common risk mitigation strategies include establishing national strategic REEs stockpiles, diversifying overseas mineral sources, and expanding recycling systems to reduce dependence on single suppliers. However, these solutions require long development cycles and substantial capital investment, limiting their effectiveness in alleviating short-term supply volatility caused by capacity concentration.

Secondly, geopolitical conflicts, export restrictions, and technological blockades are major triggers of severe supply chain disruptions in the REEs sector. As strategically critical raw materials, REEs are highly sensitive to geopolitical policy shocks, which directly disrupt supply chains and hinder downstream manufacturing. Capacity concentration and reliance on a single upstream source act as risk amplifiers, turning minor geopolitical frictions into large-scale industrial supply breakdowns. Corresponding countermeasures include transnational joint REEs resource development and multilateral mineral trade agreements to mitigate external supply shocks caused by geopolitical factors.

Thirdly, heavy pollution generated by REEs mining, mineral separation, and smelting constitutes a significant supply-side risk. The entire REEs production process produces substantial pollutants. To comply with ecological protection requirements, countries have implemented output limits and production reduction policies, directly restricting the marketable supply of REEs. Increasing environmental governance costs and stringent environmental inspections impose additional constraints, exacerbating supply volatility. To mitigate these environmental risks, stakeholders are promoting green smelting technologies and full-chain clean production methods to balance ecological conservation with a stable REEs supply.

In summary, based on the aforementioned theoretical framework and industry characteristics, the vulnerability of the REEs supply chain is determined by the interaction among risk sources, risk drivers, and mitigation measures. When the combined impact of REEs risk sources and drivers exceeds the effectiveness of mitigation measures, the vulnerability of the global REEs supply chain increases. Conversely, if mitigation measures deliver stronger restraining effects, the vulnerability of the global REEs supply chain decreases.

2.1.2 Market-based policy tools

Given China’s central role as a leading participant and dominant supplier in the global REEs supply chain, along with the policy-driven nature of its REEs industry development, China’s policies can act as either risk drivers or risk mitigators by reshaping risk sources and thereby influencing the vulnerability of the global REEs supply chain. For example, China’s early export quota system, which imposed fixed annual export caps rather than allowing open trade, restricted the global supply of REEs and served as a significant risk driver. In the early stages, China’s REEs prices were well below the production costs of other countries, fostering heavy foreign dependence on Chinese REEs with limited alternative suppliers. This high supply dependency created substantial risks of supply fluctuations and ultimately increased the vulnerability of the global REEs supply chain.

Policies are formulated to address practical challenges within the industry, and policy tools represent the specific instruments and mechanisms used for policy implementation, attracting extensive academic attention. Drawing on the policy tool framework proposed by Howlett et al. (2009) and considering the developmental characteristics of the REEs industry, China’s REEs policies can be categorized into mandatory, mixed, and voluntary tools based on the degree of government intervention. These tools typically operate synergistically as integrated policy mixes. Aligned with practical industrial development goals, China’s REEs policy implementation follows two primary approaches: mandatory tools supplemented by voluntary ones for resource regulation, and mixed tools supplemented by voluntary ones for market-oriented industrial development. This classification and operational logic provide a theoretical foundation for analyzing how market-based policy tools influence the vulnerability of the global REEs supply chain.

Combining supply chain vulnerability theory with policy tool theory, this study develops a clear transmission mechanism: policies act as either drivers or mitigators of risk sources, thereby reshaping supply chain vulnerability. Furthermore, this study categorizes China’s market-based REEs policy tools into three dimensions—technology, industry, and policy incentives—to examine their specific impacts on the vulnerability of the global REEs supply chain.

In the technological dimension, China’s market-based REEs policies emphasize industrial-chain technological innovation to mitigate risks on both the supply and demand sides, thereby reducing the vulnerability of the global REEs supply chain. Advanced mining, separation, and refining technologies enhance production efficiency and stabilize the supply of primary resources, alleviating supply-side risks. Innovations in substitution technologies help balance supply and demand, reducing demand-side pressures. Additionally, the development of recycling technologies supports the establishment of a circular REEs system and expands the supply of secondary resources, further decreasing supply-side vulnerability.

In the industrial dimension, market-based policies encourage multi-stakeholder cooperation and industrial upgrading to stabilize supply conditions and reduce the vulnerability of the global REEs supply chain. The development of industrial clusters enhances supply chain coordination and accelerates technology diffusion. An optimized industrial structure strengthens China’s pricing power, facilitating reasonable price adjustments, capacity expansion, and a stable resource supply. Improved recycling systems decrease dependence on primary resources and diversify supply channels, effectively mitigating supply-side risks.

In the policy incentive dimension, financial support, tax incentives, and talent incentives drive technological progress and industrial upgrading. These measures facilitate the effective implementation of technological and industrial policies, collectively reducing the vulnerability of the global REEs supply chain.

In 2011, the State Council of China promulgated Several Opinions of the State Council on Promoting the Sustained and Sound Development of the Rare Earth Industry, forming a top-level framework for China’s medium- and long-term REEs industry development. The principal market-based policy tools outlined in this document cover two critical dimensions: the technological and industrial dimensions. For the technological dimension, the policy advocates enterprise-led technological innovation and upgrading, promotes circular technologies for the recycling of tailings and REEs products, and expands diversified supply security systems via technological progress. It also targets breakthroughs in high-end REEs material technologies to stabilize the REEs supply chain. For the industrial dimension, the policy encourages market-oriented mergers and reorganizations of enterprises and phases out backward production capacity, thereby facilitating structural adjustment of the REEs industry and solidifying the sector’s position as a strategic basic industry. In addition, Nine Measures to Further Support the Development of New R&D Institutions in Baotou Rare Earth High-Tech Zone, issued by the Management Committee of the Baotou National Rare Earth High-Tech Industrial Development Zone in May 2019 and fully rolled out to market entities in 2020, serve as a representative special policy for incentive mechanisms. This policy provides financial grants, subsidies, and special funds for R&D centers to boost technological innovation. Meanwhile, it opens an administrative “green channel” and offers fee discounts to accelerate the settlement of enterprises and industrial projects in the zone, driving the development of the local REEs industry. Taken together, these two policy frameworks correspond to the three transmission channels of technology, industry, and policy incentives, and provide empirical policy evidence for the theoretical mechanisms proposed in this study.

2.1.3 Market-based policy tools proportion

To comprehensively assess the influence of market-based policy tools on the vulnerability of the REEs supply chain, this study aggregates the annual number of such tools employed within the REEs sector and calculates their proportion relative to the total number of REEs policy tools. This measure, hereafter referred to as the market-based policy tools proportion, directly reflects the government’s governance orientation and the intensity of its resource allocation. By adjusting the intensity and synergy of these tools across three dimensions—technology, industry, and policy incentives—it indirectly influences the vulnerability of the REEs supply chain.

Drawing on the policy tool framework proposed by Howlett et al. (2009), this study argues that a higher proportion of market-based policy tools reflects an increased emphasis on market-oriented approaches within the overall policy mix. This orientation mitigates the restrictive effects of mandatory regulations on market dynamism, enhances resource allocation for technological innovation, industrial upgrading, and policy incentives, and strengthens the synergy among these factors in reducing supply chain risks. Furthermore, the dynamic adjustment of the proportion of market-based policy tools aligns with the developmental stages of China’s REEs industry, thereby improving the precision of policy interventions targeting the sources of supply chain risk and contributing to a reduction in the vulnerability of the REEs supply chain.

Furthermore, the impact of increasing the proportion of market-based policy tools exhibits a temporal lag. In the short term, these tools prompt enterprises to rapidly adjust their decisions through targeted resource allocation, such as participating in REEs industrial parks or increasing investment in technological R&D, to mitigate localized supply chain risks. In the long term, sustained synergy among policy tools fosters technological breakthroughs in REEs and enhances the industrial ecosystem, thereby establishing a robust supply chain risk resilience framework and achieving lasting optimization of supply chain vulnerability.

2.2 Research hypotheses

Drawing on the theoretical analysis presented in Section 2.1, which explains how market-based policy tools influence the vulnerability of the REEs supply chain through technological, industrial, and policy incentive dimensions, this paper investigates the direct causal relationship between market-based policy tools (and their proportion) and the vulnerability of the REEs supply chain. Accordingly, the study proposes the following hypotheses.

2.2.1 Market-based policy tools

As illustrated in Section 2.1.2, China’s market-based REEs policy tools address vulnerabilities in the REEs supply chain through three core dimensions: REEs technology, the REEs industry, and market policy incentives. From a technological perspective, these policy tools promote three key areas of development: 1) innovative mining, separation, and smelting technologies; 2) research and development of alternative materials; and 3) low-cost, high-efficiency recycling technologies. These advancements mitigate supply-side risks by expanding available REEs resources and reduce demand-side risks by decreasing dependence on raw materials, thereby collectively lowering the vulnerability of the REEs supply chain. From an industrial perspective, market-based policies drive industrial restructuring and upgrading to achieve four objectives: 1) enhance China’s pricing power and market influence over REEs products; 2) stabilize REEs prices; 3) ensure the steady operation of existing projects and the initiation of new industrial programs; and 4) establish a comprehensive REEs recycling system. By expanding stable resource supply, these measures mitigate supply-side risks and reduce the vulnerability of the REEs supply chain. The policy incentive dimension underpins technological innovation and industrial transformation, supporting the implementation of the two aforementioned dimensions and further reducing the vulnerability of the REEs supply chain. In summary, China’s market-based REEs policies have a positive mitigating effect on the vulnerability of the global REEs supply chain. Therefore, this study proposes Hypothesis 1:

2.2.2 Market-based policy tools proportion

As shown in Section 2.1.2, an increasing share of market-based policy tools influences the vulnerability of the REEs supply chain by enhancing synergies among technological, industrial, and incentive policy tools. This impact involves a time lag: short-term resource reallocation partially mitigates supply chain risks, while long-term technological breakthroughs and optimized industrial ecosystems sustainably reduce the vulnerability of the REEs supply chain. Therefore, a higher proportion of market-based policy tools decreases the vulnerability of the global REEs supply chain in both the short and long term. The corresponding hypotheses are proposed as follows:

3 Model description and Data Sources

3.1 Model description

This study aims to examine the effects of China’s market-based REEs policy tools on the vulnerability of the global REEs supply chain. This study employs time-series data covering a 33-year period from 1991 to 2023. Consistent with established econometric methodologies, the analytical framework is specified as follows in Equation 1:

Within the model, denotes the vulnerability of the global REEs supply chain; denotes the constant term; denotes the research focus on market-based REEs policies, which is further divided into two variables: Market-Based Policy Tools () and Market-Based Policy Tools Proportion (). Additionally, denotes the concentration of REEs metal suppliers; denotes the sample variance of the Gross Domestic Product (GDP) of REEs metal importing countries; denotes geopolitical risk; and denotes the residual term of the model.

This study uses the Autoregressive Distributed Lag (ARDL) model for empirical analysis, which offers two main advantages. First, it accommodates variables that are non-stationary in their original form, unlike traditional time series methods that require differencing to achieve stationarity—often altering variable meaning and compromising results. The ARDL model allows variables integrated at different orders, provided they are stationary at level or first difference, excluding those needing higher-order differencing. Second, ARDL captures both short- and long-term dynamics between explanatory and dependent variables, enabling policymakers to assess immediate and lasting policy effects for more informed decisions.

This study develops a model of Market-Based Policy Tools () grounded in the research hypotheses previously stated and formally represented in Equation 2:

Among these, , , , ,and denote the optimal lag lengths for the dependent and independent variables, respectively.

The basic ARDL model representing the Market-Based Policy Tools Proportion () is explicitly formulated as shown in Equation 3:

Among these, , , , ,and denote the optimal lag lengths for the dependent and independent variables, respectively.

This study acknowledges that the ARDL model of has successfully passed the bounds test, with detailed results provided in Supplementary Appendix B of the supplementary materials. This transformation follows the standard derivation procedures of the Autoregressive Distributed Lag–Error Correction Model (ARDL-ECM). First, is subtracted from both sides of Equation 3, resulting in the first-difference dependent variable on the left-hand side. Next, all explanatory variables expressed in levels are decomposed into their first-difference short-term components and their corresponding lagged level terms. The lagged level terms of , , , , and are then combined to represent the long-term equilibrium relationship. After rearranging these terms, the intermediate form of the error correction model is obtained, as shown in Equation 4:

Here, Δ denotes the first-order difference of the variables. The coefficients , , , ,and are the fundamental components used to calculate the long-term multipliers, while , , , ,and denote the short-term effect coefficients. The parameters , , , ,and specify the optimal lag lengths for the dependent and independent variables, respectively.

To improve the clarity of the model coefficients, the study reformulated the relevant equation for the Market-Based Policy Tools Proportion () in the model (Kripfganz and Schneider, 2023), as explicitly presented in Equation 5:

Here, the long-term multipliers , , and are calculated as . (where ) denotes the adjustment speed and direction of the variable, as it converges to long-term equilibrium after being subjected to a short-term shock.

In determining the optimal lag order for the and models, this study employs the Akaike Information Criterion (AIC) alongside considerations of practical economic relevance. Given a sample size of 33 and to balance the trade-off between degrees of freedom loss and the ability to capture the dynamic properties of the variables, the search for lag orders is constrained to a range of one–four lags. In this process, guided by practical policy transmission logic, the lag of the core policy variable is exogenously fixed at one period, as policy interventions can rapidly influence market trade activities in the short term. In contrast, the lags of the dependent variable and other control variables are freely optimized within one–four periods to capture their gradual long-term adjustments in supply chains and trade patterns. From the perspective of economic reality, the upper bound of four lags is set to accommodate the slow adjustment characteristics of cross-border trade and global supply chain variables. Trade volumes, industrial linkages and other economic indicators require multiple periods to fully release their dynamic effects, which provides reasonable economic support for the 1–4 lag search interval. The econometric software systematically evaluates all feasible lag combinations under this constraint, identifying the optimal lag structure for the model as (4,1,4,2,3) and for the model as (4,1,4,3,3), based on the minimum AIC values. However, to avoid inconsistencies in lag orders between the two models—which could introduce specification confounds—this study standardizes the lag order for both models to (4,1,4,3,3), as illustrated in Table 1. Subsequent analysis indicates that, for the model, the signs and statistical significance of coefficients remain stable whether the lag order is set at (4,1,4,2,3) or (4,1,4,3,3). Furthermore, the model employing the (4,1,4,3,3) lag structure successfully passes relevant diagnostic and robustness checks, thereby supporting the appropriateness and reliability of this lag order selection.

3.2 Variables introduction and data sources

3.2.1 Dependent variable

The vulnerability of the REEs supply chain () is assessed using the export concentration variation index (Cui et al., 2022). This index is calculated based on the import and export countries of the relevant products and their corresponding trade data. This study focuses on REEs metals, which fall under HS code 280530. This study focuses on REEs metals (HS code: 280530) in the midstream segment of the REEs industrial chain for several reasons. First, the upstream segment, involving REEs ore (HS code: 253090), is excluded due to strict government regulations and the absence of a market-based mechanism governing mining and exploration in China, which conflicts with the study’s policy tool criteria. Second, the downstream sector includes diverse products like permanent magnet materials, phosphor materials, and hydrogen storage materials, each with different functions and varying dependence on midstream REEs metals. For example, permanent magnet materials rely heavily on REEs metals, unlike phosphor materials. This heterogeneity makes it difficult to capture the overall supply chain vulnerability by focusing on a limited subset of downstream products. Therefore, downstream products are also excluded. Since REEs metals are the fundamental raw materials for alloys and reflect China’s market-based REEs policies most clearly, they are chosen as the primary focus for assessing supply chain vulnerability.

The formula for evaluating the vulnerability of the REEs metal supply chain () involves a two-step procedure. First, the concentration index for country REEs metal exports in year is calculated, as specified in Equation 6:

Here, and denote the exporting country and importing country, respectively; represents the total number of countries importing REEs metals in year ; denotes the total volume of REEs metals imported by country from country in year ; and represents the total number of import sources for REEs metals of country in year . Therefore, in the weighted directed network, the concentration index of country corresponds to the total REEs metal exports of country , standardized by the average level of importing countries.

In the subsequent step, the standard deviation of the centrality indices for each country involved in the REEs metals trade was utilized as an indicator of REEs metals vulnerability. This metric, termed the centrality variation index, is formally defined in Equation 7:

Here, denotes the average value of the concentration indices (across different countries) for REEs metals in year ,and represents the number of exporting countries in year .

Data source: United Nations Commodity Trade Statistics Database.

This study employs the export concentration variation index to assess the vulnerability of the global REEs supply chain (). The primary rationale for using this index lies in its two-step network measurement approach, which integrates bilateral trade data between exporting and importing countries, thereby capturing the comprehensive interactive structure of the global REEs trade system. quantifies the overall trade influence of each supplier across all international import markets, while the index measures disparities in market power among supplying countries by calculating the standard deviation of their respective centrality values. REEs metals are strategically critical resources characterized by high technological processing barriers and inelastic short-term supply. The systemic risk inherent in their global supply chain predominantly arises from the uneven trade dominance among principal exporting nations. The index effectively incorporates both variations in supply-side power and the distributional characteristics of global import demand, aligning closely with the central research focus of this study. However, it is important to acknowledge that this unidimensional indicator reflects only structural imbalances within the trade network and does not capture multidimensional aspects of vulnerability, such as supply chain flexibility, market responsiveness, and recovery capacity following disruptions. These limitations suggest valuable avenues for future research.

3.2.2 Key explanatory variable

The primary explanatory variables are collectively denoted as , representing both the quantity and proportion of market-based REEs policy tools implemented in China in year . This study retrieves policy documents from the Wolters Kluwer Practical Law Database, identifying 234 valid central and local REEs policy texts (with the keyword ‘rare earths’) from 1991 to 2023. The screening of market-based policy tools follows three stages. First, using Howlett et al. (2009) classification, policies are categorized by government intervention level: mandatory, mixed, and voluntary. Second, focusing on promoting market-based REEs industry development—mainly through mixed and voluntary tools—and excluding purely government-led mandatory tools, the study targets mixed and voluntary policy tools. Third, a secondary verification ensures alignment with market-based development goals like technological innovation and industrial upgrading.

To address challenges arising from ambiguous policy classification boundaries and overlapping policy attributes, this study establishes explicit classification criteria and conducts a manual secondary verification of all policy samples. Specifically, mandatory policies are defined by stringent administrative controls, including measures such as total mining volume control, production suspension and rectification, and administrative penalties; these policies are entirely excluded from the current analysis. Mixed policies strike a balance between government guidance and market regulation, effectively reducing industry transaction costs and standardizing market order. Typical instruments include tax incentives, financial subsidies, and green channels for REEs enterprises—market-oriented interventions. Voluntary policies rely on the autonomous participation and self-regulation of market actors, thereby fully leveraging market initiative. These primarily involve the development of the REEs industry system, cultivation of key enterprises and projects, public welfare initiatives, and enterprise self-inspection. In practice, some policies incorporate both administrative constraints and market incentives, leading to classification ambiguities. To mitigate misclassification bias, this study uses the core policy objectives and principal implementation approaches as the primary criteria for manually identifying policies with overlapping attributes: policies predominantly characterized by mandatory administrative control are excluded, whereas those chiefly driven by technical, industrial, and policy incentive dimensions are included in the market-based policy sample. This approach ensures precise policy coding and the reliability of research results. A detailed classification framework is presented in Table 2.

uses two variables for quantitative analysis: the number of market-based policy tools (denoted as ) and the proportion of market-based policy tools (denoted as ). Specifically, the number of market-based policy tools () represents the annual frequency of market-based policy tools. The proportion of market-based policy tools () is defined as the ratio of the number of market-based policy tools to the total number of policy tools enacted in the same year.

Data Source: Wolters Kluwer Practical Law Database.

3.2.3 Control variables

Supplier concentration measures the degree of dependence on suppliers (Zhang et al., 2025) and is quantified using the Herfindahl-Hirschman Index (HHI), as shown in Equation 8:

Here, denotes the supplier concentration in year ; represents the export volume of REEs metals from country to the global market in year ; indicates the total global export volume of REEs metals in year ; and signifies the total number of countries involved in exporting REEs metals.

The Gross Domestic Product (GDP) demand index is calculated using the GDP data of countries importing REEs metals in year . This index is determined by computing the square root of the sample variance, as shown in Equation 9:

Here, denotes the GDP demand index for year ; denotes the GDP of the importing country in year ; denotes the average GDP of all importing countries in year ; and denotes the total number of importing countries in year .

This study uses geopolitical risk data from the Geopolitical Risk Index by Caldara and Iacoviello (2018), originally monthly. For analysis, monthly data were averaged annually following Lai et al. (2023), as shown in Equation 10:

Here, denotes the Geopolitical Risk (GPR) index for the year , while denotes the GPR index for the month within the year .

Data Source: https://www.matteoiacoviello.com/gpr.htm.

4 Empirical results

The prerequisites for using the ARDL model—unit root tests (Supplementary Appendix A), analysis of long- and short-term relationships for the proportion of market-based policy tools (Supplementary Appendix B), and model diagnostics tests, robustness checks, endogeneity discussion, and stability tests (Supplementary Appendices C,D)—are detailed in the appendices. Supplementary Appendices A-D can be found in the supplementary materials. This section presents the statistical descriptions and empirical results of the ARDL model.

4.1 Descriptive statistics

Table 3 presents the principal descriptive statistics for the variables included in this model. These variables comprise the vulnerability of the REEs metal supply chain (), China’s market-based policy variables related to the REEs sector—specifically, market-based policy tools () and their proportional representation ()—as well as REEs metal supply concentration (), GDP demand index (), and geopolitical risk ().

Table 3 shows that the mean vulnerability of the REEs metal supply chain () is 20.45, with a standard deviation of 4.36. Figure 2 illustrates a steady increase in the supply chain vulnerability index from 8.74 in 1991 to 27.01 in 2023, indicating a long-term rise in vulnerability. This trend reflects the growing complexity of global industrial chains and intensified resource competition, underscoring weaknesses in supply chain resilience amid industrial development. Consequently, developing effective strategies to reduce REEs supply chain vulnerability has become a critical global research priority.

FIGURE 2

Table 3 shows data on China’s market-based policy tools (), with a mean of 34.64 and a standard deviation of 41.09, while the mean proportion of these tools () is 0.21 with a standard deviation of 0.22. Figures 1, 3 reveal significant fluctuations in both the number and proportion of these tools from 2011 to 2023. Between 2019 and 2022, both indicators trended upward, peaking at 131 tools in 2022. Although the number slightly declined in 2023, the proportion remained above 25%. The Chinese government has been adjusting the balance of mandatory policy tools while increasing market-based tool usage. Whether these changes affect the vulnerability of the global REEs supply chain warrants further investigation.

FIGURE 3

4.2 Empirical results of the ARDL model

4.2.1 Empirical results of the market-based policy tools ()

Table 4 shows regression results assessing the impact of market-based policy tools () on the vulnerability of the REEs supply chain. The coefficient is −0.02 (p = 0.080), indicating a negative correlation that holds only at the 10% significance level. The regression coefficient associated with has a 95% confidence interval of [−0.044, 0.003], which includes zero, indicating uncertainty in the parameter estimate. Based on this limited statistical evidence, the present study can only provisionally support research hypothesis H1, suggesting that an increase in the number of market-based policy tools marginally reduces the vulnerability of the global REEs supply chain. Economic analysis shows that introducing five additional market-based policy tools within a single year results in a reduction of the average vulnerability index by only 0.10, corresponding to approximately 0.5% of the sample mean value of 20.45. This finding highlights the generally weak suppressive effect exerted by the absolute number of policy tools. From the perspective of policy transmission mechanisms, this marginally significant outcome aligns with practical considerations. Market-based policies influence supply chain risks indirectly through multiple intermediary layers, including technological advancements and industrial developments. For example, policies related to REEs recycling must first promote the advancement of recycling technologies, which in turn depend on increased availability of secondary resources to alleviate supply chain vulnerability. The presence of such multi-layered indirect transmission pathways extends the policy effect chain, making the observed weak significance in empirical results consistent with the inherent characteristics of the industry. To ensure that the observed negative association is not an artifact of the selected lag order, Supplementary Appendix C in the supplementary materials employs the Bayesian Information Criterion (BIC) to reselect the optimal lag structure for robustness checks. The re-estimation confirms that the core coefficient of retains its negative sign and remains marginally significant at the 10% level, demonstrating a degree of robustness in the observed weak negative correlation.

In summary, China’s market-based REEs policies may mitigate supply and demand risks within the REEs supply chain by fostering the synergistic interaction of technology, industry, and policy incentives. This integrated approach tends to reduce the vulnerability of the global REEs supply chain. Both empirical evidence and theoretical analysis suggest that policy tools exert a meaningful impact on ensuring the security of the global REEs supply chain.

4.2.2 Empirical results of the proportion of market-based policy tools ()

4.2.2.1 Short-term result analysis for the proportion of market-based policy tools

Table 5 shows the short-term regression results on the effect of the proportion of market-based policy tools () on the vulnerability of the REEs supply chain. The estimated coefficient is −4.56 (p = 0.005), significant at the 1% level, indicating a negative relationship. This suggests that increasing market-based policy tools reduces supply chain vulnerability in the short term, supporting hypothesis H2a. Specifically, a 10% increase in these tools (e.g., from 20% to 30%) reduces vulnerability by about 0.456, or 2.23% of the sample mean. These findings confirm the significant role of market-based policy tools in reducing the vulnerability of the REEs supply chain.

From a theoretical perspective, a growing proportion of market-based policy tools mitigates supply and demand risks by reshaping the short-term operational decisions and long-term strategic planning of REEs-related enterprises, thereby reducing overall supply chain vulnerability. This interpretation aligns with the policy tool theory proposed by Howlett et al. (2009), asserting that “market-based approaches can reduce the inhibitory effects of mandatory controls and improve the responsiveness of micro-level actors.” The development of REEs industrial parks clearly demonstrates the synergistic effects of technological advancement, industrial optimization, and policy incentives:

First, from an industrial perspective, industrial parks reduce production costs for small and medium-sized enterprises (SMEs) by providing shared infrastructure and lowering transportation expenses, thereby easing cost pressures. This cost relief stabilizes firm production and prevents capacity suspension, effectively mitigating supply-side risks. The savings can be reinvested in enterprise-specific renewable energy technology research and development. Consequently, industrial parks reduce both supply and demand risks by lowering operational costs and fostering technological innovation, thereby enhancing the stability of REEs supply networks.

Second, industrial clustering helps SMEs broaden their market reach and build regional brand credibility. This cluster effect enhances market reputation, reduces dependence on individual customers, diversifies market exposure, and mitigates demand-side risks.

Furthermore, talent support and financial incentives from the policy incentive dimension are crucial for SMEs to establish operations in industrial parks and drive technological innovation. These policies attract specialized personnel in REEs technology, strengthen the innovation foundation, and significantly reduce operational costs. Enabled by flexible, market-oriented governance, renewable energy enterprises can adjust their operational strategies accordingly, facilitating industrial upgrading and technological iteration, ultimately enhancing the resilience of the REEs supply chain.

As shown in the short-term regression results in Table 5, the coefficient of the error correction term (, equal to in Equation 5) is estimated at −1.48 and is statistically significant at the 1% level (p < 0.001). Economically, this indicates that any deviation of the vulnerability of the global REEs supply chain from its long-term equilibrium—caused by fluctuations in the proportion of China’s market-based policy tools—will be overcorrected by 1.48 times the magnitude of the initial gap in the following year, thereby restoring equilibrium. Narayan and Smyth (2006) explicitly noted that when the coefficient lies between −2 and −1, the economic system’s adjustment process toward long-term equilibrium produces oscillations around the equilibrium value with progressively diminishing amplitude, rather than a monotonic and steady convergence to the equilibrium path. Furthermore, if the coefficient is less than −2, the system exhibits unbounded explosive fluctuations and consequently loses its stable cointegration properties. Since the coefficient of −1.48 falls firmly within the stable range −2< <−1 and is significantly less than −1, this result indicates that the supply chain converges to equilibrium through oscillatory overcorrection rather than smooth, monotonic adjustment. Despite this cyclical adjustment path, the system still maintains a self-equilibrating mechanism without divergent risks. This distinctive dynamic adjustment behavior can be attributed to the unique global supply structure and strategic characteristics of the REEs industry. First, the concentration of REEs separation and processing capacity is high, with the global supply chain dependent on a limited number of suppliers; consequently, policy changes in a single country can significantly disrupt the market equilibrium. Second, REEs are critical strategic resources for national security, prompting major consumer-importing nations to maintain strategic stockpiles that can be rapidly deployed to mitigate supply chain risks. Third, as the principal global supplier, China is able to coordinate administrative regulatory measures—such as controlling total REEs mining volumes and managing export licenses—with market-based policies to implement counter-cyclical adjustments. The interplay between domestic administrative regulations and market-based tools forms an internal hedging mechanism. When combined with external hedging actions—such as the release of strategic reserves by foreign importing countries—this leads to rapid, bidirectional policy interventions at both domestic and international levels. These multiple forces collectively accelerate the equilibrium restoration process, ultimately resulting in the observed cross-year overshoot adjustment phenomenon. To exclude bias from inappropriate model setting, this study conducted lag order selection, comprehensive residual diagnostic tests, and parameter stability assessments using CUSUM and CUSUMSQ tests, as detailed in Supplementary Appendices C,D of the supplementary materials. All tests were passed satisfactorily, thereby affirming the robustness of the conclusions regarding the coefficient.

Integrating the insignificant long-term coefficients associated with market-based policy tools and the highly significant , the dynamic behavior of the global REEs supply chain can be further elucidated as follows: On one hand, although market-based policy measures effectively mitigate supply chain vulnerabilities in the short term, their impact is constrained by factors such as technological advancements, global capacity realignments, and geopolitical disruptions, thereby preventing the establishment of a stable and enduring suppressive effect. On the other hand, the relatively large and statistically significant adjustment coefficients indicate that the REEs supply chain exhibits considerable shock resistance and self-repair capabilities. When short-term policy interventions cause deviations from the long-term equilibrium, various regulatory mechanisms—including domestic policy modifications, strategic reserve releases abroad, and adjustments in the global industrial configuration—act swiftly to restore equilibrium within approximately 1 year.

In conclusion, the short-term regression results confirm that a greater proportion of market-based policy tools alleviates fragility in the REEs supply chain. These policies foster synergy among technology, industry, and policy incentives, directly influencing enterprises’ short-term decision-making to stabilize supply chain performance. This study integrates empirical findings, policy tool theory, and a three-dimensional mechanism framework, providing strong evidence that a higher share of market-based policies in China’s REEs sector will effectively enhance the short-term resilience of global REEs supply chains.

4.2.2.2 Long-term result analysis for the proportion of market-based policy tools

Table 6 shows the long-term regression results examining the effect of the proportion of market-based policy tools () on the vulnerability of the global REEs supply chain. The long-term coefficient is −1.25 (p = 0.444) and statistically insignificant. This lack of significance may result from various unobserved confounding factors, such as technological uncertainty, the evolution of global production capacity, and cross-border geopolitical shocks. Firstly, considerable uncertainty surrounds technologies related to REEs. The research and development cycles for advanced REEs separation techniques, functional materials, substitutes, and recycling technologies are lengthy and require substantial financial investment. Although short-term fiscal and tax incentives may stimulate innovation, the process of technological iteration remains highly uncertain. Consequently, relying solely on subsidies is insufficient to achieve sustained and stable innovation benefits, and policies focused exclusively on research and development support demonstrate limited long-term efficacy. Furthermore, downstream application sectors face significant risks associated with cross-category material substitution, which exacerbates the volatility of long-term policy outcomes. Even if REEs preparation and processing technologies reach maturity and are widely adopted, breakthroughs in novel materials and processes outside the sector may suppress market demand for REEs. A pertinent example is the decline in demand for REEs such as lanthanum, cerium, and europium following the widespread adoption of LED lighting technology, which replaced REEs fluorescent lamps as a primary consumption application. Over the long term, emerging cutting-edge technologies—such as REEs-free permanent magnets and innovative catalytic materials—may substitute REEs in key sectors including new energy, electronics, and catalysis. Such disruptive innovations at the application level are inherently unpredictable and continuously disrupt the long-term demand for REEs, thereby diminishing the sustained impact of policies like R&D subsidies and industrial support. This technological uncertainty is a significant reason why the long-term effects of the are statistically insignificant. Secondly, on the supply side, many countries are accelerating the development of overseas REEs mining operations and expanding smelting capacities abroad, thereby continuously reshaping the global REEs supply landscape. Relying solely on domestic industrial coordination and price regulation is insufficient to mitigate market fluctuations caused by global capacity shifts, which in turn undermines the long-term stabilizing effects of such policies. Finally, at the international level, REEs often become focal points of geopolitical competition. There is a widespread perception abroad that China uses REEs as a geopolitical tool. Trade barriers, resource protection policies enacted by various countries, and fluctuations in international public opinion continually disrupt the industrial supply chain. It is important to emphasize that the regression model used in this study includes the Geopolitical Risk Index () as a control variable to account for cross-border supply shocks. However, relying solely on the coefficient of the control variable does not sufficiently address the empirical limitation posed by the weak statistical relationship observed in the long-term effects of . These factors make unilateral domestic market policies insufficient to stabilize long-term supply and demand dynamics. Moreover, international cooperation faces inherent resistance, reducing the long-term effectiveness of isolated domestic policy measures. Considering the statistically insignificant long-term coefficients shown in Table 6, all long-term transmission channels elaborated above are merely theoretical conjectures without robust empirical evidence from our long-term regression results. We speculate that this insignificance partly stems from limitations in our sample size and observation span. Accordingly, this study avoids drawing definitive long-term policy judgments solely based on the long-term effects of .

To further confirm that the insignificant long-term effect of does not stem from insufficient early observations or local subsample bias, this study adopts recursive estimation for robustness testing. Figure 4 plots the time-varying characteristics of ’s long-term coefficient, with detailed interpretations as follows: (1) 2014–2017: In the early stage of the recursive window, the limited time-series observations lead to drastic swings in the long-term coefficient. The extremely wide 95% confidence interval reveals great uncertainty in parameter estimation, driven by the small sample size. (2) 2018–2023: As the recursive window continuously absorbs new annual observations, the sample capacity expands. The long-term coefficient gradually converges and fluctuates slightly around zero. (3) Over the full sample period, the 95% confidence interval always contains zero, suggesting there is insufficient statistical evidence to support a significant long-term impact of . This result aligns with the baseline regression outcomes. Recursive estimation serves to examine the cross-time stability of the long-term coefficient. Even with the full and adequate time-series sample, ’s long-term effect remains statistically insignificant, which effectively eliminates the concern that the insignificance originates from local subsample bias.

FIGURE 4

Collectively, this evidence suggests that China’s market-based policies in the REEs sector primarily produce transient effects without generating sustained long-term impacts. While these market-based tools can promptly mitigate short-term risks, the inherent resilience and self-correcting mechanisms of the global supply chain tend to neutralize these policy benefits over time, failing to form persistent constraints on systemic supply risks.

Based on the preceding analysis, hypothesis H2b—which posits that, in the long term, an increasing proportion of market-based policy tools in China’s REEs sector will effectively reduce the vulnerability of the global REEs supply chain—is not supported.

5 Discussion

This study analyzes the impact of China’s market-based REEs policies on global REEs supply chain vulnerability using an ARDL model. Results show that increased market-based policy tools tend to reduce supply chain vulnerability. For the proportion of market-based policy tools (), its short-term effects are significantly negative, while its long-term coefficient is negative but not statistically significant. These findings support hypotheses H1 and H2a but do not confirm H2b. The absence of significant long-term effects may be attributed to unobserved or complex factors, such as technological uncertainties, shifts in global capacity configurations, and cross-border geopolitical shocks.

Prior research has predominantly focused on China’s initial REEs export control policies (Mancheri, 2015; Zhang et al., 2024), which were characterized by administrative restrictions that increased risks within the global REEs supply chain. These policies are often interpreted in international literature as emblematic of resource nationalism in REEs governance. In contrast, the present study emphasizes the mitigating role of market-based alternative policies in reducing supply chain risks. Against the backdrop of escalating global competition for critical minerals and intensified efforts by Western nations to establish cooperative alliances for mineral supply chains, the ARDL-based empirical findings offer novel quantitative evidence that differentiates market-based regulatory mechanisms from compulsory administrative controls. The results confirm that China effectively maintains global REEs supply chain stability through market-based governance, challenging the simplistic narrative that China’s REEs policies serve solely as instruments of geopolitical competition. Unlike traditional mandatory export controls, market-based policies function by balancing supply and demand, promoting sustainable industrial operations, and diversifying global supply chain risks. Existing scholarship on the geopolitical economy of REEs predominantly focuses on restrictive resource policies, with limited quantitative empirical analysis of market-based regulatory frameworks.

This study analyzes China’s market-based REEs policies across three dimensions: technological, industrial, and policy incentives. The technological dimension covers advances in substitute technologies and secondary recycling, which reduce supply chain vulnerability by balancing supply and demand (Binnemans et al., 2018; Gueroult et al., 2018). The industrial dimension focuses on upgrading industries to maintain pricing power over REEs, improving supply stability and mitigating risks (Shen et al., 2020). Empirical evidence indirectly supports the positive impact of these dimensions in reducing the vulnerability of the global REEs supply chain. Most existing studies limit their scope to analyzing the protective effects of various policies on REEs supply security, rarely exploring the deeper developmental imperatives underlying the superficial objective of “stabilizing REEs supply.” A stable and sustainable REEs supply constitutes a fundamental material foundation for green and low-carbon industries such as wind power and energy storage and is a material prerequisite for implementing global climate governance. From the perspective of the global industrial division of labor, the photovoltaic (PV) and REEs industries share similar environmental burden characteristics. While global PV end-use applications are geographically dispersed, the majority of PV manufacturing is concentrated in China, which consequently bears most of the associated carbon emissions and pollutant discharges (e.g., sulfur) (Zheng et al., 2026a; Zheng et al., 2026b). The REEs industry exhibits comparable patterns; leveraging advanced mining, separation, and smelting technologies, China undertakes most global REEs production and processing activities. The environmental costs—including wastewater, solid waste, and substantial carbon emissions—are predominantly localized within China. In this context, China’s market-based REEs policies have dual environmental benefits: they incentivize REEs recycling and green smelting technology research and development, thereby promoting industrial restructuring and the elimination of outdated capacity, which reduces pollution and carbon emissions; concurrently, they stabilize the REEs supply chain, continuously supporting critical raw materials for clean energy sectors such as wind power and energy storage. Collectively, these policies alleviate local environmental pressures within the REEs industry while promoting global low-carbon industrial development, thereby contributing to global environmental governance and long-term climate objectives.

The three-dimensional analytical framework of market-based policies developed here can be extended to evaluate the generalizability of empirical findings across different national contexts and critical mineral sectors. For mineral-rich developing countries, China’s practical experience in technological innovation, industrial integration, and supportive incentive policies offers valuable lessons. However, replicating this policy package requires a comprehensive industrial system and a substantial endowment of mineral resources. In contrast, most European and American countries primarily import minerals and lack these conditions. When applying this framework to other mineral supply chains exposed to geopolitical and environmental risks—such as lithium and cobalt—the core logic of market-based risk mitigation remains theoretically applicable. Nonetheless, due to significant disparities in resource distribution and industrial structures among countries, quantitative conclusions about policy effects require targeted empirical validation.

6 Conclusion

This study employed an Autoregressive Distributed Lag (ARDL) model to analyze the impact of China’s market-based REEs policies on the vulnerability of the global REEs supply chain. The key findings are: First, China’s market-based REEs policies tend to reduce supply chain vulnerability by synergistically promoting technological innovation, optimizing industrial structure, and implementing policy incentives. These measures mitigate risks on both supply and demand sides. Second, increasing the proportion of these policies has a significant short-term effect in lowering vulnerability, as enterprises can quickly adjust operations through coordinated technological, industrial, and policy efforts. Third, the long-term suppressive effect of the proportion of these policies on supply chain vulnerability is statistically insignificant. This outcome is attributed to the fact that long-term policy effects depend on indirect transmission via continuous technological innovation, industrial iteration, and ecological governance—processes that are susceptible to disruption by external factors such as global capacity expansion, technological uncertainty, and geopolitical shocks.

This study makes a significant academic contribution by employing the ARDL model to quantitatively estimate the short- and long-term dynamic effects of market-based REEs policies on the vulnerability of the global REEs supply chain. It addresses a notable research gap, as existing literature predominantly focuses on administrative export quota restrictions and lacks time-series empirical evidence regarding market-based tools.

Overall, China’s market-based REEs policies have effectively reduced the vulnerability of the global REEs supply chain, providing valuable experiential insights for managing risks in the global critical mineral supply chain. However, due to constraints imposed by technological innovation cycles, global capacity fluctuations, and geopolitical shocks, relying on a single market-based policy tool is insufficient to achieve stable, long-term risk mitigation. Instead, a diversified policy portfolio—including regulation of the REEs industry, ecological and environmental protection, and coordination of foreign trade—is essential.

It is important to acknowledge that the short-term policy recommendations presented here are supported by substantial empirical evidence from the short-term estimates of the proportion of market-based policy tools (). In contrast, the medium- and long-term industrial development strategies proposed in this study rely purely on theoretical industrial logic and cannot be empirically validated by the statistically insignificant long-term coefficients of . Accordingly, this paper avoids drawing definitive long-term policy conclusions based on the long-term effects of .

Based on these findings, the study proposes two practical, phased, and detailed policy recommendations.

6.1 Improve the domestic market-based policy system

6.2 Advancement of international cooperation in the global REEs sector

Lead Implementing Agencies: Ministry of Commerce, Ministry of Natural Resources, Ministry of Foreign Affairs.

While some countries express concern that China might leverage REEs resources as a geopolitical bargaining tool, empirical evidence indicates that China’s past adjustments to REEs export policies were primarily motivated by environmental considerations—specifically, controlling pollution from REEs mining and processing—and supporting downstream high-end REEs industries, rather than political or economic coercion. It is imperative that countries abandon zero-sum perspectives and embrace multi-stakeholder collaboration to build an integrated, full REEs industry chain. This approach would alleviate international concerns, ensure a stable REEs supply, and promote a transition from sole reliance on Chinese REEs imports to coordinated global development of the REEs industry. Specific strategies include joint exploration and development of overseas REEs resources; collaborative research and development on key technologies such as separation, smelting, REEs substitution, and recycling; and the joint establishment of unified global environmental and product quality standards for REEs. Given significant institutional risks arising from disparities in environmental regulations and industrial protection policies among countries, a 3-year short-term bilateral cooperation pilot is recommended initially, with dynamic optimization of cooperation details through meetings every 12–18 months. Upon maturation of institutional alignment, a medium-to long-term (3–8 years) plan could promote the establishment of a multilateral global REEs governance mechanism. Through multi-level international collaboration, REEs can be transformed from a source of geopolitical conflict into a platform for global low-carbon transition and mutual benefit.

References

• 1

  BalaramV. (2019). Rare earth elements: a review of applications, occurrence, exploration, analysis, recycling, and environmental impact. Geosci. Front.10 (4), 1285–1303. 10.1016/j.gsf.2018.12.005

  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  BartekováE.KempR. (2016). National strategies for securing a stable supply of rare earths in different world regions. Resour. Policy49, 153–164. 10.1016/j.resourpol.2016.05.003

  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  BinnemansK.JonesP. T.MüllerT.YurramendiL. (2018). Rare earths and the balance problem: how to deal with changing markets?J. Sustain. Metallurgy4, 126–146. 10.1007/s40831-018-0162-8

  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  BodeC.WagnerS. M. (2015). Structural drivers of upstream supply chain complexity and the frequency of supply chain disruptions. J. Operations Manag.36, 215–228. 10.1016/j.jom.2014.12.004

  • CrossRef
  • Google Scholar
  • View reference in article

• 5

  CaiX. M.ZhengS. X.ZhangX.YeZ. H.LiuC.TanZ. L. (2024). The impact of CO₂ emission synergy on PM₂.₅ emissions and a dynamic analysis of health and economic benefits: a case study of china’s transportation industry. J. Clean. Prod.471, 143405. 10.1016/j.jclepro.2024.143405

  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  CaldaraD.IacovielloM. (2018). Measuring geopolitical risk. Int. Finance Discuss. Pap.1222, 1–66. 10.17016/ifdp.2018.1222

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  ChaiS.ZhangZ. C.GeJ. P. (2020). Evolution of environmental policy for China’s rare earths: comparing central and local government policies. Resour. Policy68, 101786. 10.1016/j.resourpol.2020.101786

  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  CharalampidesG.VatalisK. I.ApostoplosB.Ploutarch-NikolasB. (2015). Rare earth elements: industrial applications and economic dependency of Europe. Procedia Econ. Finance24, 126–135. 10.1016/S2212-5671(15)00630-9

  • CrossRef
  • Google Scholar

• 9

  CuiX. M.XiongW. T.YangP. P.XuQ. Y. (2022). Measuring global supply chain fragility: evidence from trade network analysis. Stat. Res.39 (8), 38–52. 10.19343/j.cnki.11-1302/c.2022.08.003

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  EggertR.WadiaC.AndersonC.BauerD.FieldsF.MeinertL.et al (2016). Rare earths: market disruption, innovation, and global supply chains. Annu. Rev. Environ. Resour.41, 199–222. 10.1146/annurev-environ-110615-085700

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  FilhoW. L.KotterR.ÖzuyarP. G.AbubakarI. R.EustachioJ. H. P. P.MatandirotyaN. R. (2023). Understanding rare earth elements as critical raw materials. Sustainability15 (3), 1919. 10.3390/su15031919

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  GarcíaM. V. R.KrzemieńA.del CampoM. Á. M.ÁlvarezM. M.GentM. R. (2017). Rare earth elements mining investment: it is not all about China. Resour. Policy53, 66–76. 10.1016/j.resourpol.2017.05.004

  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  GueroultR.RaxJ. M.FischN. J. (2018). Opportunities for plasma separation techniques in rare earth elements recycling. J. Clean. Prod.182, 1060–1069. 10.1016/j.jclepro.2018.02.066

  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  GuoQ.MaiZ. S. (2024). How do seasonal, significant events, and policies affect China’s REE export prices? Based on deep learning perspective. Resour. Policy96, 105205. 10.1016/j.resourpol.2024.105205

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  HamedM. M.TuranH. H.ElsawahS. (2024). Balancing supply diversification and environmental impacts: a system dynamics approach to de-risk rare earths supply chain. Resour. Policy92, 105038. 10.1016/j.resourpol.2024.105038

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  HanA. P.GeJ. P.LeiY. L. (2016). Vertical vs. horizontal integration: game analysis for the rare earth industrial integration in China. Resour. Policy50, 149–159. 10.1016/j.resourpol.2016.09.006

  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  Hayes-LabrutoL.SchillebeeckxS. J. D.WorkmanM.ShahN. (2013). Contrasting perspectives on China’s rare earths policies: reframing the debate through a stakeholder lens. Energy Policy63, 55–68. 10.1016/j.enpol.2013.07.121

  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  HowlettM.PerlA.RameshM. (2009). Studying Public Policy: Policy Cycles and Policy Subsystems. Oxford: Oxford University Press.

  • Google Scholar
  • View reference in article

• 19

  HuH.TanZ. L.LiuC.WangZ.CaiX. M.WangX.et al (2022). Multi-timescale analysis of air pollution spreaders in Chinese cities based on a transfer entropy network. Front. Environ. Sci.10, 970267. 10.3389/fenvs.2022.970267

  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  IlankoonI. M. S. K.DushyanthaN. P.MancheriN.EdirisingheP. M.NeethlingS. J.RatnayakeN. P.et al (2022). Constraints to rare earth elements supply diversification: evidence from an industry survey. J. Clean. Prod.331, 129932. 10.1016/j.jclepro.2021.129932

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  JowittS. M. (2022). Mineral economics of the rare-earth elements. MRS Bull.47, 276–282. 10.1557/s43577-022-00289-3

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  KeilhackerM. L.MinnerS. (2017). Supply chain risk management for critical commodities: a system dynamics model for the case of the rare earth elements. Resour. Conservation Recycl.125, 349–362. 10.1016/j.resconrec.2017.05.004

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  KripfganzS.SchneiderD. C. (2023). Ardl: estimating autoregressive distributed lag and equilibrium correction models. Stata J.23 (4), 983–1019. 10.1177/1536867X231212434

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  LaiF. J.XiongD. P.ZhuS.LiY. Z.TanY. Z. (2023). Will geopolitical risks only inhibit corporate investment? Evidence from China. Pacific-Basin Finance J.82, 102134. 10.1016/j.pacfin.2023.102134

  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  LaiC. H.WangX. L.LiH. K.ZhouY. B. (2024). Unleashing the power of closed-loop supply chains: a Stackelberg game analysis of rare earth resources recycling. Sustainability16 (12), 4899. 10.3390/su16124899

  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  LeeY.DacassT. (2022). Reducing the United States’ risks of dependency on China in the rare earth market. Resour. Policy77, 102702. 10.1016/j.resourpol.2022.102702

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  LeiZ. R.ZhangX. L.LiW. B.HanY. X. (2025). Rare earth processing in China. Mineral Process. Extr. Metallurgy Rev.47, 254–277. 10.1080/08827508.2024.2449254

  • CrossRef
  • Google Scholar
  • View reference in article

• 28

  LiZ. S.HamidiA. S.YanZ. M.SattarA.HazraS.SoulardJ.et al (2024). A circular economy approach for recycling electric motors in the end-of-life vehicles: a literature review. Resour. Conservation Recycl.205, 107582. 10.1016/j.resconrec.2024.107582

  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  MaJ.LiM. L. (2021). The feedback effect of rare earth development policy on economy and society. E3S Web Conf.228, 01014. 10.1051/e3sconf/202122801014

  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  MancheriN. A. (2015). World trade in rare earths, Chinese export restrictions, and implications. Resour. Policy46, 262–271. 10.1016/j.resourpol.2015.10.009

  • CrossRef
  • Google Scholar
  • View reference in article

• 31

  McLellanB. C.CorderG. D.AliS. H. (2013). Sustainability of rare earths-An overview of the state of knowledge. Minerals3 (3), 304–317. 10.3390/min3030304

  • CrossRef
  • Google Scholar
  • View reference in article

• 32

  McNultyT.HazenN.ParkS. (2022). Processing the ores of rare-earth elements. MRS Bull.47, 258–266. 10.1557/s43577-022-00288-4

  • CrossRef
  • Google Scholar
  • View reference in article

• 33

  NarayanP. K.SmythR. (2006). What determines migration flows from low-income to high-income countries? An empirical investigation of Fiji–U.S. migration 1972–2001. Contemp. Econ. Policy24 (2), 332–342. 10.1093/cep/byj019

  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  PanA.FengS. S.HuX. Y.LiY. Y. (2021). How environmental regulation affects China’s rare earth export?PLOS One16 (4), e0250407. 10.1371/journal.pone.0250407

  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  PawarG.EwingR. C. (2022). Recent advances in the global rare-earth supply chain. MRS Bull.47, 244–249. 10.1557/s43577-022-00305-6

  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  RiddleM.MacalC. M.ConzelmannG.CombsT. E.BauerD.FieldsF. (2015). Global critical materials markets: an agent-based modeling approach. Resour. Policy45, 307–321. 10.1016/j.resourpol.2015.01.002

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  SchneiderL.BergerM.Schüler-HainschE.KnöfelS.RuhlandK.MosigJ.et al (2014). The economic resource scarcity potential (ESP) for evaluating resource use based on life cycle assessment. Int. J. Life Cycle Assess.19, 601–610. 10.1007/s11367-013-0666-1

  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  ShenY. Z.MoomyR.EggertR. G. (2020). China’s public policies toward rare earths, 1975–2018. Mineral. Econ.33, 127–151. 10.1007/s13563-019-00214-2

  • CrossRef
  • Google Scholar
  • View reference in article

• 39

  U.S. Geological Survey (USGS) (2024). Mineral commodity summaries. Available online at: https://www.usgs.gov/centers/nmic/mineral-commodity-summaries (Accessed September 20, 2025).

  • Google Scholar
  • View reference in article

• 40

  WagnerS. M.BodeC. (2006). An empirical investigation into supply chain vulnerability. J. Purch. Supply Manag.12 (6), 301–312. 10.1016/j.pursup.2007.01.004

  • CrossRef
  • Google Scholar

• 41

  WangC. Z. (2023). “Rare earth research, production, policy, and future development,” in Theory and Application of Rare Earth Materials. Editor WangC. Z. (Singapore: Springer), 351–368. 10.1007/978-981-19-4178-8_20

  • CrossRef
  • Google Scholar

• 42

  WübbekeJ. (2013). Rare earth elements in China: policies and narratives of reinventing an industry. Resour. Policy38 (3), 384–394. 10.1016/j.resourpol.2013.05.005

  • CrossRef
  • Google Scholar

• 43

  YanG. L.LiZ. X. (2019). Global political economy of rare earths: changing positions of major market actors including China, European Union, Japan and United States. IOP Conf. Ser. Earth Environ. Sci.295 (5), 052022. 10.1088/1755-1315/295/5/052022

  • CrossRef
  • Google Scholar

• 44

  ZhangL.GuoQ.ZhangJ. B.HuangY.XiongT. (2015). Did China’s rare earth export policies work? Empirical evidence from USA and Japan. Resour. Policy43, 82–90. 10.1016/j.resourpol.2014.11.007

  • CrossRef
  • Google Scholar

• 45

  ZhangT. T.ZhangP. F.PengK.FengK. S.FangP.ChenW. Q.et al (2022). Allocating environmental costs of China’s rare earth production to global consumption. Sci. Total Environ.831, 154934. 10.1016/j.scitotenv.2022.154934

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 46

  ZhangH. W.CaoH. L.GuoY. Q. (2024). The time-varying impact of geopolitical relations on rare earth trade networks: what is the role of China’s rare earth export restrictions?Technol. Forecast. Soc. Change206, 123550. 10.1016/j.techfore.2024.123550

  • CrossRef
  • Google Scholar

• 47

  ZhangH.HuM.JiangS. Y. (2025). Profit or growth? The impacts of supplier dependence and customer dependence on SMEs’ performance. Sustainability17 (3), 1302. 10.3390/su17031302

  • CrossRef
  • Google Scholar

• 48

  ZhengS. X.HeT.LiuY. Z.CaiX. M.LuK. P.ZhangL. X.et al (2026a). Spatially decoupled sulfur emissions in global photovoltaic supply chains and their vulnerability to trade barriers. Environ. Impact Assess. Rev.121, 108554. 10.1016/j.eiar.2026.108554

  • CrossRef
  • Google Scholar

• 49

  ZhengS. X.YeZ. H.LuK. P.HeT.LiuY. Z.ZhangL. X.et al (2026b). Accounting for Non-CO₂ greenhouse gases in global PV trade: implications for climate responsibility and policy. Sustain. Prod. Consum.62, 55–69. 10.1016/j.spc.2025.12.012

  • CrossRef
  • Google Scholar