Assessment of Green Development Levels and Exploration of Regional Differences in Countries Along the Belt and Road

Abstract

Green development is an important path to breaking the “growth-pollution” paradox and realizing the United Nations 2030 Agenda for Sustainable Development. Most of the countries along the “Belt and Road” are developing countries with fragile ecological environments, and the pressure of economic activities on environmental resources exceeds the global average, so the countries have a strong will to promote green development. Accurately grasping the level of green development and the dynamic evolution trend of BRI countries can help to explore the implementation effect of policies at a certain stage, and on this basis, exploring regional correlations and differences can help to provide data support for future regional cooperation and policy formulation. Based on this, this study comprehensively evaluates the level of green development in the BRI countries by constructing a multi-dimensional evaluation index system and then explores the correlation and regional differences of green development in the countries along the route by utilizing the Moran index and the Thiel index. The results indicate that the green development levels of BRI countries have shown an upward trend; the spatial association exhibited a fluctuating upward trend, with significant changes in local clustering characteristics. Overall disparities in green development levels were relatively small, with intra-group differences being the primary contributor to these disparities, and West Asia and the Middle East displayed larger variations in green development levels. This study provides policy recommendations for promoting green development in BRI countries.

1. Introduction

In the face of the twin demands for economic transformation and environmental sustainability, numerous countries are vigorously engaging in green development strategies [1]. The United States has introduced the Green New Deal and enacted the American Clean Energy and Security Act (ACESA). Japan has developed a comprehensive plan under the “Green Growth Strategy” [2]. The European Union has integrated a number of green economy initiatives into strategic documents like “Europe 2020” and the “Resource Efficiency Roadmap” [3]. China has consistently implemented the development concepts of innovation, coordination, green development, openness, and sharing. Notably, since the commencement of the “Belt and Road” initiative (BRI), through the establishment of low-carbon demonstration zones, material support, experience and technology sharing, as well as professional training under the guarantee of institutional frameworks, China has significantly quickened the pace of green and low-carbon transformation in the relevant countries [4].

The BRI involves a large number of countries. As a key representative of emerging economies, the green development of BRI countries exerts a far-reaching influence on the achievement of global sustainable development goals. Currently, however, BRI countries contribute over 55% of global CO₂ emissions. The environmental and resource pressures stemming from the economic activities of these countries are higher than the global average [5]. Despite the abundance of natural resources, these countries face challenges such as complex terrain, fragile ecological environment, frequent natural disasters, insufficient economic development, irrational industrial structure, and relatively low income due to poor infrastructure. Therefore, building a green BRI is crucial to promoting sustainable development in these regions.

The idea of green development emerged in the 1960s. The American scholar Boulding introduced the Spacecraft Economic Theory, and in 1989, the British economist David Pearce coined the term “green economy” [6,7]. They all emphasized that economic development should not simply pursue production but should be carried out within the limits of the environment and human beings’ own tolerance and focus on the efficient use of resources and recycling. As green development has gained widespread attention, different organizations and scholars have outlined definitions of green development that, while different in focus, all contain the core idea of reconciling economic development and environmental protection. For example, OECD considers green development not only as a way to achieve economic growth and progress, but also as an effective way to prevent environmental degradation, the sharp decline of biodiversity and the unsustainable use of resources [8]. The World Bank emphasizes that green development implies a model of economic growth that is both environmentally friendly and socially inclusive and that its core lies in the efficient use of natural resources, minimizing polluting emissions and thus mitigating the impact on the environment [9]. Wang and Zhang state that “green development” is a new development model that attains sustainable development by protecting the ecological environment within the constraints of ecological and environmental capacities and the resource carrying capacity [10]. Hu and Zhou assert that green development is the second generation of sustainable development, highlighting the systemic, holistic, and coordinated nature of economic, social, and natural systems [11].

With the continuous progress of the Belt and Road Initiative (BRI), the attention to related issues has grown. On one hand, scholars indicate that the BRI can mobilize capital and build infrastructure more swiftly than projects supported by traditional-development-financing organizations due to its faster implementation speed, thereby promoting future trade. Secondly, in the long run, the BRI is anticipated to have positive effects on energy transition [12]. Thirdly, research shows that after the implementation of the BRI, green development in BRI countries has improved significantly [13]. However, some scholars have expressed concerns about potential problems in BRI implementation. They suggest that while bringing early value, it might cause unsustainable debt in recipient countries, lack transparent project decision-making, and harm the local and global environments [14].

Most scholars hold a dialectical view of the BRI, recognizing its value while identifying challenges and optimization strategies. For example, Dong et al. (2016) stressed the strategic significance of promoting green development along the BRI but pointed out problems such as insufficient environmental regulation systems for overseas investments and imperfect regional cooperation mechanisms. They proposed a strategic framework for promoting green BRI development [15]. Xie (2017) analyzed the opportunities and challenges for green BRI construction in domestic and international contexts, emphasizing the necessity of ecological cooperation to ensure project safety, enhance China’s international image, and promote ecological civilization concepts. Policy suggestions were put forward to promote green BRI development [16]. Sun (2017) noted the fragile ecological conditions in some BRI countries and the lag in environmental legislation, emphasizing the urgency of establishing environmental legal frameworks for green BRI construction [17]. Fernando Ascensão (2018) argued that, although new infrastructure can boost socio-economic development, projects passing through environmentally sensitive areas need strict, region-specific guidelines to reduce biodiversity impacts [18]. Alice C. Hughes (2019) emphasized the need for comprehensive assessments of ecological impacts in large-scale infrastructure projects, advocating adaptive planning to minimize impacts on critical biodiversity areas [19]. Jon Elkind (2019) examined the BRI’s environmental implications, suggesting that, with strict environmental and social safeguards, BRI projects could greatly improve development prospects in participating countries [5]. Divya Narain et al. (2020) revealed that BRI economic corridors cross 150,000 km² of priority conservation areas, urging the immediate adoption of best-practice safeguards to protect biodiversity [20]. Mao et al. (2024) suggested that the Chinese government should formulate differentiated low-carbon economic cooperation policies based on the differences in the development levels of BRI countries and take measures such as strengthening the management of outward-investing enterprises, signing multilateral agreements, and establishing a benefit-sharing mechanism to promote low-carbon cooperation [21].

Against this background, the “Green BRI” concept has gained traction. Scholars have started to quantitatively assess the green development levels in BRI countries to inform policy-making. Huang et al. (2017) measured green total factor productivity in BRI countries using the SBM-DDF Luenberger productivity index, analyzing its spatiotemporal characteristics and influencing mechanisms [22]. Du et al. (2019) conducted a quantitative analysis of green development in 45 BRI countries through energy efficiency, carbon emission efficiency, and green performance metrics [23]. Lan (2020) established a Green BRI evaluation system integrating economic, industrial, sustainability, and environmental indicators, applying factor analysis to assess the green development levels of 42 BRI countries [24]. Cheng and Ge (2020) developed a 2015 green development index system for 49 BRI countries, using multi-level linear evaluation methods to explore development levels and correlations [25].

In conclusion, scholars have thoroughly explored the conceptual framework of green development and the potential advantages and challenges of BRI implementation. Based on this, extensive research has been carried out on green development issues in BRI countries, providing significant reference value for this study. However, current research has obvious drawbacks. First, there is a lack of systematic assessment. Most studies focus on individual countries or regions and fail to systematically evaluate the overall green development levels of BRI countries. Second, dynamic analysis is insufficient. Many studies use static analysis methods and cannot fully reveal the spatiotemporal evolution patterns of green development levels. Third, there is also a lack of quantitative analysis of the correlation and regional differences in the level of green development in the BRI countries. These research gaps seriously limit a comprehensive and dynamic understanding of green development in BRI countries.

To overcome these limitations, this study poses the following questions: How did the green development levels of BRI countries change over time from 2011 to 2020? How are the spatial correlations and localized aggregation characteristics? What are the regional differences and their main sources? Accordingly, this research aims to construct a multi-dimensional green development indicator system, combining the entropy weight method and TOPSIS method to accurately quantify the dynamic changes in green development levels across BRI countries, exploring the spatial correlation and localized agglomeration characteristics of green development in BRI countries using the Moran index. The Theil index will be used to quantify regional differences in green development levels. Compared with previous studies, this study systematically and dynamically evaluates the level of green development, spatial correlations and regional differences, and major sources of green development in the BRI countries. These efforts fill important gaps in existing research, provide key data support and practical guidance for promoting green BRI construction, and contribute to the achievement of sustainable development goals.

2. Materials and Methods

2.1. Study Area and Data Availability

The BRI has exerted a far-reaching influence globally, drawing significant attention from the international community, and it is an open international network of regional economic cooperation, and as such has no precise spatial scope; therefore, the study takes 65 countries, which are frequently mentioned in the literature, as the study area [26,27]. For the purposes of scientific research, this study roughly divides these countries into six regions based on their geographical attributes: East Asia, Central Asia, ASEAN, South Asia, Middle Eastern Europe, and West Asia–Middle East (Figure 1). Due to severe data limitations for some countries, after thorough consideration, we selected 54 BRI countries for this study (

Table 1. Regional zones and countries of the BRI region in this study.

Regions	Name of Nation	Number
ASEAN	Brunei, Cambodia, Timor-Leste, Lao people’s Democratic Republic, Indonesia, Malaysia, Myanmar, Philippines, Singapore, Thailand, Vietnam	11
Central Asia	Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan, Uzbekistan	5
East Asia	China, Mongolia, Russian Federation	3
Middle Eastern Europe	Albania, Bulgaria, Belarus, Croatia, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Serbia, Slovakia, Slovenia, the Czech Republic, the Republic of Moldova, Ukraine	16
Southern Asia	Afghanistan, Bangladesh, Bhutan, India, Maldives Nepal, Pakistan, Sri Lanka	8
West Asia–Middle East	Armenia, Azerbaijan, Bahrain, Egypt, Georgia, Iran, Israel, Lebanon, Qatar, Turkey, the United Arab Emirates	11

).

2.2. Construction of the Green Development Evaluation Level System

Scientific assessment of the current situation and changes in the level of regional green development levels can offer a scientific basis for relevant decision-making, thus promoting green development. At present, the construction of the green development evaluation index system mainly follows three approaches: First is a green GDP accounting system. This includes initiatives like the natural resource accounting introduced by the Norwegian Ministry of the Environment in 1974 as a tool for the central government to monitor and manage the nation’s natural resource stock [28]. The second is a multi-indicator measurement system for green development. For example, the United Nations Economic and Social Commission for Asia and the Pacific has an ecological efficiency indicator system that encompasses sectors such as agriculture, industry, manufacturing, and services. The OECD has developed a green growth measurement framework with primary indicators including natural resources, environmental productivity, quality of life, and policy response. The UNEP has devised a set of indicators for gauging the green economy, with an emphasis on policy interventions, human well-being, and equality [8,29,30]. The third is an integrated green development index system. This includes the Environmental Performance Index (EPI) created by Yale University and Columbia University. The Chinese Academy of Sciences’ Sustainable Development Strategy Research Group proposed an indicator system to evaluate sustainable development capabilities [31,32]. These indicator systems can assess diverse aspects of green development. However, in comparison, the comprehensive evaluation indicator system is more appropriate for evaluating the overall level of regional green development [33].

By summarizing the indicator systems in the relevant literature, it is evident that the focus of indicator selection varies according to different scales and development statuses of the regions. For instance, in 2015, Wang et al. evaluated the green development state in nine Chinese cities in the Pearl River Delta. They constructed an evaluation indicator system using 23 indicators from five aspects: improvement of the living environment, urban pollution control and utilization, enhancement of ecological efficiency, optimization of economic growth, and development of innovation potential [33]. Han et al. examined the green growth degree in ASEAN countries from 2010 to 2019, using 17 measures from three domains: resource efficiency, socioeconomic progress, and environmental and health competitiveness [1]. Cheng and Ge evaluated the green development levels of BRI countries in 2015, selecting 15 indicators from three aspects: economic and social development, natural resource consumption, and ecological environment competitiveness [25]. These three studies cover different regional scales. Small-scale regions, due to relatively easy access to comprehensive data, can address more detailed issues and employ more specific indicators. For instance, in the nine cities of the Pearl River Delta, besides the commonly used indicators of economy, resources, and ecology, the study also focused on living environment and urban development potential. In regions like ASEAN, with a relatively solid economic foundation, the emphasis is more on the sustainability of economic growth and environmental issues, resulting in fewer indicators. These three studies cover different regional scopes and provide an important reference for the construction of the indicator system of this study.

This study takes the three key dimensions of “economy, society, and environment”, emphasized by the sustainable development theory, and the two crucial dimensions of “resource carrying capacity” and “ecological and environmental capacity”, which are closely related to the green development theory, as the core guiding principles [34,35]. Meanwhile, indicators that have been proven significant for green development in existing studies, such as “degree of external openness” and “renewable energy consumption,” are included. Also, the regional characteristics of BRI countries are fully considered to construct an indicator system for evaluating green development levels [36,37,38,39]. This indicator system covers three dimensions: socio-economic development, resource utilization efficiency, and environmental competitiveness, with a total of 17 indicators (

Table 2. Evaluation system of green development level of the countries along the BRI.

Target Level	Indicators	Unit	Impact
Economic and social development	GDP growth rate	%	+
	GDP per capita	USD/capita	+
	Population growth rate	%	−
	Ratio of tertiary industry to GDP	%	+
	Ratio of primary industry to GDP	%	−
	Announced unemployment rate	%	−
	Foreign direct investment (net inflows of GDP)	%	+
	Government expenditure on education, ratio of GDP	%	+
Efficiency of resource and energy use	Energy intensity level of primary energy	MJ/$2011 PPP GDP	−
	Adjusted savings: energy depletion (% of GNI)	%	−
	Tree cover change	%	+
	Renewable energy consumption (% of total final energy consumption)	%	+
Competitiveness of environment	Greenhouse gas emissions (kt of CO2 equivalent)	Kt	−
	PM2.5 air pollution, mean annual exposure	μg/m3	−
	CO2 emissions	metric tons per capita	−
	Methane emissions (kt of CO2 equivalent)	Kt	−
	Nitrous oxide emissions (thousand metric tons of CO2 equivalent)	Kt	−

Note: “+” indicates a positive impact and “−” indicates a negative impact.

).

The first dimension is socio-economic development, which is characterized by the current state of economic development and the level of sustainability. The economic growth rate reflects the speed of expansion of the total economy, as well as its vitality and dynamics; GDP per capita is a key indicator of the region’s per capita economic output and standard of living, reflecting the extent to which individuals have benefited from the economic results; too rapid a population growth may increase the pressure on the distribution of society and resources. Reasonable and diversified industrial structure is conducive to the optimal allocation of resources and economic stability and innovation, and the process of green transformation is a manifestation of the harmonious development of the economy and the environment. International economic cooperation reflects the degree of integration and influence of the region in the global economic pattern. The labor market situation is related to the efficiency and quality of production, and the balance of supply and demand and the quality and rationality of employment structures are of great significance; the national investment in human resources development determines the cultivation of talents and the capacity of knowledge innovation, and the investment in social progress establishes the social foundation of economic development.

The second dimension is resource utilization efficiency, which is characterized by energy utilization efficiency and ecological resource protection. Energy utilization efficiency reflects the effectiveness of technology, process, and management; energy loss reflects the perfection of the industrial chain and the maturity of technology; the structure of energy consumption reflects the strategic orientation and diversification of supply; the high proportion of clean energy indicates the response to green development and the reduction in dependence on fossil energy; the change in the status of forest resources is related to the ecological protection, and changes in the area of forest resources reflect the efforts of the countries along the route to build up ecological construction and protect resources.

The third dimension is the status of environmental protection, measured by the emission of key pollutants. Key pollutant emissions reflect the energy consumption structure and transformation, production patterns, technology promotion and innovation, and pollutant treatment processes in the countries along the route. For example, the dominance of fossil energy, the lack of the technological management of enterprises to reduce emissions, the dependence of agriculture on chemical fertilizers and the lack of environmental technology, and the backwardness of industrial processes all lead to high emissions, which prompts the countries along the route to promote the upgrading and transformation of industrial enterprises with green technology, improve resource utilization efficiency, and reduce pollutant emissions.

2.3. Data Sources

The World Bank Open Dataset (WBOD), the BP Statistical Review of World Energy, the National Bureau of Statistics of China (NBSC) (https://www.stats.gov.cn, accessed on 5 September 2023), and the Belt and Road Portal of China (https://www.yidaiyilu.gov.cn, accessed on 5 September 2023) provided the original data for this study. The WBOD covers global economic, demographic, environmental, and other fields of data, aiming to promote global data sharing and development research. The BP Statistical Review of World Energy focuses on the energy sector, providing key data on global energy production, consumption, and other data over the years, helping energy research and analysis; the official website of the NBSC serves as an authoritative platform for the release of statistics and information and contains Chinese macroeconomic and social development data for all sectors. The official website of the National Bureau of Statistics of China, as an authoritative statistical information release platform, contains macroeconomic and social development data of various industries in China. The China Belt and Road Portal focuses on information related to the construction of the Belt and Road, covering data on trade, investment, and cooperation projects, which provides strong support for the study. For some missing data, linear interpolation and mean summation methods were used to complete the dataset.

The principle of the “interpolation method” is to estimate the missing values based on the changing trends of the existing data. Taking time-series data as an example, the data of “PM2.5 air pollution, mean annual exposure” for the Belt and Road countries have only been updated to 2019, and the data for 2020 are missing. By analyzing the characteristics such as the growth rate and the change slope of these data from 2011 to 2019, we constructed a linear function model and then calculated the missing value for 2020 to ensure the continuity and the rationality of the trend in the time dimension. The “mean value of both sides” fills in the missing values with the average of the adjacent two years of similar data. As can be seen from

Table 3. Data processing information.

Indicators	Available Data	Required Imputation	Number of Countries	Processing Methods
GDP per capita growth (annual %)	2011–2020	Yes	1	Mean value of both sides
Ratio of value added in services to GDP	2011–2020	Yes	1	Mean value of both sides
Government expenditure on education, ratio of GDP	2011–2020	Yes	15	Interpolation method and/or mean value of both sides
Adjusted savings: energy depletion (% of GNI)	2011–2020	Yes	1	Mean value of both sides
PM2.5 air pollution, mean annual exposure	2011–2019	Yes	54	Interpolation method

, this method was used for data interpolation of all data except “PM2.5 air pollution, mean annual exposure”. This method is mainly applicable when there are definite statistical data at both ends of the missing value. Under this premise, the average of the data in the two adjacent years is used to fill in the missing value. Through the rational application of these two methods, we successfully constructed a complete dataset, laying a solid foundation for further in-depth research.

2.4. Evaluation Method

This research measures the degrees of green development in the BRI countries using a mix of the entropy weight approach and the TOPSIS approach. Prior to assigning weights to each indication, all indicators are first normalized using the entropy weight approach. Next, each nation’s degree of green development is ranked and quantified using the TOPSIS methodology. The measuring findings of green development levels are made more objective and logical by combining the advantages of both approaches [40].

In the entropy weight method, indicator weights are derived from the data variation of each measurement indicator, minimizing subjective biases in the weighting process. Based on the impact of the indicators on green development, the indicators were first categorized into positive and negative indicators. Secondly, the data of the secondary indicators were standardized and processed as follows [41]:

$$\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e} \mathrm{i}\mathrm{n}\mathrm{d}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{o}\mathrm{r}:X_{ij}={\displaystyle \frac{x_{ij}-min\left\{x_{ij}\right\}}{max\left\{x_{ij}\right\}-min\left\{x_{ij}\right\}}}$$

(1)

$$\mathrm{N}\mathrm{e}\mathrm{g}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e} \mathrm{i}\mathrm{n}\mathrm{d}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{o}\mathrm{r}:X_{ij}={\displaystyle \frac{max\left\{x_{ij}\right\}-x_{ij}}{max\left\{x_{ij}\right\}-min\left\{x_{ij}\right\}}}$$

(2)

where X_ij represents the standardized value of the j-th indicator for the i-th evaluation object.

Then, the entropy weight method is applied to determine the weights of the indicators.

$$w_j={\displaystyle \frac{1-e_j}{{\sum }_{j=1}^p\left(1-e_j\right)}}$$

(3)

The variables e_j, w_j, and n represent the entropy, weight value, and number of items to be valued, respectively, of the j-th indication.

The TOPSIS approach offers the benefits of straightforward computation and realistic findings by quantitatively ranking each measurement item by comparing its relative distance to the best and worst solutions. First, the normalized entropy weight matrix is calculated; next, the optimal and worst solutions are identified [42]. Next, these solutions’ Euclidean distances to the target object are calculated. Finally, the comprehensive evaluation index is calculated:

$$C_j={\displaystyle \frac{r_j^{-}}{\left(r_j^{+}+r_j^{-}\right)}}$$

(4)

The nearer an object is to the optimal scheme, represented as Cj, the higher the value it has. And a higher value indicates a better-performing evaluated object.

Spatial autocorrelation analysis can effectively identify the spatial correlation of green development in the countries along the Belt and Road, determine whether there are significant spatial clustering or dispersion characteristics, and reveal spatial heterogeneity, which mainly includes global spatial autocorrelation and local spatial autocorrelation. Among them, global autocorrelation is used to measure the distribution characteristics of the variables in the whole regional space, which is generally measured with the help of Moran’s I index. When the global Moran’s I is in (0, 1], it indicates positive spatial correlation, and the level of green development is significantly concentrated; when the global Moran’s I is in [−1, 0), it indicates negative spatial correlation, and the level of green development is significantly discrete, and when the global Moran’s I = 0, it indicates a random distribution, and the samples are independent of each other [43]. The calculation formula is as follows:

$$I={\displaystyle \frac{\sum _{i=1}^n\sum _{j=1}^n{w}_{ij}\left(X_i-\overline{X}\right)\left(X_j-\overline{X}\right)}{w_{ij}}}$$

(5)

$$s^2=\sum {\left(X_i-\overline{X}\right)}^2/l$$

(6)

where X_i and X_j are the observed values for regions i and j, respectively; there $\overline{X}$ is the mean value; l is the total number of samples; and there $w_{ij}$ is the spatial weight matrix.

Local autocorrelation reflects the correlation between two neighboring units and describes the local spatial differences in the region, and the local Moran index is used to reflect the local autocorrelation of the study object by LISA agglomerative plot with Moran scatter plot [44]. The calculation formula is as follows:

$${ I}_i=x_i{\sum }_j{w}_{ij}{x}_j$$

(7)

where I_i is the local Moran’s I index for country i and x_i and x_j are normalized forms of the green development levels for regions i and j, respectively and $w_{ij}$ is spatial weight matrices.

The Thiel index is mainly applied for analyzing regional disparities or inequality. In this research, the Thiel index is utilized to explore the variations in green development levels both within and between the Belt and Road regions. The formula is as follows [45]:

$$T={\displaystyle \frac{1}{N}}{\sum }_{i=1}^N{\displaystyle \frac{y_i}{\overline{y}}}\mathrm{l}\mathrm{o}\mathrm{g}\left({\displaystyle \frac{y_i}{\overline{y}}}\right)$$

(8)

In the formula, T denotes the Thiel index, which takes values ranging from 0~∞, with 0 indicating a uniform distribution, in which the higher the value the greater the degree of inequality; N represents the number of countries; y_i represents the green development index of nation i; and $\overline{y}$ represents the average index of the N countries’ green development.

Furthermore, the total differences may be broken down into intra-group (T_w) and inter-group (T_b) differences using the Thiel index. The formula is as follows [45]:

$$T={\displaystyle \frac{1}{n}}{\sum }_{k=1}^K\sum _{i\in n_k}{\displaystyle \frac{y_i}{\overline{y}}}log{\displaystyle \frac{y_i}{\overline{y}}}={\sum }_{k=1}^K{\displaystyle \frac{n_k}{n}}{\displaystyle \frac{y_k}{y}}{\displaystyle \frac{1}{n_k}}\sum _{i\in n_k}{\displaystyle \frac{y_i}{y_k}}log{\displaystyle \frac{y_i}{y_k}}+{\displaystyle \frac{1}{n}}{\sum }_{k=1}^K\sum _{i\in n_k}{\displaystyle \frac{y_k}{\overline{y}}}log{\displaystyle \frac{y_k}{\overline{y}}}$$

(9)

$$T=T_w+T_b={\sum }_{k=1}^K{f}_k{\displaystyle \frac{y_k}{\overline{y}}}T\left(z^k\right)+{\sum }_{k=1}^K{f}_k{\displaystyle \frac{y_k}{\overline{y}}}log{\displaystyle \frac{y_k}{\overline{y}}}$$

(10)

The formula indicates that countries are partitioned into K groups (k = 1, 2, …, K); the number of countries within a group is denoted by n_k(${\sum }_{k=1}^K{n}_k=N$). y_k represents the average value of the green development level for countries in group k; f_k is the proportion of the number of samples in group k to the total number of countries; and T(z^k) is the Thiel index of group k.

3. Results

3.1. Overall Characteristics of the Level of Green Development in Countries Along the BRI

Drawing on the accessible data, we computed the green development levels of 54 BRI countries (Figure 2). The green progress of each nation is vital for enhancing green development across the Belt and Road region and serves as a key foundation for formulating targeted policies and enabling effective cooperation. Figure 2 graphically represents the green development levels of these countries in 2011, 2014, 2017, and 2020.

Regarding spatial distribution patterns, countries with high green development levels exhibit both spatial clustering and point-like distribution. Clustering is mainly concentrated in the Middle Eastern European region, while point-like distribution is mainly found in the West Asia–Middle East region and the ASEAN region. In terms of dynamics, in 2011 and 2014, the distribution pattern was “higher in the east and lower in the west”. By 2017 and 2020, this pattern shifted to “higher in the west and lower in the east”. This shift is mainly because Middle Eastern European countries have actively adjusted their economic structures and vigorously developed green industries in recent years. For example, Poland, the Czech Republic, and other countries have made substantial progress in new energy vehicle manufacturing, clean energy power generation, and other fields, significantly boosting their green development levels. Consequently, there has been a geographical shift in the center of gravity of green development.

Considering the overall trend, the average green development levels of Belt and Road countries in 2011, 2014, 2017, and 2020 are 0.242, 0.262, 0.301, and 0.280, respectively, showing an overall upward trend, with a more notable growth from 2011 to 2017. This indicates that, as global environmental awareness continues to rise, BRI countries increasingly recognize the importance of green development. The relevant policies they have formulated and implemented, such as setting up special funds for green development and encouraging enterprises to use clean energy and improve resource efficiency, have yielded remarkable results. However, the average value slightly decreased in 2020, mainly due to the huge impact of the COVID-19 pandemic on the global economy. This not only affected national economic development but also foreign trade, potentially reducing investment in green development.

Specifically, in 2011 and 2014, Qatar ranked highest in green development levels. In 2017 and 2020, Singapore replaced Qatar as the country with the highest green development level (Figure 3). Qatar’s success largely stems from its substantial investment in renewable energy research and development, supported by its strong financial resources. The country has actively engaged in the solar energy, wind energy, and other clean energy fields, establishing numerous large-scale research and development centers and demonstration projects. Simultaneously, Qatar has formulated and strictly enforced a series of environmental protection policies, imposing extremely high requirements on enterprise production emission standards and resource utilization efficiency, thus promoting the development of economic activities towards high efficiency and low pollution. Singapore, benefiting from its geographical advantages, has amassed a large amount of initial capital through long-term development. Based on this, the country vigorously develops high-tech, low-carbon industries, attracting a large number of the world’s top environmental protection technology enterprises and professionals, providing strong technical and intellectual support for green development. In contrast, many countries, such as Egypt, Armenia, and Iran, still have a relatively low level of green development. This is mainly due to their single-structure economies, economic development models, and low technological levels, resulting in limited financial and technological capabilities, making it difficult to allocate sufficient resources to green industry development. The inability to support large-scale environmental improvement measures has led to slow progress in green development.

It is also remarkable that, despite the significant global impact of the COVID-19 pandemic at the end of the study period, 40 countries achieved positive growth in green development levels during the study period, indicating that national policies and initiatives related to green development can still have a positive effect. Hungary witnessed the largest increase in green growth, attributed to a series of environmental and green development policies. Mongolia, on the contrary, experienced a significant decline in green development. Mongolia’s economy features a single-track development approach and over-reliance on mineral exports. Since 2011, global mineral prices have fluctuated, and Mongolia’s economy has remained sluggish, with its total trade volume dropping sharply. Economic hardships have led to a substantial reduction in government and corporate investment in environmental protection and green development, hindering the progress of related projects and causing a significant decline in the green development level [46].

3.2. Spatial Correlation of Green Development Levels in Countries Along the BRI

We conducted a global autocorrelation analysis of the spatial correlation characteristics of the green development level of the BRI countries and tested the credibility of the analysis through the p-value and Z-value, and the results are shown in Figure 4. The results show that Moran’s I index, a measure of green development level of BRI countries, is positive from 2011 to 2020. Except for 2011, 2012, and 2014, when the Z-value is <1.65, the Z-value of the other years is >1.65, with a credibility of 90%, and from 2016 to 2020, the p-value is <0.05, and the Z-value is >1.96, with a credibility of 95%, which indicates that the green development level of the BRI countries shows a more significant positive spatial distribution. Moran’s I index shows an overall fluctuating upward trend, of which the years 2011–2014, 2014–2017, and 2017–2019 all show an increasing and then decreasing trend. Within the years with a credibility level of 90% and above, the Moran’s I index was the smallest in 2013, indicating the weakest agglomeration in that year, and the Moran’s I index was the largest in 2020, indicating the strongest agglomeration in that year.

By analyzing the spatiotemporal transition characteristics of Local Indicators of Spatial Association (LISA) (Figure 5), we found that the United Arab Emirates and Bahrain in West Asia and the Middle East maintained long-term hotspot clustering features from 2011 to 2017, indicating that these two countries sustained relatively high levels of green development over an extended period, thereby forming persistent hotspot clusters. With the accelerated growth of green development levels in Middle Eastern Europe, hotspots shifted to this region, and countries such as Poland, the Czech Republic, and Slovakia exhibited hotspot clustering characteristics. This suggests that these countries enhanced their green development levels through the adoption of green development strategies and measures, while mutually influencing and promoting neighboring countries, forming new hotspot regions. The Central Asian countries, along with neighboring Georgia and Azerbaijan, maintained long-term coldspot clustering characteristics, indicating that these countries face numerous challenges in green development, with relatively lower development levels. Their geographical, economic, and environmental similarities contributed to the formation of stable coldspot clusters in green development levels. Additionally, Turkey and Pakistan also exhibited coldspot clustering features at certain times. The “low-high clustering” characteristic only appeared in specific years. Subsequently, as China and India continuously improved their green development levels and Mongolia’s green development level declined, the “low-high clustering” feature became less prominent. On the other hand, Malaysia, with its relatively lower green development level in ASEAN, displayed a long-term “low-high clustering” pattern, reflecting its lag compared to neighboring countries. However, as Malaysia’s green development level improved and the gap with surrounding nations narrowed, this characteristic gradually diminished. Similarly, the “high-low clustering” feature briefly emerged in Georgia and Azerbaijan, indicating significant disparities in green development levels between these two countries during specific periods. Later, as Georgia’s green development level declined, the gap with neighboring countries gradually narrowed, and this characteristic also faded.

3.3. Assessment of Differences in Levels of Green Development

We calculated and analyzed the Theil index to explore the influence of within-group and between-group differences on the overall disparities in the green development levels of BRI countries. A higher Theil index implies a larger gap in the green development levels among countries, while a lower index indicates more equal green development levels. As depicted in Figure 6, the variation in green development levels among West Asia–Middle East countries is notably higher than in other regions. This is mainly because countries like Israel, Qatar, and the United Arab Emirates in this region have sustained a relatively high green development level and either maintained it well or enhanced it to some degree. Conversely, many West Asia–Middle East countries have low green development levels. By 2020, five of the bottom-ranking countries are from this region. Consequently, the fluctuation in green development levels within this region is much greater than in others.

Regarding trends, the internal gap in the Middle Eastern European and ASEAN regions shows a fluctuating upward tendency, while that in other regions exhibits a fluctuating downward trend. This is because, in both the Middle Eastern European and ASEAN regions, some countries have experienced a relatively rapid growth in green development levels, while others have grown more slowly. In the Middle Eastern European region, Hungary and Estonia have witnessed an almost continuous increase in green development levels over the past decade. In contrast, countries such as Belarus and Serbia have had extremely slow growth rates and achieved only minimal improvements. In the ASEAN region, Singapore’s green development level has long been among the highest in the entire Belt and Road region and continues to increase steadily despite some fluctuations. Conversely, countries like Myanmar and Timor-Leste have seen negative growth in their green development levels. This pattern of development has widened the internal gap in the Middle Eastern European and ASEAN regions. Additionally, the regional disparity in East Asia decreased significantly in 2012–2013, mainly due to a substantial decline in Mongolia’s green development level.

To delve deeply into the degree and main sources of differences in green development among BRI countries, we calculated and decomposed the overall differences. The results are presented in Figure 7. From 2011 to 2020, the Theil index of the green development level in BRI countries ranges from 0.044 to 0.049, suggesting that the overall difference is relatively small. Upon further decomposition of the overall variation, it becomes evident that the impact of intra-group variation on the overall variation is substantial. Despite some fluctuations during this period, the proportion of intra-group variance has consistently remained above 85.96%. This implies that the intra-group difference is a key factor contributing to the overall difference in the green development level of BRI countries. Notably, although the contribution of inter-group differences to the overall differences is relatively minor, the inter-group differences display a remarkably upward trend from 2011 to 2020. This phenomenon indicates that the disparities in green development levels among BRI countries are gradually widening.

When conducting an in-depth analysis in combination with the evolving correlations in green development indicators and the practical progress of the Belt and Road Initiative, the underlying causes can be attributed to the following mechanisms: As the BRI steadily advances, member states actively engage in extensive and in-depth green development cooperation with neighboring countries. Such cooperation significantly promotes intra-group resource sharing, technological exchange, and the mutual adoption of best practices. As a result, the green development gaps within most regions are narrowed, and consequently, the contribution rate of within-group differences to overall disparities decreases. Simultaneously, due to the diverse resource endowments, industrial structures, policy orientations, and different levels of participation in BRI green development cooperation across regions, inter-group disparities show a fluctuating upward trend under the combined influence of multiple factors.

4. Discussion

4.1. Drivers and Constraints of Green Development in Different Countries

Based on the results of the green development assessment, it can be seen that the level of green development in each country is characterized by various types and distinctive features.

The first category is that of multifaceted synergy-driven countries, represented by Singapore and Israel. These countries have accumulated strong economic strength through long-term development and have significantly raised their scientific and technological level through sustained investment in scientific research, thereby promoting their green development [47,48]. For example, Singapore has significantly improved the efficiency of green economy transformation through policy synergies between digital technology and financial services [25], while Israel has formed a cross-industry synergistic network based on the pivot of agricultural science and technology and clean energy technological innovations, confirming the multiplier effect of science and technology inputs on green development. This “economy-technology” dual-wheel drive mechanism has enabled both countries to realize low-carbon technological breakthroughs and industrial structure upgrading while maintaining economic growth. The second category is the single-economy-driven type, such as Qatar and the United Arab Emirates, which rely mainly on abundant resources to reach a high level of economic development, but have a single model of development, relying on the extraction of resources for export to accumulate wealth, which is susceptible to price fluctuations and leads to a fragile economic structure. In the long run, this monolithic development model will pose a potential threat to green development. The third category is the double pressure type of economy and environment, such as China, Russia, and other countries. These countries face the double pressure of economic development and environmental protection. For example, China has experienced large-scale industrialization and urbanization and has a large manufacturing industry [25]; Russia has maintained its development by virtue of its rich resources and advantages in traditional industries such as energy and military industry. However, rapid economic development has brought about higher pollutant emissions, which have put greater pressure on the environment and become a key factor restricting green economic development. Although both countries have taken a series of environmental governance and green transformation initiatives, many difficulties and obstacles on the way forward still need to be overcome in order to realize the coordinated development of the economy and the environment. The fourth category is that of economic and technological constraints, such as those of Tajikistan, Kyrgyzstan, and Afghanistan. These countries have low economic and technological levels and a single industrial structure [25]. Although some countries are rich in resources, they have limited technological level, weak extraction and processing capacity, and can only export low-value-added primary products. Due to the lack of diversified industrial support, their risk-resistant capacity is poor, and their weak economic foundation limits investment in education, scientific and technological research and development, environmental protection, and other areas, forming a vicious circle and making the level of economic development a constraint on their level of green development.

4.2. Analysis of the Causes of Green Development Differences in BRI Regionals

Larger regional disparities imply more unbalanced green development among countries within a region. Take the West Asia–Middle East region as an example; it shows much higher variations in green development levels than other regions. This is mainly because countries such as Israel and Qatar have long maintained high green development levels. They have used their unique strengths and continuous investments in this area to make further progress. Conversely, many other countries in the region are far behind. For example, in 2020, five of the ten countries with the lowest green development rankings were from the West Asia–Middle East region, leading to significant intra-group disparities.

Substantial differences in the growth rates of green development among countries within regions have caused the internal gaps to widen. In Middle Eastern Europe, for instance, Hungary has energetically promoted environmental and green development policies, investing a great deal in renewable energy. This has effectively spurred the growth of green industries, keeping its green development on an upward path. Estonia has centered on a digitally enabled green economy. By using information technology to optimize energy management and promote green production, it has achieved remarkable results. In contrast, in countries like Belarus and Serbia, where traditional high-energy-consumption and high-pollution industries are still dominant, there are serious technical and financial obstacles in the transition to green industries, resulting in slow green development growth. In the ASEAN region, Singapore, taking advantage of its strategic location and strong long-term economic base, has vigorously developed high-tech and low-carbon industries. It has attracted top global environmental technology companies and professionals, enabling continuous innovation and maintaining a high level of green development. However, countries like Myanmar and Timor-Leste, restricted by lower economic development levels and an over-reliance on single-industry structures, show slow or even negative growth in green development. These large differences in green development growth rates between countries have worsened intra-regional disparities, increasing the gap in green development levels in both the Middle Eastern European and the ASEAN regions.

Severe fluctuations in the green development levels of individual countries within a region can cause significant changes in regional disparities. For example, the East Asia region saw a notable reduction in disparities from 2012 to 2013. This was mainly because Mongolia’s green development level dropped sharply due to an economic downturn, while China and Russia maintained relatively stable development. Mongolia’s sudden decline narrowed the gap between it and other East Asian countries, thus reducing the overall disparity in the region. This shows how local setbacks or surges in green development can have a large impact on regional inequality patterns.

Regional disparities in green development levels arise from differences in countries’ resource endowments, their participation in the BRI, and the spatial proximity effects that affect their development paths. Specifically, differences in green development levels are mainly due to different resource endowments and the effectiveness of policy implementation and adaptation across regions. Countries with higher green development levels usually have stronger economic bases and use their unique advantages to formulate and carry out suitable green development policies. A decomposition analysis of regional disparities along the BRI shows that intra-regional differences are dominant but gradually decreasing, while inter-regional differences are increasing. This change is driven by two main factors: There is uneven participation in BRI green development cooperation with some countries actively participating in BRI collaborations, seizing opportunities to boost their green development, while others are less involved. Constraints on inter-regional cooperation–geographical distance, political relations, and cultural differences limit deeper cooperation between regions compared to within-region cooperation, increasing inter-regional disparities. This changing trend has important implications for formulating green development policies and cooperation models within the BRI framework.

4.3. Recommendations for BRI Countries to Advance the Level of Green Development

4.3.1. Based on the Realities of the Country and the Precision of the Policy

Based on the assessment results and data, this study classifies the green development models of the BRI countries into four categories and puts forward policy recommendations accordingly:

(a) Multi-dimensional synergy-driven countries with good economic and scientific and technological foundations should strengthen “economy-technology” synergy, which on the one hand, increases the investment in emerging green technology, including the layout of artificial intelligence and environmental protection fusion; quantum computing to help energy optimization and other cutting-edge research and development; and supporting universities and research institutions to carry out projects and training of talents, and on the other hand, expands policies to guide the transformation and application of green science and technology achievements to encourage business practices and to promote the new momentum of the green industry.

(b) Single-economy-driven countries should accelerate economic diversification, reduce dependence on a single resource, increase investment in clean energy research and development and application, build research and development centers, introduce advanced foreign technologies, increase the proportion of clean energy in the energy consumption structure, and build a sustainable green development model.

(c) Economy–environmental dual-pressured countries need to deepen environmental governance and green transformation. In terms of environmental governance, it is necessary to increase the investment in environmental protection infrastructure, improve sewage treatment and garbage disposal facilities, and strengthen environmental supervision and law enforcement; in terms of green transformation, it is necessary to rely on scientific and technological innovation to enhance the greening of industries and encourage enterprises to adopt cleaner production technologies to reform traditional industries, promote industrial upgrading, and realize the coordinated development of the economy and the environment.

(d) Economically constrained countries should strengthen international cooperation to introduce technologies and funds, cooperate with international financial institutions to seek special green loans, carry out technical cooperation with developed countries or enterprises, and introduce advanced resource extraction and processing technologies to improve the efficiency of extraction and enhance added-value. For example, the establishment of resource deep-processing industrial parks to attract foreign capital and technology will drive the development of the local economy and break the vicious circle of economic and green development.

4.3.2. Building Regional Synergistic Governance Mechanisms

Regional synergistic governance mechanisms need to be constructed to stabilize the cooperation framework, such as the establishment of a transnational green development information platform and a joint monitoring and early warning system to promote data-sharing and risk-sharing; at the same time, policy uncertainties should be reduced through the signing of a multilateral green investment agreement. On this basis, the clustering effect should be optimized to promote the transfer of experience, for example, by setting up a green technology hub in a high-clustering area (Middle Eastern Europe), exporting technology and management experience to a low-clustering area (Central Asia), and implementing a “twinning” scheme that guides countries with high levels of development in providing targeted support to countries with low levels of development, so as to avoid increasing polarization.

4.3.3. Focus on Internal Variance Governance and Dynamic Adjustment

In view of the fact that regional differences are mainly due to internal imbalances, priority should be given to narrowing intra-regional gaps, for example, through the establishment of a special fund for green development, with a focus on supporting green infrastructure in lagging regions such as Central Asia and South Asia, and coordinating the development of regionally harmonized standards to reduce policy friction. In addition, a dynamic monitoring and adaptive policy adjustment mechanism should be established, a green development relevance index should be released regularly to quantitatively assess progress, flexible policy tools should be designed, and the transition path should be simulated through multilateral scientific research cooperation to support scientific decision-making, with the ultimate goal of curbing the widening of disparities in the short term and promoting balanced and synergistic development in the long term.

5. Conclusions

In the context of the global drive for sustainable development, this research constructs a multi-dimensional evaluation index system for green development levels. The system encompasses socioeconomic development, resource utilization efficiency, and environmental protection dimensions. It provides a scientific and systematic framework for analyzing green development in BRI countries. By exploring the core aspects of green development, this study offers a reference for future research and evidence-based policy-making guidance. The research findings indicate that BRI countries with higher levels of green development exhibit two distinct spatial distribution patterns: clustered and dispersed. Clustered patterns are predominantly observed in Middle Eastern Europe, while dispersed patterns are found in ASEAN and West Asia–Middle East regions. Notably, despite global economic slowdowns and the COVID-19 pandemic during the study period, 83.3% of the countries demonstrated improvements in green development. With deepening cooperation in green development, the spatial association of green development levels among the BRI countries has strengthened. However, due to varying growth rates across nations, localized clustering patterns have undergone corresponding changes. Overall regional disparities in green development levels along the Belt and Road were relatively small, with intra-group differences being the primary source of such disparities. In conclusion, this study comprehensively reveals the multi-dimensional characteristics of green development in BRI countries, providing robust empirical and theoretical support for advancing national green development strategies and guiding the construction of a green BRI framework.

There are still some issues that need to be addressed in future research. The main challenges are as follows: First, the limited channels for data acquisition have resulted in incomplete coverage of the research scope and constrained timeliness. Second, there is an insufficient inclusion of abstract driving factors such as the level of participation in the Belt and Road Initiative. Third, although the entropy weight–TOPSIS method has its own advantages, it still has some problems. The entropy weight method mainly relies on the degree of data variability to determine the weights, which leads to its excessive dependence on data as well as the possibility of ignoring the inherent importance of the indicators; the TOPSIS method, which sorts the samples by calculating the distance between each sample and the ideal solution and the negative ideal solution, is also more sensitive to small changes in the indicator data. To address these three issues, subsequent studies plan to adopt the following solutions:

1. Expand data sources through remote sensing technology: by processing and extracting remote sensing images, data such as vegetation coverage and water resource reserves will be obtained to further improve the indicator system and enhance its accuracy;

2. Incorporate the “Five Connectivity” index: This index, comprising policy coordination, infrastructure connectivity, unimpeded trade, financial integration, and people-to-people bonds, will be used to measure the participation of BRI countries. This will further analyze the significance of the initiative in promoting green development among these nations;

3. The data dependence of entropy weighting method can be reduced by introducing subjective weights of experts and data smoothing techniques. Secondly, the important matrix of indicators can be constructed, and the weight adjustment coefficients can be set according to the theoretical framework to take into account the intrinsic importance of the indicators. Lastly, in the future, we will try to use Box–Cox transform, rank transform, and other robust data transformations, as well as the introduction of the fuzzy TOPSIS method to reduce the sensitivity of TOPSIS method to small changes of the data.