Analyzing the Spatiotemporal Pattern and Interaction of SDGs for Sustainable Development in Inner Mongolia

Abstract

Increased global climate change and contradictions in human–land relationships has aroused awareness of studies on regional sustainable development. Whether SDGs and their interactions are suitable for analyzing the fine scale of regional differentiation of sustainable development, especially in ecologically sensitive regions, was still in suspense. This study analyzed the inter-annual changes and regional clustering of SDGs and the interactions among SDGs, and among their indicators, at both provincial and municipal levels in Inner Mongolia. We found the sustainable development was relatively higher in the east than in the west. SDG5, SDG6, SDG10, SDG11, and SDG15 got cold spots in the west and hot spots in the east. For most SDGs and indicators, synergies outweigh trade-offs. Improvement measures should focus on the indices with strong synergies such as SDG3 and SDG7, and SOC2, ECO1, ECO4, and ECO5. Special attention should be put on those with trade-offs such as ECO15 and ENV20 to be aware of their opposite effects. SDG5 and ECO2 were the most important in the entire network that need attention. Considering only singular or a few SDGs may not be feasible when assessing sustainable development because other goals or indicators may play roles. Reasonable improvements for sustainable development needed to clarify synergies and trade-offs among SDGs and indicators.

1. Introduction

Since the 21st century, in the process of rapid social and economic developments, urban expansion, deforestation, farmland expansion, grassland degradation, and desertification have caused significant changes in land use and population distribution. Problems which greatly hinder regional sustainable development such as contradictions in human–land relations, ecological dysfunction, excessive resource consumption, and the intensification of the global greenhouse effect have become increasingly apparent [1]. The problem of uneven and insufficient development between regions has also become increasingly prominent. Therefore, sustainable development has become the current and future core development goal both domestically and internationally.

On the occasion of the expiration of the Millennium Development Goals (MDGs) of United Nations member states in 2015, the draft “Changing Our World: 2030 Agenda for Sustainable Development” [2] was approved at the summit, which sets sustainable development goals (SDGs) and aims to provide guidance for sustainable development in countries from 2015 to 2030. On the basis of MDGs, SDGs have established 17 main goals and 169 specific goals and set multiple detailed indicators for each goal, comprehensively covering the three major fields of society, economy, and environment [3,4] (Figure 1). SDGs provided a unified measurement system for achieving comprehensive and sustainable development of social progress, economic development, and environmental friendliness. In 2016, the United Nations’ Sustainable Development Action Network launched standards for evaluating the implementation of SDGs by countries. This evaluation system displays the overall progress of each target on a global scale through four color codes: red, orange, yellow, and green [5]. SDGs have been applied to assess regional sustainability development in many countries such as USA, China, the United Kingdom, India, Japan, etc. [6]. Due to the fact that SDGs are mainly applicable to guiding sustainable development on a global scale, special regions should construct a suitable indicator system based on their actual situation. In recent years, many studies have focused on achieving SDGs in different areas. For example, Xu et al. [7] quantitatively evaluated the overall scores of sustainable development goals in China over the past 15 years and assessed the overall scores of sustainable development goals in each province. The results showed that although the scores of each region have increased over the years, there are significant spatial and temporal differences among different regions. Wang et al. [8] explored the spatiotemporal changes in environmental sustainability at the national and provincial levels in China from 2010 to 2018. Dong et al. [9] implemented a local assessment framework for SDG11.6.2 and quantitatively evaluated the sustainable development status of air quality in Beijing. In addition, some scholars combined national major issues such as ecological civilization construction, beautiful China construction, and green development with SDGs in their research on sustainable development and constructed various evaluation index systems with Chinese characteristics. For example, He et al. [10] constructed an evaluation index system for ecological civilization construction in Sichuan Province based on the SDGs and the Pressure State Response (PSR) model. Cheng et al. [11] combined SDG indicators and beautiful China evaluation indicators to construct an indicator system centered on water resource carrying capacity from the five major systems of society, economy, ecology, coordination, and water resources.

Formulating a sustainable development indicator system is critical and effective for analyzing regional sustainability development. SDGs were generally accepted by researchers to compose the indicator system [5,6]. However, interactions among SDGs are complicated. The actions taken by one goal may strengthen or offset those taken by another goal [12,13]. SDGs constitute a globally applicable and comprehensive evaluation indicator framework. When formulating a sustainable development indicator system based on SDGs, due to the susceptibility of the composite index to changes in its internal indicators, there may be mutual offsetting among different goals. This means that even if the overall score increases, only some goals may improve, while the deterioration of other goals may be offset or ignored [5]. The number of evaluation indicators involved also affects the relationship between SDGs. With the increase in indicator amount, the interaction between SDGs shows non-linear changes. Their interaction degree varies among different levels of sustainable development. With the increase or decrease in sustainable development, the interaction may be more positive or negative. For example, at a moderate level of sustainable development, there was generally a more positive interaction between SDGs [14]. The interaction may be direct or indirect. Considering both aspects has a positive significance for more accurate identification of their interaction relationships and the formulation of more comprehensive policy measures [15]. The scores for sustainable development goals are not spatially consistent between the overall and individual components. For example, the uneven development in different regions of China has led to inconsistent achievement of sustainable development goals. For the whole, this inconsistency may result in offsetting and masking the low scores of certain sustainable development goals in individual regions. Studying the differences and similarities between regional sustainable development goals can provide a reliable basis for managers to formulate related policies [16]. To achieve comprehensive sustainable development, it is necessary to consider all SDGs and their interactions. It is also beneficial for special goals in regional sustainability development. For example, considering SDG6 and SDG14 is of great significance for controlling water pollution in China [17]. Further consideration of the interaction between indicators of SDGs is conducive to more rational allocation of resources to important indicators [18].

However, currently, there is limited research on the interactions between different SDGs and the related indicators, especially in ecologically sensitive regions like Inner Mongolia at both provincial and municipal levels. This study constructed an indicator evaluation system suitable for Inner Mongolia based on SDGs, evaluated the sustainable development level and regional differences at the provincial and municipal levels, and analyzed the inter-relationships among SDGs and among indicators. The study aimed to provide theoretical basis and practical support for achieving regional sustainable development in Inner Mongolia.

2. Study Area

Inner Mongolia (97°12′~126°04′ E, 37°24′~53°23′ N) is located in the northern China (Figure 2). The total length of the boundary line is about 4221 km, and the total area is about 1.183 million km², accounting for 12.3% of China’s land area. It is the third largest provincial-level administrative region in China. The terrain of Inner Mongolia is mainly composed of plateaus, with rich and diverse types of landforms. The elevation of the entire region ranges from 85 m to 3526 m, with an average elevation of 1000 m. The main mountain ranges include the Greater Khingan Mountains, the Yinshan Mountains, and the Helan Mountains. Inner Mongolia has a vast territory and sparse population, with a wide span from east to west. The western region exhibits a temperate continental climate, while the eastern region has a mid-temperate monsoon climate. Winter is cold and dry, while summer is short and warm. Precipitation is mainly concentrated in summer, with strong winds in spring and autumn. Precipitation in autumn decreases and temperatures significantly decrease [19]. Inner Mongolia is rich in natural resources, especially grassland resources, with an available grassland area of approximately 680,000 km², making it the largest grassland area in China. Due to significant changes in temperature and precipitation from west to east, the distribution characteristics of desert, desert grassland, typical grassland, meadow grassland, and forest are presented in sequence (Figure 2). According to the conclusion of the second national wetland resource survey, the total area of wetlands in Inner Mongolia has reached 60,106 km², ranking third in the country [20]. Inner Mongolia is an important area for sustainability development because of the following reasons. Firstly, Inner Mongolia has abundant natural resources with many types of ecosystems and high biodiversity. It is an important ecological barrier in northern China and a typical ecologically fragile area, which is sensitive to global environmental changes. Problems such as grassland degradation, desertification, and drought have aggravated the situation. Secondly, its economic and social development has put pressure on the local environment [21]. In addition, Inner Mongolia’s grain production ranks among the top in China, making it the main grain producing area, and it is known as the “granary outside the Great Wall”. Therefore, its food security is crucial to China. Thus, a comprehensive sustainability development of social–economic–environmental systems is crucial to this area.

3. Materials and Methods

3.1. Principles for Constructing a Sustainable Development Evaluation Indicator System

To standardize the selection of indicators, this study summarized the following principles for constructing an indicator system for assessing the sustainable development based on previous studies [22,23]: (1) Comprehensive and accessible principles—indicators that are easy to obtain and calculate, with no repetition are preferable during the selection. (2) The principles of scientificity and comparability—the system should be constructed based on clarifying the definition, calculation methods, and reliable data sources for each indicator to ensure comparability between different years. (3) The principles of systematicity and hierarchy—based on SDGs, the selected indicators should not only meet the specific requirements proposed in SDGs, ensuring that each indicator is independent and logical, but also fully describe various aspects of the sustainable development subsystems in Inner Mongolia, and be as concise and hierarchical as possible.

3.2. Selection of Evaluation Indicators

Based on the above principles and previous research [8,23,24], we established an evaluation indicator system of sustainable development for Inner Mongolia with local characteristics (

Table 1. Evaluation indicator system of sustainable development in Inner Mongolia.

Subsystem	Goal	Indicator	Indicator Number	Role *
Society	SDG1	Population proportion covered by unemployment insurance (%)	SOC1	Positive
	SDG1	Insurance participation ratio (%)	SOC2	Positive
	SDG1	The proportion of education expenditure to fiscal budget expenditure (%)	SOC3	Positive
	SDG2	Grain yield per hectare (kg/ha)	SOC4	Positive
	SDG2	Growth rate of grain production (%)	SOC5	Positive
	SDG2	Per capita arable land area (km2/10,000 people)	SOC6	Positive
	SDG3	Basic medical insurance coverage rate (%)	SOC7	Positive
	SDG3	Engel’s coefficient (%)	SOC8	Negative
	SDG3	Death toll from traffic injuries per 100,000 people	SOC9	Negative
	SDG3	The proportion of local medical and health expenditure to fiscal budget expenditure (%)	SOC10	Positive
	SDG3	Number of people involved in medical services per 10,000 people	SOC11	Positive
	SDG3	Number of beds in medicine treatment institutions per 10,000 people	SOC12	Positive
	SDG4	Enrollment rate of school-age children (%)	SOC13	Positive
	SDG4	The ratio of the number of teachers to the number of students in higher education institutions (%)	SOC14	Positive
	SDG5	Mobile phone penetration rate (unit/100 households)	SOC15	Positive
	SDG5	The proportion of female employees (%)	SOC16	Positive
	SDG6	Popularity rate of rural sanitary toilets (%)	SOC17	Positive
	SDG7	Urban gas penetration rate (%)	SOC18	Positive
	SDG8	Work injury insurance coverage rate (%)	SOC19	Positive
	SDG11	Bus ownership per 10,000 people	SOC20	Positive
	SDG11	The ratio of built-up area growth rate to population growth rate (%)	SOC21	Positive
	SDG11	Population density (person/km2)	SOC22	Positive
	SDG11	Per capita construction land area (km2/10,000 people)	SOC23	Positive
	SDG17	Internet penetration rate (unit/100 households)	SOC24	Positive
Economy	SDG2	Per capita disposable income of rural residents (¥)	ECO1	Positive
	SDG7	Energy intensity (ton of standard coal/10,000 ¥)	ECO2	Negative
	SDG8	Real per capita GDP annual average growth rate (%)	ECO3	Positive
	SDG8	Per capita GDP (¥)	ECO4	Positive
	SDG8	Per capita salary of employed personnel (¥)	ECO5	Positive
	SDG8	Urban registered unemployment rate (%)	ECO6	Negative
	SDG8	Annual growth rate of tourism industry revenue (%)	ECO7	Positive
	SDG8	The proportion of tourism industry revenue to GDP (%)	ECO8	Positive
	SDG9	Highway passenger volume (10,000 people)	ECO9	Negative
	SDG9	Highway freight volume (10,000 tons)	ECO10	Positive
	SDG9	Number of researchers per million population	ECO11	Positive
	SDG10	Annual growth rate of per capita disposable income of urban residents (%)	ECO12	Positive
	SDG10	Annual growth rate of per capita disposable income of rural residents (%)	ECO13	Positive
	SDG10	The proportion of total wages of urban unit employees to GDP (%)	ECO14	Positive
	SDG11	Land use rate (%)	ECO15	Positive
	SDG17	The proportion of local general public budget revenue to GDP (%)	ECO16	Positive
	SDG17	The proportion of total exports to GDP (%)	ECO17	Positive
Environment	SDG2	Fertilizer application intensity (t/km2)	ENV1	Negative
	SDG6	Per capita water resources (m3)	ENV2	Positive
	SDG6	Popularization rate of safe drinking water (%)	ENV3	Positive
	SDG6	Urban sewage treatment rate (%)	ENV4	Positive
	SDG6	Water consumption per 10,000 yuan of GDP (m3/10,000 ¥)	ENV5	Negative
	SDG6	Water resource pressure (proportion of total water use to total water resources) (%)	ENV6	Negative
	SDG6	Change rate of water area (%)	ENV7	Positive
	SDG11	Harmless treatment rate of household waste (%)	ENV8	Positive
	SDG11	Annual mean concentration of PM2.5 (mg/m3)	ENV9	Negative
	SDG11	Per capita park green space area (m2)	ENV10	Positive
	SDG12	Per capita industrial wastewater discharge (tons/person)	ENV11	Negative
	SDG12	Per capita industrial sulfur dioxide emissions (kg/person)	ENV12	Negative
	SDG12	Total sulfur dioxide emissions per unit GDP (t/billion ¥)	ENV13	Negative
	SDG13	Carbon dioxide emissions (million tons)	ENV14	Negative
	SDG13	Total SO2 emissions (t)	ENV15	Negative
	SDG13	Carbon intensity per unit GDP (t/10,000 ¥)	ENV16	Negative
	SDG15	Forest coverage rate (%)	ENV17	Positive
	SDG15	The proportion of wetland area to total land area (%)	ENV18	Positive
	SDG15	Afforestation area (thousand ha)	ENV19	Positive
	SDG15	The proportion of desertified land to total land area (%)	ENV20	Negative
	SDG15	The proportion of forest ecosystem nature reserve area to nature reserve area (%)	ENV21	Positive
	SDG15	The proportion of grassland ecosystem nature reserve area to nature reserve area (%)	ENV22	Positive
	SDG15	The proportion of the area of wildlife nature reserves to the area of nature reserves (%)	ENV23	Positive

* Positive indicator promotes sustainability development and vice versa.

). Firstly, based on in-depth analysis of the natural ecological characteristics and the economic and social development status of Inner Mongolia, the suitability of various SDG indicators for the region was judged to ensure that the selected indicators were in line with the economic development level, social governance structure, and the resource and environmental conditions of the region. Secondly, the evaluation index system was constructed based on the sustainable development strategy and the region’s own development plan and goals. For example, the requirements proposed in Inner Mongolia’s 14th Five Year Plan include economic aspects, the targets for regional GDP and per capita income growth of residents. The construction of ecological civilization proposes to increase forest coverage and reduce energy consumption, and rural revitalization proposes to improve rural living standards, enhance agricultural modernization, and improve people’s livelihoods to ensure their living standards. The constructed evaluation indicators should not only accurately reflect the current development status of the region, but also effectively describe and monitor the changing trends of its key development needs. Thirdly, a potential evaluation index system for sustainable development that was in line with the actual situation in Inner Mongolia was constructed. Then, based on the specific situation of Inner Mongolia, some unreliable indicators were removed and some reliable indicators were supplemented, and the final sustainable development evaluation index system was established (Table 1).

The indicator system was divided into four levels: the first level consisted of three subsystems: social, economic, and environmental. The second level was SDGs corresponding to each subsystem. The third level contained the specific indicators selected for each goal. About 15 out of the 17 SDGs were selected, excluding irrelevant SDG14 (protecting marine ecology) and SDG16 (institutional justice). They included: 7 social goals such as SDG1 (poverty alleviation), SDG2 (hunger alleviation and food security), SDG3 (good medical conditions), SDG4 (quality education), SDG5 (gender equality), SDG7 (sustainable energy use), and SDG11 (sustainable cities); 4 economic goals such as SGD8 (decent work), SGD9 (infrastructure and sustainable industrialization), SDG10 (social equality), and SDG12 (sustainable consumption and production models); and 3 environmental goals such as SDG6 (clean water and environmental sanitation), SDG13 (addressing climate change), and SDG15 (sustainable use of terrestrial ecosystems). The SDGs were further classified into 24 social indicators, 17 economic indicators, and 24 environmental indicators (Table 1).

3.3. Data Collection and Process

The evaluation indicator data included environmental monitoring data, geospatial data, land use/land cover data, and socio-economic statistical data, from 2001 to 2020. The statistical data came from various statistical yearbooks, as well as various league city statistical yearbooks and statistical bulletins in Inner Mongolia. There may be differences in data from different sources, so we selected those with two or more data sets with the same values. Some indicator data can be directly obtained, while other data need to be calculated based on their actual meaning through the joint calculation of several other data items. The missing data were supplemented using the average method of neighboring years and the kernel grey prediction method. Land use/land cover data came from the data sharing service system of the Chinese Academy of Sciences (https://data.casearth.cn/, accessed on 4 August 2024) [25]. The geospatial data came from the interpretation results of Landsat TM/ETM remote sensing images from 2001 to 2020. The images were downloaded from geospatial data cloud websites (http://www.gscloud.cn/, accessed on 4 August 2024) and were interpreted to obtain classification results. According to the definition in the SDGs’ metabase (https://unstats.un.org/sdgs/metadata/, accessed on 4 August 2024), some required indicators such as the ratio of land use rate to population growth rate were calculated. The CO₂ emission data were derived from the carbon emission inventory of Chinese prefecture level cities compiled by Chen et al. [26], which was based on the revised version of China’s winter energy data in 2015. The popularization rate of rural sanitary toilets and the popularization rate of safe drinking water were calculated by Xu et al. [27].

3.4. Calculating SDGs Scores at the Provincial and League Levels

The scores of SDGs were calculated using arithmetic means [5]. To reflect the principle of “not leaving behind any goal” required by SDGs [7,14], this study assigned the same weight to each goal, and the calculation formula was as follows:

$${SDG}_{it}\left(N_i,N_{ij},x_{ijt}^{\prime }\right)={\Sigma }_{i=1}^{N_i}{\displaystyle \frac{1}{N}}{\sum }_{j=1}^{N_{ij}}{\displaystyle \frac{1}{N_{ij}}}{x}_{ijt}^{\prime }$$

(1)

where SDG_it represented the sustainable development score of region i in year t; N_i was the total number of SDGs in region i; N_ij represented the indicator number of SDG_j in region i; and $x_{ijt}^{\prime }$ represented the indicator scores of SDG_j in region i in year t.

3.5. Using SDGs Dashboards to Evaluate the Development Levels of Different Goals by Regions

The SDGs Dashboard is a method of utilizing existing data to display the overall implementation progress of 15 SDGs through a color-coding system of red, yellow, and green traffic lights. The SDGs scores were divided into four colors according to the interval, as shown in Figure 3. The green represented that the score was in the best range. Its score was higher than the sum of the average and standard deviation of the SDGs scores of all cities, indicating an excellent state of the sustainable development level. The yellow color represented the interval with the second-best score, which was lower than the sum of the average and standard deviation of all the SDGs scores, but higher than the mean SDGs score, indicating a good state of the sustainable development level. The brown color represented the interval with the second lowest score, which was lower than the mean SDGs score, but higher than the minus between the average score and standard deviation of all the SDGs, indicating a moderate state of the sustainable development level. The red color represented the interval with the worst score, which was lower than the minus of the average and standard deviation of all the SDGs scores, indicating a poor state of the sustainable development level.

3.6. Sen Trend Analyses and Mann–Kendall Test

Sen trend analysis is a robust non-parametric statistical method [8] that can be used to detect the trend of changes in long-term time series data. It has strong resistance to measurement errors and outliers. Unlike linear regression methods that require data to follow a normal distribution, Sen trend analysis estimates long-term trends by calculating the median of the data series. The Mann–Kendall test is a non-parametric statistical method that does not require data to satisfy the assumption of a normal distribution or linear trend. It can handle missing values and outliers in the data. Therefore, this method has been widely applied in the trend significance test of long-term time series data. The formula was as follows [28]:

$$\beta =Median\left({\displaystyle \frac{x_j-x_i}{j-i}}\right),1<i<j<n$$

(2)

where β represents the change slope of the sustainable development level, β > 0 indicating an upward trend in the sustainable development level and β < 0 indicating a decreasing trend in the sustainable development level; $x_i$ and $x_j$ represents time series data, representing the sustainable development scores for the ith and jth years, respectively; and n represents the length of the time series (2001–2020).

The statistical formula S for Mann–Kendall’s trend test was as follows:

$$S={\sum }_{i=1}^{n-1}{\sum }_{j=i+1}^{n}sgn(x_j-x_i)$$

(3)

where $sgn(x_j-x_i)$ was the symbol function, which was defined as

$$sgn\left(x_j-x_i\right)=\left\{\begin{array}{c}1, x_j-x_i>0\\ 0, x_j-x_i=0\\ -1, x_j-x_i<0\end{array}\right.$$

(4)

$$Var\left(S\right)={\displaystyle \frac{n\left(n-1\right)\left(2n+5\right)-{\sum }_{i=1}^{n}t(i-1)(2i+5)}{18}}$$

(5)

The calculation formula for the standardized statistic Z was

$$Z=\left\{\begin{array}{c}{\displaystyle \frac{S-1}{\sqrt{Var\left(S\right)}}},S>0\\ 0,S=0\\ {\displaystyle \frac{S+1}{\sqrt{Var\left(S\right)}}},S<0\end{array}\right.$$

(6)

where S was calculated by Formula (3). When n < 10, if S > 0, the sequence showed an upward trend; when S = 0, no trend; when S < 0, the sequence showed a downward trend. When n ≥ 10, the statistic S tended towards a normal distribution.

3.7. Spearman Correlation Analysis and Network Analysis

The range of Spearman correlation coefficient was [−1,1], which can reflect the monotonic relationship between indicators. According to the threshold of the correlation coefficient (r), the correlation between SDGs was divided into synergistic effects (r > 0.6), trade-off effects (r < −0.6), and unclassified effects (−0.6 ≤ r ≤ 0.6) [29]. This study set a p-value less than 0.05 to ensure statistical significance of the synergistic and trade-off relationship between SDGs and between the indicators [30].

Social network analysis was used to determine the interaction between SDGs and between the indicators. A social network is composed of a series of nodes and their interconnected lines, forming a complete network of relationships [31]. In a network structure, the node centrality is determined by calculating the number of other nodes directly connected to it. This indicator reflects the influence and importance of the node in the network. The higher the centrality of a node, the more it is directly connected to other nodes, and its position in the network becomes more significant [32]. The betweenness centrality determines the importance of a node by calculating the number of shortest paths that it is connected to. The greater the betweenness centrality of a node, the more significant its impact on other nodes [33]. Proximity centrality is evaluated by calculating the sum of the shortest path lengths from a node to all other nodes in the network, which is used to determine the importance of a node in the network. The higher the proximity centrality, the closer the node is to the center of the network, and the shorter the distance from other nodes [34]. The eigenvector centrality not only measures the number of connections between a node and other nodes, but also the importance of expected vector nodes. Therefore, a larger eigenvector centrality not only represents their own importance, but also represents the importance of other nodes associated with them [35]. The K-kernel evaluates whether the node is located in a group with high clustering in the network. The larger the K-kernel value, the more links the node has with other nodes, and the higher the cohesion between these nodes [35]. This study used Ucinet 6.5.6 software to calculate the node centrality, betweenness centrality, proximity centrality, eigenvector centrality, and K-kernel between various sustainable indicators in Inner Mongolia and used Gephi 0.10.1 software for visualization to construct an analysis network.

3.8. Hot and Cold Spots Analysis

This study used the Getis-Ord Gi* tool on the ArcMap platform to identify cold and hot spots in a region and obtain their p-values and z-values. Areas with a hot spot indicated high scores for SDGs in surrounding cities, reflecting high sustainable development for SDGs in the area. On the other hand, cold spots indicated low scores for SDGs in surrounding cities, reflecting low sustainable development for SDGs in the area. The confidence levels were classified based on statistical significance into three types: hot spot areas with 99%, 95%, and 90% confidence levels, cold spot areas with 99%, 95%, and 90% confidence levels, and areas with indistinct features. The calculation formula was as follows:

$$G_i^{\mathrm{*}}={\displaystyle \frac{\sum _{j=1}^n{w}_{i,k}{x}_j-\overline{X}\sum _{j=1}^n{w}_{i,j}}{S\sqrt{\frac{\left[n\sum _{j=1}^n{x}_j^2-{\left(\sum _{j=1}^n{w}_{i,j}\right)}^2\right]}{\mathrm{n}-1}}}}$$

(7)

$$\overline{X}={\displaystyle \frac{\sum _{j=1}^n{x}_j}{\mathrm{n}}}$$

(8)

$$S=\sqrt{{\displaystyle \frac{\sum _{j=1}^n{x_j}^2}{\mathrm{n}}}-{\left(\overline{X}\right)}^2}$$

(9)

where i and j represented region i and goal j (SDG_j), respectively. $G_i^{\mathrm{*}}$ was calculated as the z score, $x_j$ represented the score of SDG_j, n represented the number of cities, and $w_{i,k}$ represented spatial weight, calculated by the distance from city i to city k.

4. Results

4.1. Inter-Annual Change of SDG Scores at the Provincial Level

According to

Table 2. Sen slope trend and MK significance test of SDG scores from 2001 to 2020.

SDGs	Trend	Sen Slope (per Year)	Z Value
SDG1	Increasing	3.540 **	4.185
SDG2	Increasing	2.651 **	4.445
SDG3	Increasing	5.616 **	5.568
SDG4	Decreasing	−0.838 **	−2.727
SDG5	Increasing	3.055 **	5.808
SDG6	Increasing	3.135 **	4.185
SDG7	Increasing	5.873 **	5.483
SDG8	Increasing	2.613 **	5.353
SDG9	Increasing	5.003 **	5.353
SDG10	Insignificance	−0.998	−1.59
SDG11	Increasing	3.815 **	5.288
SDG12	Insignificance	1.34	1.395
SDG13	Insignificance	1.068	1.136
SDG15	Insignificance	−0.05	−0.487
SDG17	Increasing	1.672 **	2.693

Note: * means 0.05 significance level, ** means 0.01 significance level.

, most SDGs had significant upward trends in scores, while SDG4 had a certain downward trend. The SDG4 score decreased significantly from 2001 to 2020 with a smaller annual mean change rate of 0.82% than other scores of SDGs. However, its Sen slope was −0.838/a, which reached a very significant level, indicating a strong significant downward trend in SDG4 scores. The growth rate and average annual growth rate of SDG3 were the highest among all SDGs, with a score of only 4.65 in 2001, and 94.28 in 2020. Its lowest score appeared in 2002, only 2.18, and the highest score appeared in 2017, reaching 95.73. The Sen slope of the SDG3 score was relatively large and had passed the strongly significant test, reaching 5.616/a. SDG7 had the highest Sen slope of all the SDGs, reaching 5.873/a, and its growth rate was also significant, increasing from 19.31 in 2001 to 94.53 in 2020. The scored trends of SDG10, SDG11, SDG13, and SDG15 were not significant.

4.2. SDGs Scores at the Municipal Level in 2020

It can be inferred from Figure 4 that each city had at least one sustainable development goal as “red”. The cities with only one ‘red’ goal were Hohhot (SDG3), Baotou (SDG17), Chifeng (SDG8), Hulunbuir (SDG13), and Xilingol League (SDG3) located in the central and eastern regions. Cities with the most ‘red’ goals were Wuhai (SDG6, SDG11, SDG12, and SDG15) and Alxa League (SDG4, SDG11, SDG12 and SDG15) located in the western region. Thus, their sustainable developments were relatively poor. The cities with “red” scores in SDG1, SDG4, SDG6, SDG10, SDG11, SDG12, and SDG15 were all located in the western region. The SDGs in the four eastern cities scored the most in “green”. The cities with “green” scores in SDG6, SDG10, and SDG15 were all located in the eastern region. The sustainable development level in the western region was relatively uneven. In the case of having the most “red” SDGs, there were also more SDGs scoring “green”. The development in the central region was relatively balanced, with the most SDGs scoring “yellow” and “brown”.

4.3. Hot and Cold Spots of SDGs at the Municipal Level in 2020

Figure 5 shows the hot and cold spots of SDGs. It can be inferred from Figure 5 that SDG1, SDG2, and SDG13 did not have cold spots. Compared to other cities, cities with more SDGs with cold spots were Wuhai (SDG6, SDG7, SDG10, SDG11, SDG12, SDG15, and SDGs), Alxa League (SDG11, SDG12, SDG15, and SDGs), and Baotou (SDG5, SDG6, and SDG10), located in the western and central regions. Of all the cities, Wuhai had the most cold spots of SDG6, SDG7, SDG10, SDG11, SDG12, SDG15, and SGDs as a whole. Cities with 2–3 hot spots were Baotou (SDG4 and SDG9), Chifeng (SDG5 and SDG11), Xing’an League (SDG6, SDG10, and SDG15), and Hulunbuir (SDG10 and SDG15) located in the central and eastern regions. Among all the SDGs, SDG5, SDG6, SDG10, SDG11, and SDG15 all had cold spots appearing in the western region and hot spots appearing in the eastern region. SDG8 and SDG9 had mostly hot spots in the western or central regions.

4.4. Interactions between SDGs and between the Indicators at the Principal Level in 2020

4.4.1. Interactions between Indicators

It can be inferred from Figure 6 that for most SDGs, synergies outweigh trade-offs. Except for SDG4, SDG10, and SDG15, all other SDGs were positively correlated with each other. Especially SDG1, SDG3, SDG5, SDG7, SDG8, SDG9, and SDG11 showed strong positive correlations with each other, displaying synergistic relationships. The synergistic effect between SDG3 and SDG7 was the strongest, with a correlation coefficient of 0.98 and passing the strong significant level of 0.001. Strong synergistic effects also occurred between SDG1, SDG7, SDG3, SDG9, SDG8, SDG5, and SDG11, with correlation coefficients ranging from 0.87 to 0.98. The synergistic effects between SDG2, SDG6, SDG17, and other SDGs were weak. The positive correlations between SDG12 and SDG13 were strongly significant. However, their relationships with other SDGs were insignificant. SDG4 and most other SDGs had strongly negative correlations, but with no significant synergistic or trade-off relationships because the correlation coefficients were small. In the development of achieving other SDGs, the development of SDG4, SDG10, and SDG15 could not achieve improvements and may even be hindered. The achievements of other SDGs would be mutually promoted.

Figure 7 shows the interlinkages among different SDGs based on the network analysis.

Table 3. SDGs ranked by eigenvector centrality measured from network analysis.

SDGs	Hub Centrality	Betweenness Centrality	Proximity Centrality	Eigenvector Centrality	K-Kernel
SDG5	11	18.25	1	0.328	8
SDG2	9	0.536	0.846154	0.321	8
SDG3	9	0.536	0.846154	0.321	8
SDG6	9	0.536	0.846154	0.321	8
SDG7	9	0.536	0.846154	0.321	8
SDG8	9	0.536	0.846154	0.321	8
SDG9	9	0.25	0.846154	0.321	8
SDG11	9	0.25	0.846154	0.321	8
SDG1	8	0.286	0.785714	0.291	8
SDG17	8	0.25	0.785714	0.291	8
SDG10	1	0	0.52381	0.037	1
SDG15	1	0	0.52381	0.037	1
SDG12	1	0	1	0	1
SDG13	1	0	1	0	1
SDG4	0	0	0	0	0

shows SDGs ranked by eigenvector centrality measured from the network analysis. It can be inferred from Figure 6 and Table 3 that SDG5 had the highest hub centrality in the entire network, reaching 11, indicating a strong connection between SDG5 and other SDGs. Its betweenness centrality and proximity centrality were also the highest among all SDGs, reaching 18.25 and 1, respectively. In addition, the centrality and K-kernel of the feature vectors of SDG5 were also the highest, reaching 0.328 and 8, respectively. Thus, SDG5 was at the core of the entire network and was the most important indicator. It was in a group with high cohesion. The SDGs associated with SDG5 were also important throughout the entire network. SDG2, SDG3, SDG6, SDG7, SDG8, SDG9, and SDG11 occupied important positions in the entire system, with high centralities and K-kernel values. SDG12 was only directly related to SDG13, so its hub centrality, proximity centrality and K-kernel values were all 1, and the betweenness centrality and eigenvector centrality were 0. SDG4 had no obvious connection with other SDGs and was independent of the overall network. Thus, it could not achieve synergy or trade-off with others.

4.4.2. Interactions among Different Indicators of SDGs

As shown in Figure 8, most of the indicators showed significant positive correlations. About 33.58% of the indicators had a synergistic effect, 25.99% had trade-off effects, and 40.43% had insignificant relationships between them. The proportion of synergies between indicators was significantly higher than that of trade-offs. Among them, SOC2, ECO1, ECO4, and ECO5 had the strongest synergistic effects with correlation coefficients close to 1. There were also strong synergistic effects between SOC2, SOC19, and SOC23, with correlation coefficients reaching 0.998. The trade-off between ECO15 and ENV20 was the strongest, with a correlation coefficient close to −1. There were also strong trade-offs between SOC14 and other indicators such as SOC2, SOC19, SOC23, ECO1, ECO4, and ECO5, with correlation coefficients ranging from −0.986 to −0.988.

Figure 9 and

Table 4. Top 30 indicators ranked by eigenvector centrality measured from network analysis.

Indicator	Hub Centrality	Betweenness Centrality	Proximity Centrality	Eigenvector Centrality	K-Kernel
ECO2	52	27.513	0.948	0.147	36
SOC11	51	12.563	0.932	0.147	36
ENV13	51	12.563	0.932	0.147	36
SOC4	50	5.194	0.917	0.146	36
SOC9	50	5.194	0.917	0.146	36
SOC2	49	1.788	0.902	0.146	36
SOC7	49	1.788	0.902	0.146	36
SOC8	49	1.788	0.902	0.146	36
SOC10	49	1.788	0.902	0.146	36
SOC12	49	1.788	0.902	0.146	36
SOC13	49	1.788	0.902	0.146	36
SOC15	49	1.788	0.902	0.146	36
SOC17	49	1.788	0.902	0.146	36
SOC18	49	1.788	0.902	0.146	36
SOC19	49	1.788	0.902	0.146	36
SOC23	49	1.788	0.902	0.146	36
SOC24	49	1.788	0.902	0.146	36
ECO1	49	1.788	0.902	0.146	36
ECO4	49	1.788	0.902	0.146	36
ECO5	49	1.788	0.902	0.146	36
ECO8	49	1.788	0.902	0.146	36
ECO10	49	1.788	0.902	0.146	36
ECO11	49	1.788	0.902	0.146	36
ENV3	49	1.788	0.902	0.146	36
ENV4	49	1.788	0.902	0.146	36
ENV8	49	1.788	0.902	0.146	36
ENV10	49	1.788	0.902	0.146	36
ENV17	49	1.788	0.902	0.146	36
SOC14	49	1.788	0.902	0.146	6
ENV14	49	1.788	0.902	0.146	6

show the results of the social network analysis. The color of the icon in Figure 8 represents hub centrality, i.e., the number of indicators that it connected. The size of the icon represents the importance of the indicator. A larger icon means higher importance. The red line represents a trade-off between indicators, with a negative correlation coefficient. The green line represents a synergistic effect between indicators, with a positive correlation coefficient. It can be seen from Figure 8 that ECO2 had the highest degree of hub centrality, reaching 52, indicating strong correlations between ECO2 and other indicators. Its betweenness centrality and proximity centrality were also the highest among all the indicators, reaching 27.513 and 0.948, respectively. In addition, the centrality and K-kernel of the feature vectors of ECO2 were also the highest, reaching 0.147 and 36, respectively. Thus, ECO2 was at the core of the entire network and was the most important indicator. It was in a group with high cohesion of 36 indicators. The SDGs associated with ECO2 were also important throughout the entire network. SOC11, ENV13, SOC4, and SOC9 were important in the entire network too, with a hub centrality of 50 or 51, indicating their connection to 50 or 51 indicators. The other centralities and K-kernel of these indicators were also high. ENV2 was only related to ENV6. SOC3, SOC5, SOC21, ENV9, ENV19, and ENV23 were independent of the overall network, having no synergy or balance with other indicators.

5. Discussions

5.1. Increasing Trend of Sustainable Development in Inner Mongolia

The SDGs score is influenced by geographical conditions, economic conditions, urbanization level, and climate conditions [7,15,35]. This study found that from 2001 to 2020, Inner Mongolia has steadily improved its economic and social conditions with the support of national policies. Among them, the growth rate of clean energy in Inner Mongolia was the largest, which was closely related to the country’s strong investment in energy security in Inner Mongolia [36]. At the same time, Inner Mongolia has significantly improved its medical and health sector due to increased financial support from local governments for healthcare. Basic medical care for residents was guaranteed, and the increase in beds for medical workers and institutions also ensured the well-being of local residents. With the investment in transportation infrastructure construction and the vigorous introduction and cultivation of scientific and technological innovation talents, Inner Mongolia has made significant progress in industrial innovation and infrastructure construction. Most regions had the best sustainable development level in 2018. However, due to the impact of the COVID-19 pandemic that began in 2019, social and economic development was restricted, medical service supply was short, and food security risks increased [37], leading to a decline in sustainable development in most regions in 2019 and 2020.

Challenges in education also occurred in Inner Mongolia. For example, high quality teaching staff could not meet the rapidly growing number of students. In the future, emphasis should be placed on cultivating or introducing high-quality teaching staff to strengthen the overall education level in Inner Mongolia. At the same time, Inner Mongolia has made poor progress in promoting social equality, building sustainable cities and communities, addressing climate change and its impacts, and protecting, restoring, and promoting sustainable use of terrestrial ecosystems. The growth rate of household income was unstable, and the high emissions of air pollutants remain an urgent issue that needs to be address. These problems should be emphasized and improved in the future to improve the sustainable development of the region.

5.2. Obvious Regional Differentiation of Sustainable Development in Inner Mongolia

Inner Mongolia faced problems of imbalanced and uncoordinated regional development, severe spatiotemporal distribution imbalance of water resources, and unsuitable distribution of population and arable land [26]. Due to the deep inland areas, insufficient water resources, severe desertification, and scarcity of natural resources, the western region of Inner Mongolia has experienced poor economic development and insufficient social services [38]. Therefore, the SDGs in the western region had poor scores. The eastern region has abundant water resources and natural resource reserves, and the development of agriculture and animal husbandry is better than that in the central and western regions. Therefore, the overall SDGs score was better. However, the eastern region still faced the problem of backward socio-economic development.

In 2020, there was still a situation where the scores of various SDGs in Inner Mongolia were generally higher in the eastern region than in the western region, especially in gender equality, clean water and environmental sanitation, social equality, sustainable cities and sustainable use of terrestrial ecosystems, which showed a clustering of high and low values in the east and west, respectively. In terms of gender equality, clean drinking water, social equality, urban and human community sustainability, as well as the protection and restoration of terrestrial ecosystems, the completion of the goals in the cities of the eastern region was relatively good. Most environmental indicators were related to water resources and vegetation. Previous studies have shown that water resources may be the main obstacle to sustainable development in arid areas [39]. Under arid and semi-arid climate conditions, the achievement of SDGs in Inner Mongolia was also limited by water resources. The western region of Inner Mongolia lacks water resources, and its vegetation degradation has been severe [40]. Vegetation degradation may further cause soil erosion and ultimately hinder local sustainable development. The eastern region is rich in water resources, and the vegetation improvement has been relatively good from 2001 to 2020 [40]. The restoration of vegetation may lead to an increase in precipitation by affecting water flux, further improving the local water resource situation [41], resulting in an increase in SDGs scores. Decent work and infrastructure and sustainable industrialization have achieved better results in the western and central regions, and their indicators were both related to economic development. The western and central regions of Inner Mongolia are rich in mineral resources. With the continuous progress of technology, mineral resources are gradually being exploited and utilized, and the local economy has achieved great development. The economic development in the eastern region was relatively poor. Therefore, different SDGs exhibit significant regional clustering due to their different development situations and resource endowments in each city.

5.3. Reasonable Improvement Required Clarifying Coordination and Trade-Offs among SDGs

Due to the close relationships between SDGs and between the indicators, which may reflect collaborative progress or conflicting trade-offs, it is impossible to achieve them simultaneously. Therefore, it is necessary to clarify the complex relationship between these goals and indicators in order to achieve all sustainable development goals by 2030 as soon as possible [42]. This study found that there were more synergistic effects between goals and indicators than trade-offs, which was consistent with previous research results [15,29,43]. However, unlike previous studies on different provinces where the proportion of trade-offs between indicators was very small, Inner Mongolia had a larger proportion of trade-offs between various indicators. Therefore, in the future, Inner Mongolia should pay more attention to considering the importance of different indicators, finding reasonable solutions, and making effective decisions.

Among the relationships of SGDS, the connections between gender equality and the other goals were the most extensive and mainly reflected in synergies. Therefore, Inner Mongolia can make progress together with the other 11 goals by focusing on promoting gender equality. In addition, good medical conditions and sustainable energy use had the strongest synergistic effect with the other SDGs among all target pairs. Therefore, improving medical conditions and sustainable energy use are also important to regional sustainable development because they can provide positive effects on other SDGs. More attention should be paid to the development of quality education. Although its correlation coefficients with other goals were relatively small and had not yet reached a balanced level, its negative correlations with most SDGs were strongly significant. Thus, it may have a negative impact on the achievement of other SDGs, and then affect the overall level of regional sustainable development.

During the sustainable development of Inner Mongolia, the following indicators were also important: energy intensity, number of health workers per 10,000 people, total sulfur dioxide emissions per unit GDP, grain yield per unit, and number of deaths from traffic injuries per 100,000 people, because they had strong synergistic effects on other indicators. Some indicators can also jointly promote development. These indicators included insurance participation ratio, per capita disposable income of rural residents, per capita GDP, and per capita salary of employed personnel. More attention should be paid to the trade-off relationships between the indicators, because one indicator’s improvement may deteriorate another’s development. The trade-offs occurred between land use rate and the proportion of desertified land to total land area, and between the ratio of the number of teachers to the number of students in higher education institutions and other indicators such as insurance participation ratio, work injury insurance coverage rate, per capita construction land area, per capita disposable income of rural residents, per capita GDP, and per capita salary of employed personnel. The results of the coordination and balancing effects between SDGs and between indicators in regions with different levels of development also had significant differences. For example, developed regions always had a smaller proportion of balancing effects compared to other regions [29]. Therefore, when exploring their relationship, further consideration should also be given to the local development situation.

5.4. Position and Future Prospects of This Study

This study clarified the interactions among SDGs and among their detailed indicators at both provincial and municipal levels in the ecologically sensitive region of Inner Mongolia. It provided a novel idea to think about SDGs in assessing regional sustainability development, because there are complicated coordination or trade-off inter-relationships among different goals and among different indicators. Our results indicated that considering single or a few SDGs may be not feasible when assessing sustainable development because other related goals or indicators may play roles. In this study, we found that for most SDGs and indicators, synergies outweigh trade-offs. Improvement measures should focus on the indices with strong synergies such as good medical conditions and sustainable energy use and insurance participation ratio, per capita disposable income of rural residents, per capita GDP, and per capita salary of employed personnel. Special attention should be put on those with trade-offs such as land use rate and the proportion of desertified land to total land area, to be aware of their opposite effects. Gender equality and energy intensity were the most important in the entire network that needed concentration.

Although the indicators and goals were selected based on the characteristics of Inner Mongolia, the index system was general for all the cities. Achieving a sustainable development of a social–economic–environmental system is the final goal for all the countries and cities in the world. Thus, clarifying the status and interactions among SDGs and their indicators is the foundation for further exploration of the detailed implementation of each goal. The results of this study provided a preliminary conclusion about the inter-relationships among SDGs and the indicators. In the future, more studies should be carried out to verify their interactions in different regions, to derive general rules for improving sustainable development worldwide.

6. Conclusions

(1)

At the provincial level, compared to other goals, the development of good medical conditions (SDG3) and sustainable energy use (SDG7) increased obviously, and the development of challenges in education (SDG4) decreased significantly from 2001 to 2020.

(2)

Different goals distributed divergently in different regions. Gender equality (SDG5), clean water and environmental sanitation (SDG6), social equality (SDG10), sustainable cities (SDG11), and sustainable use of terrestrial ecosystems (SDG15) got cold spots in the western region and hot spots in the eastern region. Decent work (SDG8) and infrastructure and sustainable industrialization (SDG9) got hot spots in the western regions.

(3)

For most SDGs and their indicators, synergies outweigh trade-offs. Of all the relationships among SDGs, the synergistic effect between good medical conditions (SDG3) and sustainable energy use (SDG7) was the strongest, and gender equality (SDG5) was the most important in the entire network. A total of 33.58% of the indicators had synergies, and 25.99% of the indicators had trade-offs. Of all the indicators, insurance participation ratio (SOC2), per capita disposable income of rural residents (ECO1), per capita GDP (ECO4), and per capita salary of employed personnel (ECO5) had the strongest synergies with each other, and land use rate (ECO15) and the proportion of desertified land to total land area (ENV20) had the strongest trade-off. Energy intensity (ECO2) was the most important indicator in the entire network.