EN中文

Regional Sustainability >

2026 , Vol. 7 >Issue 1: 100295

DOI: https://doi.org/10.1016/j.regsus.2026.100295

Full Length Article

Coupling dynamics of SDGs in Tajikistan from 2001 to 2023

  • Ranna HAZIHAN a, b ,
  • DU Hongru , a, b, c, * ,
  • HE Chuanchuan d ,
  • Kobiljon Khushvakht KHUSHVAKHTZODA e ,
  • Bobozoda KOMIL e, f
Expand
  • aXinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Urumqi, 830011, China
  • bUniversity of Chinese Academy of Sciences, Beijing, 100049, China
  • cState Key Laboratory of Ecological Safety and Sustainable Development in Arid Lands, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Urumqi, 830011, China
  • dSchool of Public Administration, Xinjiang Agricultural University, Urumqi, 830052, China
  • eNational Academy of Sciences of Tajikistan, Dushanbe, 734000, Tajikistan
  • fInstitute of Economics and Demography, National Academy of Sciences of Tajikistan, Dushanbe, 734000, Tajikistan
* E-mail address: (DU Hongru).

Received date: 2025-08-29

  Revised date: 2025-10-25

  Accepted date: 2026-01-06

  Online published: 2026-03-11

Fold

Abstract

Since the United Nations launched the Sustainable Development Goals (SDGs) in 2015, global implementation has steadily advanced, yet prominent challenges persist. Progress has been uneven across regions and countries, with Tajikistan representing a typical example of such disparities. Based on 81 SDG indicators for Tajikistan from 2001 to 2023, this study applied a three-level coupling network framework: at the microscale, it identified synergies and trade-offs between indicators; at the mesoscale, it examined the strength and direction of linkages within four SDG-related components (society, finance, governance, and environment); and at the global level, it focused on the overall SDG interlinkages. Spearman’s rank correlation, sliding window method, and topological properties were employed to analyze the coupling dynamics of SDGs. Results showed that over 70.00% of associations in the global SDG network were of medium-to-low intensity, alongside extremely strong ones (|r| value approached 1.00, where r is the correlation coefficient). SDG interactions were generally limited, with stable local synergy clusters in core livelihood sectors. Network modularity fluctuated, reflecting a cycle of differentiation, integration, and fragmentation, while coupling efficiency varied with the external environment. Each component exhibited distinct functional characteristics. The social component maintained high connectivity through the “poverty alleviation-education-healthcare” loop. The environmental component shifted toward coordinated eco-economic governance. The governance-related component broke interdepartmental barriers, while the financial component showed weak links between resource-based indicators and consumption/employment indicators. Tajikistan’s SDG coupling evolved through three phases: survival-oriented (2001-2012), policy integration (2013-2018), and shock adaptation (2019-2023). These phases were driven by policy changes, resource industries, governance optimization, and external factors. This study enriches the analytical framework for understanding the dynamic coupling of SDGs in mountainous resource-dependent countries and provides empirical evidence to support similar countries in formulating phase-specific SDG promotion strategies.

Key words： Sustainable; Development; Goals (SDGs); Climate change; Coupling network analysis; Spearman’s rank correlation; Synergic relationships; Trade-off relationships; Tajikistan

Cite this article

Ranna HAZIHAN , DU Hongru , HE Chuanchuan , Kobiljon Khushvakht KHUSHVAKHTZODA , Bobozoda KOMIL . Coupling dynamics of SDGs in Tajikistan from 2001 to 2023[J]. Regional Sustainability, 2026 , 7(1) : 100295 . DOI: 10.1016/j.regsus.2026.100295

1. Introduction

In 2015, the United Nations formally introduced 17 Sustainable Development Goals (SDGs) in the 2030 Agenda for Sustainable Development. By integrating the economy, society, and environment, the SDGs create a comprehensive development framework addressing key issues such as poverty, hunger, health, education, gender equality, and climate action. The initiative seeks to eradicate poverty, achieve global prosperity, and foster coordinated efforts to tackle global challenges by 2030 (United Nations, 2015). The SDGs emphasize the “integrated and indivisible” interrelated nature among goals, where each goal involves three pillars to varying degrees. This reflects the essential consensus that sustainable development is the coordinated evolution of the economic, social, and environmental pillars (United Nations, 2015; Mishra et al., 2024).
However, the global progress of SDGs is marked by significant imbalance and contradictions. The Sustainable Development Report 2024, published by the United Nations Sustainable Development Solutions Network (SDSN), indicates that by 2023, only 15.00% of SDG targets had achieved mid-term progress (e.g., poverty reduction in some Asian countries and universal primary education in Nordic countries), 48.00% of targets were stagnant or even regressing (e.g., climate change response and inequality reduction), while progress on the remaining targets remained sluggish (Sachs et al., 2024). This disparity is particularly pronounced across different development trajectories. Developed countries lead in areas such as social equity (e.g., SDG 10 on Reducing Inequality) and environmental governance (e.g., SDG 14 on Ocean Conservation), yet fall short in Global Cooperation (SDG 17). For instance, developed countries have only fulfilled 60.00% of their climate financing commitments to developing countries, which has constrained the coordination of global climate action to a certain extent (Pickering et al., 2015; Alayza and Caldwell, 2021). Low-income developing countries face pressures from both survival needs and limited development capacity. Core indicators for basic needs, such as SDG 1 (No Poverty) and SDG 2 (Zero Hunger), have continued to decline due to conflicts, economic slowdowns, pandemics, and unprecedented desert locust outbreak in East Africa. The global extreme poverty rate even rose in 2020 for the first time in decades, with projections indicating that 0.59×109 people will remain in extreme poverty by 2030 and at least 3.00×109 people cannot afford a healthy diet, highlighting the challenges between persistent poverty and SDGs (Bongaarts, 2021; Soto, 2024). Resource-dependent countries in Central Asia, Latin America, and other regions worldwide generally fall into a “development paradox” between economic growth and ecological protection: their economic growth (SDG 8 on Decent Work and Economic Growth) relies on resource-based industries such as mineral development and hydropower construction, which is accompanied by ecological problems like soil pollution and biodiversity loss (SDG 15 on Life on Land), forming a vicious cycle of “growth equals destruction” (Eisenmenger et al., 2020; Leal Filho et al., 2023).
Among resource-dependent countries grappling with SDG contradictions, the particularity of mountainous resource-dependent countries deserves attention. These countries typically exhibit both extreme ecological fragility and high economic dependence on resources, where climate change, disaster risks, and poverty intertwine. Additionally, the spatial heterogeneity of population distribution and resource endowments further complicates the coordinated development of SDGs (Mackey et al., 2024). As a typical alpine resource-dependent country, Tajikistan exhibits SDG advancement patterns that closely align with these challenges. The 2024 Sustainable Development Report assessed its progress, placing Tajikistan in the upper-middle range among the 167 participating countries regarding overall SDG scores. However, significant disparities exist in the advancement of individual goals. Substantial progress has been made in basic livelihood and service sectors, indicating breakthroughs in survival-related objectives. Conversely, notable shortcomings persist in ecological conservation and industrial upgrading goals (Sachs et al., 2024). Tajikistan’s SDG contradictions are particularly illustrative. Ecologically and economically, 93.00% of its territory is mountainous areas, with ecosystems highly sensitive to climate change. Glacial melting and mountain hazards directly threaten water security and biodiversity (Gaforzoda and Yuldashev, 2023). Hydropower development and mineral exports account for over 40.00% of gross domestic product (GDP), making resource-based industries both the core driver of economic growth and the primary cause of ecological degradation (Migranyan and Krishtal, 2024). Socially and developmentally, 60.00% of its population is concentrated in limited river valley plains; the scarcity of available land leads to regional development imbalances, and the mismatch between people’s livelihood needs (e.g., youth employment, skills training, and basic medical care) and institutional supply is prominent (Strokova and Ajwad, 2017; Ghimire et al., 2023; Abdullozoda et al., 2024). This pattern—where progress on certain livelihood objectives is significant while ecological and industrial goals lag, and resource dependency rigidly conflicts with ecological fragility—makes the country an ideal case study for analyzing the coupling dynamics of SDGs in mountainous resource-dependent nations.
Academic consensus has emerged that the SDGs are not a collection of isolated goals, but rather a complex system characterized by interrelated objectives and overlapping multidimensional attributes. Previous scholars have conducted extensive foundational research on this core characteristic, establishing a crucial framework for subsequent exploration (Blanc, 2015; Fu et al., 2019; Cernev and Fenner, 2020). In terms of analyzing the core characteristics of the SDG system, Blanc (2015) systematically examined synergies (positive correlations and mutual reinforcement among indicators) and trade-offs (negative correlations and mutual constraints among indicators), demonstrating that advancing a single goal inevitably triggers chain reactions across others. Nilsson et al. (2016) further clarified that the 17 SDGs form an organic whole through direct or indirect connections. For instance, the popularization of education (SDG 4, Quality Education) can promote gender equality (SDG 5, Gender Equality) by improving female literacy rates (Damanik, 2025), while energy-intensive economic growth (SDG 8, Decent Work and Economic Growth) may exacerbate carbon emissions (SDG 13, Climate Action) (Hickel, 2019). Based on data from 166 countries, Wu et al. (2023) refined this complexity into three dimensions: the dependence of the economy and society on resources and climate, the cross-target transmission of environmental issues, and the sacrifice of equality in the development process, enriching the concrete understanding of interrelated characteristics. Within the classification framework, the three-pillar framework of social, economic, and environmental dimensions provides a clear logical structure for systematic analysis, serving as a foundational paradigm for understanding the essence of sustainable development (Mensah, 2019). However, with in-depth research, some studies have pointed out its limitations. For example, it overlooks the restrictive nature of the environment on human activities, potentially leading to imbalances between economic and environmental objectives (Holden et al., 2017). To overcome this limitation, Fu et al. (2019) employed a matrix method to classify the SDGs into three categories—basic needs, expected goals, and governance guarantees—seeking to align with the functional attributes of the goals. This attempt has provided an important insight for optimizing classification frameworks.
In terms of method application, the academic community has developed a variety of tools to quantify the interaction relationships among SDGs, such as system mapping, network analysis, and weighted graph models (Zelinka and Amadei, 2019; Cernev and Fenner, 2020; Pham-Truffert et al., 2020). Among these, network construction based on correlation analysis has emerged as a mainstream approach due to its ability to accurately characterize connections. For example, based on data from 227 countries, Pradhan et al. (2017) identified core hub goals such as SDG 1 (No Poverty) and SDG 4 (Quality Education) through correlation analysis, confirming the method’s capability to capture key connections. Weitz et al. (2018), Sebestyén et al. (2019), and Wu et al. (2022) further improved the accuracy of identifying the degree of synergy/trade-offs and transmission efficiency by calculating network node eigenvector centrality and network topological properties (e.g., connectivity and transitivity), which quantify network structural characteristics. Cutting et al. (2024) verified the effectiveness of key connection identification and global event impact assessment through network methods. A review of 30 SDG interaction analysis methods by Horvath et al. (2022) also shows that the academic community has formed a diversified method system. However, there is still no universally optimal solution. Furthermore, numerous studies have focused on the importance of indicator-level analysis; for instance, by splitting indicators under SDG 6 (Clean Water and Sanitation), such as “water quality compliance rate” (environmental attribute) and “access to drinking water” (social attribute), more accurate connection analysis can be achieved (Roy and Pramanick, 2018; Balata et al., 2022).
Despite significant progress in existing research on SDG complexity analysis and methodological optimization, there remain several areas for expansion in the research on the SDG coupling mechanisms in mountainous resource-dependent countries in Central Asia. First, region-specific research is relatively scarce. Most existing network analyses focus on global cross-country comparisons or large-scale regional assessments, resulting in insufficient empirical analysis of SDG coupling laws under the multidimensional conflicts of ecological fragility, economic dependence, and livelihood constraints. This makes it difficult to capture the unique impacts of special mountainous geographical environments (e.g., glacial melting) and resource-based economic structures on goal interconnections (Mackey et al., 2024). Second, although some studies emphasize indicator-level analysis, certain literature still classifies units based solely on SDG target numbers (e.g., categorizing SDG 6 entirely under the environmental dimension), failing to sufficiently refine classifications based on indicator substance. This fails to fully align with the multidimensional nature of the goals (Roy and Pramanick, 2018; Balata et al., 2022). Third, quantitative analysis of institutional regulatory effects is lacking. While Dawes (2022) identifies policies as key tools for coordinating SDG conflicts, most existing studies treat institutions as “external intervention variables” rather than core components within coupling networks. Consequently, it is impossible to quantify the amplification effect of policies on synergistic effects and the control threshold of policies on trade-off effects. Fourth, the capture of long-term dynamic coupling laws is inadequate. Most existing studies adopt static cross-sectional data or short-period panel data, rarely tracking the complete development cycle of a single country, e.g., from a survival crisis to quality improvement. This hinders the ability to reveal the transformation characteristics of coupling dynamics at different stages.
Therefore, this study took Tajikistan as a case, constructing a three-level coupling network analysis based on 81 SDG indicators covering four components: social, financial, environmental, and governance-related dimensions. Combining Spearman’s rank correlation analysis, the sliding window method, and topological attribute quantification, this study revealed the phased characteristics of SDG coupling, the functional differentiation of components, and the dynamic mechanisms. The aim of this study is to enrich the analytical paths for the dynamic coupling of SDGs in mountainous resource-dependent countries, offering a reference perspective for similar countries’ SDG system research. Given that Tajikistan shares common characteristics in resource endowments and development stages with other mountainous resource-rich countries in Central Asia and parts of Latin America, the research conclusions can offer empirical references for these countries to formulate SDG promotion strategies adapted to their respective developmental phases.

2. Data and methods

2.1. Study area

The Republic of Tajikistan is located in the southeastern Central Asia, bordering Kyrgyzstan to the north, Uzbekistan to the west, Afghanistan to the south, and China to the east, with a land area of 143,100 km2. The territory is mountainous, known as “The Land of High Mountains”. Tajikistan features multi-ethnic cohabitation, and most of its population is concentrated in plains and river valleys. Available land resources are relatively scarce, but water and mineral resources have huge development potential. Since 2000, Tajikistan’s domestic economy has grown steadily and rapidly. Industrialization and urbanization have driven regional development and improved living standards. However, due to geographical limitations and economic development stage constraints, the country still faces challenges like unbalanced regional development, limited employment capacity, and over-reliance on resource-based industries.

2.2. Analytical framework

Focusing on the coupling dynamics of SDGs in Tajikistan, this study developed a “Microscale-Mesoscale-Global” three-level coupling network analysis framework, and applied this framework to analyze the system’s coupling laws and dynamics by examining the network characteristics of each level.
At the microscale level, SDGs served as network nodes. The synergistic or trade-off relationships between indicators were the sources of coupling dynamics, and the strength of connections between nodes determined the magnitude of such dynamics. At the mesoscale level, indicators were grouped into four sub-networks by social, financial, environmental, and governance-related dimensions. The interaction of indicators within each sub-network and cross-domain interaction between sub-networks acted as the intermediate link for dynamic transmission, which helped identify the specific role of each domain in coupling. At the global level, all indicator and component connections were integrated to form a complete coupling network. Its temporal scale changes and network connection strength reflected the overall characteristics of sustainable development and the state of coupling dynamics. By analyzing the structural changes, fluctuations in connection strength, inter-level interactions, and differences in node importance of the three-level networks, this study clarified the evolutionary laws and dynamic sources of SDG system coupling in Tajikistan.

2.3. Indicator selection and network construction

2.3.1. Indicator selection and preprocessing

The Tajikistan’s SDG indicator system in this paper included 81 indicators, with data from 2001 to 2023. For selection, 44 indicators were prioritized from UN SDGs to ensure standardization and comparability. Due to the lack of continuous time-series data for some SDGs during the study period, 37 supplementary indicators were added to reflect the country’s socio-economic structure, geographical characteristics, and resource-environmental features. All indicators were evenly assigned to four components: social (21 indicators), financial (28 indicators), environmental (14 indicators), and governance-related (18 indicators) components.
It should be noted that this study does not include indicators related to SDG 14 (Life Below Water), as Tajikistan, a landlocked country, has no available data on marine areas. For SDG 10 (Reduced Inequalities), several indicators were not directly included primarily due to a lack of continuous time-series data from 2001 to 2023. Importantly, core aspects of social equity—such as gender equality and urban-rural disparity—are addressed through related indicators under other SDGs. Although these indicators are not classified under SDG 10, they exemplify the integrated yet distributed nature of the SDG indicator framework.
The data mainly came from publicly available sources, including the World Bank (https://data360.worldbank.org/en/search) and the Food and Agriculture Organization of the United Nations (https://www.fao.org/faostat/en/#data). The indicators of each component are presented in Table 1. During preprocessing, the interpolation methods were used to fill missing data, and all indicators were normalized.
Table 1 Description of the selected indicators and identifiers.
Component type Indicators and identifiers
Social component The proportion of the population below the poverty line of 6.85 USD/d at 2017 purchasing power parity (SDG 1_S01), dietary energy supply (SDG 2_S02), prevalence of malnutrition (SDG 2_S03), the percentage of stunted children under 5 years of age (SDG 2 _S04), prevalence rate of anemia among pregnant women (SDG 2_S05), infant mortality rate (SDG 2_S06), adolescent fertility rate (SDG 2_S07), maternal mortality ratio (SDG 2_S08), tuberculosis incidence rate (SDG 3_S09), life expectancy at birth (SDG 3_S10), preschool enrollment rate (SDG 4_S11), primary school enrollment rate (SDG 4_S12), secondary school enrollment rate (SDG 4_S13), tertiary education enrollment rate (SDG 4_S14), male-female enrollment ratio in primary and secondary schools (SDG 4_S15), overall youth literacy rate (SDG 4_S16), higher education, teaching staff (SDG 5_S17), access to clean fuels and cooking technologies (SDG 7_S18), urban population growth rate (SDG 11_S19), the proportion of slum dwellers within the urban population (SDG 11_S20), and population using the internet (SDG 17_S21).
Financial component Household final consumption expenditure (SDG 1_F01), current health expenditure (SDG 3_F02), female labor force participation rate (SDG 5_F03), female employment in industry (SDG 5_F04), primary energy intensity (SDG 7_F05), renewable energy consumption (SDG 7_F06), GDP per employed person (SDG 8_F07), GDP growth rate (SDG 8_F08), GDP per capital growth (SDG 8_F09), total unemployed persons (SDG 8_F10), total unemployed women (SDG 8_F11), international tourism receipts (SDG 8_F12), merchandise trade (SDG 8_F13), employees in the service sector (SDG 8_F14), industrial value added (SDG 9_F15), manufacturing value added (SDG 9_F16), research and development expenditure (SDG 9_F17), railway passenger traffic (SDG 9_F18), air transport passenger volume (SDG 9_F19), ratio of actual government expenditure to original approved budget (SDG 16_F20), net barter terms of trade index (SDG 17_F21), merchandise exports to high-income economies (SDG 17_F22), exports of goods and services (SDG 17_F23), merchandise imports from developing economies in East Asia and the Pacific (SDG 17_F24), net official development assistance received (SDG 17_F25), net bilateral aid flows from Development Assistance Committee donors, Republic of Korea (SDG 17_F26), net inflows of foreign direct investment (SDG 17_F27), and net foreign assets (SDG 17_F28).
Governance-related component Grain import dependency ratio (SDG 2_G01), Political Stability and Absence of Violence/Terrorism (SDG 2_G02), physician density (SDG 3_G03), hospital beds (SDG 3_G04), out-of-pocket expenditure as a percentage of current health expenditure (SDG 3_G05), professional and technical institutions (SDG 4_G06), tertiary education institutions (SDG 4_G07), proportion of women in national parliaments (SDG 5_G08), Women, Business and Law Index score (SDG 5_G09), population using at least basic drinking water services (SDG 6_G10), population using safely managed drinking water services (SDG 6_G11), population using at least basic sanitation services (SDG 6_G12), population practicing open defecation (SDG 6_G13), electrification rate (SDG 7_G14), scientific and technical journal articles (SDG 9_G15), railway line density (SDG 9_G16), standard deviation of the Control of Corruption index (SDG 16_G17), and tax revenue (SDG 16_G18).
Environmental component Grain production (SDG 2_E01), per capita food supply variability (SDG 2_E02), proportion of freshwater withdrawal to available water resources (SDG 6_E03), total renewable internal freshwater resources (SDG 7_E04), per capita forest area (SDG 11_E05), adjusted net savings, excluding particulate emission damages (SDG 12_E06), total natural resource rents (SDG 12_E07), adjusted savings: mineral resource depletion (SDG 12_E08), total greenhouse gas emissions (SDG 13_E09), annual average exposure to PM2.5 air pollution (SDG 13_E10), total emissions from crop residue burning (SDG 13_E11), renewable internal freshwater resources per capita (SDG 15_E12), permanent cropland (SDG 15_E13), and permanent meadows and pastures (SDG 15_E14).

Note: SDG, Sustainable Development Goal; GDP, gross domestic product. Abbreviations are defined in the table and are consistent in subsequent figures and tables.

2.3.2. Network construction method

In this study, we constructed the network using Spearman’s rank correlation analysis, a commonly used method in SDG network analyses (Pradhan et al., 2017; Wu et al., 2022). In the Spearman’s rank correlation analysis, nodes represented the SDG indicators, and the edges represented indicator-pair links; edge weights corresponded to the absolute correlation coefficient (|r|) value, reflecting the strength of the association, with higher absolute values indicating stronger relationships between two indicators. The positive and negative signs of r indicated synergistic promotion relationships and trade-off constraint relationships, respectively (Zhang et al., 2022). Thus, two network types were formed: a synergy network for positive coefficients and a trade-off network for negative coefficients.
For network construction, we adopted screening criteria that require |r|≥0.10 and a statistical significance of P<0.05 after false discovery rate (FDR) correction. The selection of this threshold was based on two pieces of evidence. First, |r|≥0.10 covered 93.29% of actual associations, with strong correlations accounting for over 60.00% (Fig. 1). Second, further analysis of weak associations with |r| (between 0.10 and 0.30) showed that their weak correlation stems from the long-term indirect transmission characteristics of Tajikistan’s SDGs and constraints of its development stage, rather than meaningless random associations.
Fig. 1. Distribution of correlation strength in Tajikistan’s Sustainable Development Goal (SDG) network. Statistical significance was defined as a P<0.05 after false discovery rate (FDR) correction. r, correlation coefficient.
For example, the r between SDG 2_S05 (prevalence rate of anemia among pregnant women) and SDG 12_E07 (total natural resource rents) was 0.10. As a resource-dependent economy, Tajikistan relies on natural resource rents for fiscal revenue; these revenues can only indirectly improve maternal anemia by supporting public health investment, and the long transmission chain leads to low association strength. The r between SDG 5_G09 (Women, Business and Law Index score) and SDG 8_F08 (GDP growth rate) was -0.30, showing a weak negative correlation. Enhancing women’s business rights could boost economic vitality by expanding entrepreneurship and increasing jobs. However, due to Tajikistan’s development stage, where women’s business participation remains low and their entrepreneurship is concentrated in small-scale, low-GDP-weight industries, the positive effect of enhanced women’s business rights on economic growth has not yet manifested. The current weak negative correlation is thus transitional, with the positive effect expected to emerge as participation rises and the economic structure diversifies. The positive driving effect of improved women’s rights on economic vitality is a globally verified mechanism (Naveed et al., 2023), with a stronger effect in economically backward regions (Zhao, 2022).
Although these associations are weak, they suggest potential interlinkage patterns within the SDG system; an overly stringent threshold would lead to the omission of such valuable information. This aligns with the core proposition of weak tie theory, particularly recent studies that have further validated the theory’s applicability to complex networks. These studies emphasize that weak associations are not arbitrary; instead, they facilitate the transmission of implicit information across distinct components of a complex system, thereby uncovering latent systemic connections (Mrowinski et al., 2024). Meanwhile, FDR correction (P<0.05) effectively controlled false positives, ensuring that associations with |r|≥0.10 have statistical reliability.
All network analyses and visualizations were implemented using the ‘NetworkX 3.3 package’ in Python 3.13.5 (Hagberg et al., 2008).

2.3.3. Sliding window method

In this study, the sliding window method was used to characterize the medium- and long-term dynamic evolution of the SDG coupling network in Tajikistan during 2001-2023, with a focus on verifying the rationality of a fixed 5-a window through tests. The coefficient of variation (CV) of the indicator pair was calculated using the following formula:
$\text{CV}=\frac{\text{SD}}{\left|{\mu }_{ar}\right|}$
where SD is the standard deviation of the |r| of the indicator pair; and μₐᵣ is the mean of |r| of the indicator pair.
First, among indicator pairs with moderate to strong correlations, 23.56% have a CV value less than 0.30. This proportion is substantially higher than the 0.05% observed for weakly correlated pairs. This contrast suggests that more strongly correlated SDG indicators may have a greater tendency toward stability, potentially providing a more reliable foundation for core couplings within the SDG system.
Second, 82.30% of unstable strong correlations involve highly volatile indicators concentrated in the financial (39.28%) and governance-related (25.18%) subsystems. This is directly related to regional development characteristics, ruling out fluctuations caused by the sample size of the 5-a window.
Third, when comparing 3-a, 5-a, 7-a, and 10-a windows, the trend consistency rate of strong correlations reaches 61.42%, and the trend overlap rate between the 5-a window and the 7-a (or 10-a) window exceeds 76.00%, confirming that the dynamic trends it captures are reliable.
In conclusion, the 5-a window not only ensures the stability of core correlations and the reliability of trends but also is compatible with the characteristics of single-country time-series data, conforming to the methodological consensus in single-country sustainable development research (Dai et al., 2022).

2.4. Analysis methods

2.4.1. Analysis of network coupling strength

We selected four topological properties for analysis to quantify the coupling strength of the global network and component networks. Detailed calculation formulas for each property are shown in Table 2. Connectivity reflects the network’s overall connection density. The denser the actual associations between indicators, the higher the connectivity, indicating more frequent interactions in the SDG network. Transitivity measures if local nodes form connected, stable clusters. For example, if indicator A is associated with indicator B and indicator B is associated with indicator C, indicator A can easily connect with indicator C. A higher transitivity value means more such clusters (with no gaps in internal connections) and stronger local coupling. Modularity indicates whether the network has self-contained groups. These groups have close internal connections but little interaction with external nodes. A higher value makes independent subgroups more prominent, with coupling effects mainly transmitted internally. Coupling efficiency measures the smoothness of the coupling effect transmission. The closer the ratio to 1, the fewer detours in transmission, resulting in faster, more efficient overall linkage of the SDG network.
Table 2 Connotations and calculation methods of topological properties.
Property Connotation Calculation formula Reference
Connectivity (C) Network coupling strength (reflecting overall connection tightness) $C=\frac{{\displaystyle \sum {w}_{\text{observed}}}}{{\displaystyle \sum {w}_{\text{max}}}}$,
where ∑wobserved is the sum of the weights of all edges observed in practice; and ∑wmax is the theoretical maximum total link weight for the same number of nodes (assuming all possible edge weights are 1).		 Wu et al. (2022); Hu et al. (2023)
Transitivity (T) Local stability of the network $T=\frac{3 \times \text { number of triangles }}{\text { Number of connected triples }}$,
where number of triangles is the number of actual triangles formed (where all three nodes are pairwise connected); and number of connected triples is the number of possible triangles (triplets of nodes that could potentially form a triangle).		 Wu et al. (2022); Hu et al. (2023)
Modularity (Q) Network inhomogeneity
$Q=\frac{1}{2m}{\displaystyle \sum _{\begin{array}{c}i=1\\ j=1\end{array}}^{n}\left({A}_{ij}-\frac{{k}_{i}{k}_{j}}{2m}\right)}\delta ({c}_{i},{c}_{j})$
,
where m is the total edge weight of the network; n is the total number of nodes; Aij is the edge weight between nodes i and j (0 if not connected); ki and kj are the total connection weight of nodes i and j belong, respectively; ci and cj are the subset or community to which nodes i and j belongs, respectively; and δ(ci, cj) is the Kronecker delta function. If ci=cj, the value equals to 1; otherwise, it is 0.		 Hu et al. (2023)
Coupling efficiency (Coupling E) Transmission speed of coupling effects
$\text{Coupling }E=\frac{{t}_{\text{ideal}}}{{t}_{\text{observed}}}=\frac{n(n-1)}{{\displaystyle \sum _{i=1}^{n}\mathrm{min}}\left({\displaystyle \sum _{i\ne j\in G}\frac{1}{{w}_{ij}}}\right)}$
,
where tideal is the ideal traversal time; tobserved is the actual traversal time; n is the number of nodes; G represents the network under study; wij is the edge weight between nodes i and j; and 1/wij represents the transmission time between nodes i and j.		 Yuan et al. (2021)

2.4.2. Node centrality analysis

Eigenvector centrality was used to measure a node’s influence within the network. Its calculation considers not only the number of connections the node itself has but also the centrality of its adjacent nodes—this helps capture the indirect influence a node gains by associating with other important nodes (Hu et al., 2023). The formula is:
${x}_{i}=\frac{1}{\lambda }{\displaystyle \sum _{j=1}^{n}{M}_{i,j}}{x}_{j}$
where xi and xj denote the eigenvector centrality values of nodes i and j, respectively; n is the total number of nodes; Mi, j denotes the adjacency matrix element (if node i is connected to node j, it is 1; otherwise, it is 0); and λ denotes the largest eigenvalue of matrix M.
This study identified the top three nodes as network hubs each year. These hubs influenced other important nodes via dense connections and served as the core for propagating synergistic effects. Ranking nodes by centrality clarifies their relative importance and helps identify key drivers of changes in network coupling strength.

3. Results

3.1. Coupling relationships of Sustainable Development Goals (SDGs) at the national level

From 2001 to 2023, in Tajikistan’s SDG coupling network, the number of synergistic relationships (1112 pairs) and trade-off relationships (1044 pairs) between indicators were comparable, reflecting a balanced state between mutual promotion and constraint. The average |r| values were 0.75 and 0.76 for synergistic relationships and trade-off relationships, respectively, at relatively high levels, indicating that both types of relationships play significant roles in driving the SDG system (Fig. 2).
Fig. 2. Heatmap of the synergistic and trade-off relationships among 81 SDG indicators of Tajikistan from 2001 to 2023. The meaning of the heatmap colors is as follows: the bluer the color, the greater the positive value of the Spearman’s rank r of the indicator pair, indicating a stronger synergistic and promoting relationship between the indicators, with the synergistic effect as the main feature; the yellower the color, the greater the negative value of the Spearman’s rank r of the indicator pair, indicating a stronger trade-off and restrictive relationship between the indicators, with more prominent trade-off characteristics.
Specifically, among the synergistic relationships, there were 330 pairs of strong synergy with an average r value of 0.96 (|r|≥0.87), and 782 pairs of weak synergy with an average r value of 0.67 (0.10≤|r|<0.87). Among the trade-off relationships, strong trade-offs included 298 pairs with an average r value of -0.97 (|r|≥0.91), and weak trade-offs included 746 pairs with an average r value of -0.68 (0.10≤|r|<0.91). The proportion of medium- to low-intensity relationships exceeded 70.00%, but the |r| value of high-intensity relationships approached to 1.00, showing extremely strong interactions among certain SDG goals.
Strong synergistic relationships were mainly concentrated in three core clusters: livelihood development (SDG 3: Good Health and Well-being; SDG 4: Quality Education; and SDG 5: Gender Equality), survival guarantee (SDG 1: No Poverty; SDG 2: Zero Hunger; and SDG 6: Clean Water and Sanitation), and development drivers (SDG 4: Quality Education; SDG 8: Decent Work and Economic Growth; and SDG 9: Industry, Innovation and Infrastructure).
Strong trade-off relationships focused on three key conflicts: development versus climate constraints (SDG 8: Decent Work and Economic Growth; SDG 13: Climate Action; and SDG 7: Affordable and Clean Energy), survival versus ecological limits (SDG 2: Zero Hunger and SDG 15: Life on Land), and consumption versus sustainable transformation (SDG 12: Responsible Consumption and Production; SDG 8: Decent Work and Economic Growth; and SDG 13: Climate Action).
This long-term coexistence of strong associations reflected Tajikistan’s post-war development needs (i.e., addressing livelihood gaps and prioritizing survival) and development driver cultivation from 2001 to 2023, while also illustrating the specific impact of its national constraints (i.e., resource dependence and ecological vulnerability) on SDG target coupling dynamics.

3.2. Temporal evolution of the SDG coupling network

3.2.1. Dynamic trend of network scale

From 2001 to 2023, the total number of edges and nodes in Tajikistan’s SDG coupling network fluctuated in phases (Fig. 3), showing an overall trend of early stability, followed by contraction, gradual recovery to a peak, and eventual decline.
Fig. 3. Dynamic changes of the SDG coupled network structure in Tajikistan from 2001 to 2023. Edge refers to the number of significant associations filtered by Spearman’s rank correlation; synergy corresponds to the number of positive correlation edges; trade-off corresponds to the number of negative correlation edges; node refers to the number of indicators included in the network within the sliding window.
Among them, the number of nodes and synergistic edges peaked during 2014-2018, driven by the coordinated guidance of a series of Tajikistan’s policy documents. The core documents included the National Development Strategy of the Republic of Tajikistan for the Period to 2015 (Republic of Tajikistan, 2007), its preceding phased poverty reduction strategies (Phase II 2007-2009 and Phase III 2010-2012), and the Living Standards Improvement Strategy of Tajikistan for 2013-2015 (Republic of Tajikistan, 2013a). They integrated cross-domain resources in economy, society, and livelihoods, focused on core goals supporting sustainable development (e.g., energy independence, food security, and human development). By leveraging government-civil society collaboration to align resource allocation with SDG priorities, these policies effectively enhanced the positive correlations between complementary indicators, ultimately strengthening the synergies of SDGs by centering on these key fields.
After 2021, the network simplified. Weak connections (especially trade-off edges) were “squeezed out” due to external shocks such as the 2020 pandemics and the 2022 global energy crisis, and this phenomenon exposed the shortcomings of insufficient resilience in Tajikistan’s SDG coupling system. These external shocks directly undermined the stability of non-core connections within the system, which in turn reduced the overall coupling strength of the sustainable development network; ultimately, the network contracted and focused on core connections, as the contribution of non-core connections was significantly reduced.

3.2.2. Temporal characteristics of global network topological attributes

Tajikistan’s SDG global network has long maintained low connectivity, indicating that the SDGs are in a state of low interaction. The trade-off network’s value was lower than the synergy network’s value during most periods, confirming a synergy-dominated pattern (Fig. 4).
Fig. 4. Connectivity trend changes in the SDG global network in Tajikistan from 2001 to 2023. Synergy connectivity reflects the intensity and scope of positive associations between SDG indicators in the network, while trade-off connectivity reflects the intensity and scope of negative SDG associations.
The phased changes of this characteristic are related to policy advancement and external shocks. From 2014 to 2018, the connectivity of the synergy network reached its peak, corresponding to the implementation of the Living Standards Improvement Strategy of Tajikistan for 2013-2015 (Republic of Tajikistan, 2013a) and overlaying on the goal of “synergizing industrialization and energy independence” outlined in the National Development Strategy of the Republic of Tajikistan for the Period until 2030 (Ministry of Defence of the Republic of Tajikistan, 2016). This strategy prioritizes cross-sectoral integration (e.g., energy and employment) as a core task. Under its framework, the “CASA-1000 Energy Programme” links power, employment, and trade sectors, significantly boosting interaction intensity across SDG fields (Kojo and Sattar, 2018).
After 2020, the connectivity plummeted to 0.08. The main reason was the impact of the global pandemic in 2020. In 2020, Tajikistan’s remittance income decreased, and resources in livelihood-related fields shrank. Meantime, the government shifted its focus to pandemic prevention, which significantly reduced interactions in non-core fields.
Changes in the synergy network’s transitivity, modularity, and coupling efficiency reflect the phased characteristics of the coupling dynamics of Tajikistan’s SDG system. These characteristics are manifested in the dynamic responses of local coupling stability, inter-field synergy integration capacity, and synergy effect transmission efficiency.
Synergy network’s transitivity has remained around 0.90 for a long time (Fig. 5a). It reflects the tightness of the synergistic relationship within the local clusters of the SDG system and its resistance to interference. This stability is concentrated in core livelihood fields. From Tajikistan’s development reality, livelihood fields such as energy, food security, healthcare, and education have long been the government’s policy priorities. Even in the face of external shocks, the synergy links in these fields remain stable (>0.80). For example, in 2008, the government safeguarded social security and agricultural investment through the Plan of Measures to Counter the Financial Crisis (Republic of Tajikistan, 2008), preventing the breakdown of synergy between food production and livelihood subsidies. The Strategy for Countering Corruption in the Republic of Tajikistan for the period of 2013-2020 (Republic of Tajikistan, 2013b) standardized resource allocation and ensured stable investment in livelihood fields. These consistent policies helped form anti-interference local synergy clusters for “energy-food security” and “healthcare-education”, which became the core carriers of coupling stability.
Fig. 5. Changes in global network coupling strength characteristics. (a), transitivity trend changes of the synergy network; (b), modularity trend changes of the synergy network; (c), coupling efficiency trend changes of the synergy network.
Synergy network’s modularity reflects the tightness of links between different SDG fields. Its three-phase change of “differentiation-integration-fragmentation” corresponds to the dynamic adjustment of coupling integration capacity (Fig. 5b). From 2001 to 2012, modularity stayed at a high level of 0.50-0.60. This was because policies focused on breakthroughs in single fields (e.g., phased poverty reduction strategies, Phase II 2007-2009, and Phase III 2010-2012); they only targeted poverty governance without cross-field linkage, leading to obvious barriers between SDG fields and weak integration capacity. From 2014 to 2018, modularity dropped to 0.34. This was supported by the cross-field integration of the Strategy for Improving Living Standards (2013-2015) (Republic of Tajikistan, 2013a) and the linkage of the “CASA-1000 Energy Programme”, which broke down field segmentation and promoted cross-field integration of economic and social resources. From 2019 to 2023, modularity rose to 0.68. Due to the combined impact of the 2020 pandemic and the 2022 global energy crisis, the government concentrated resources on core fields such as electricity and food, leading to the breakdown of coupling links in non-core fields.
Synergy network’s coupling efficiency reflects the transmission speed of synergy effects. High fluctuations in the range of 0.36-3.08 reflect changes in coupling transmission efficiency, which is closely related to the external environment (Fig. 5c). It peaked from 2004 to 2008, coinciding with the early phase of the 2007-2009 poverty reduction strategy (Resolution of the Government of Tajikistan, 2007). During this period, the positive synergy effects of agricultural output growth and livelihood improvement spread rapidly through “government-community” linkages, resulting in high transmission efficiency. A pronounced low coupling efficiency occurred from 2008 to 2012, primarily driven by the global financial crisis. Declining resource allocation efficiency during this time hindered the transmission of synergy effects. It rebounded briefly after 2016, as the national development strategy promoted cross-field synergy. Coupling efficiency declined again after 2020 due to the pandemic, slowing down the transmission of synergy effects.
In conclusion, through the analysis of global network topological properties, we can outline the evolution trajectory of Tajikistan’s SDG coupling dynamics: in the early stage (2001-2012), coupling stability was maintained through livelihood policies; in the middle stage (2013-2018), coupling integration was strengthened via comprehensive policies and international cooperation; and in the later stage (2019-2023), affected by external shocks, resources shrank toward core fields, and the system shifted from “overall coupling” to “local coupling in core fields”.

3.3. Coupling strength within components

3.3.1. Social component

The connectivity (Fig. 6a and b) and transitivity (Fig. 6c) of the social component have long remained at the highest level, manifested as a “poverty alleviation-education & healthcare-infrastructure” triangular connection loop. The interconnections among these three livelihood indicators provide stable support for SDG coupling.
Fig. 6. Changes in the coupling strength characteristics of component networks. (a), connectivity trend changes of the synergy network; (b), connectivity trend changes of the trade-off network; (c), transitivity trend changes of the synergy network; (d), modularity trend changes of the synergy network; (e), coupling efficiency trend changes of the synergy network; (f), coupling efficiency trend changes of the trade-off network. Transitivity, by topological connotation, measures local nodes’ ability to form stable connected clusters, and modularity reflects the independence of network functional subgroups—both necessary for identifying positive, stable subgroups supporting SDG advancement in the synergy network. Whereas the trade-off network centers on negative conflict intensity—rendering properties like transitivity and modularity analytically valueless—these structural metrics were calculated only for the synergy network.
From the perspective of indicator linkage, the proportion of the population below the poverty line (SDG 1_S01) is inherently tied to food security indicators (dietary energy supply (SDG 2_S02) and the percentage of stunted children under 5 years of age (SDG2_S04)). The income level of poor groups directly determines their ability to access food, while an insufficient dietary energy supply leads to child stunting. This interaction is an inevitable result of meeting survival needs, and its frequency continues to increase as livelihood security progresses.
Universal education (preschool enrollment rate (SDG 4_S11) and primary school enrollment rate (SDG 4_S12)) and health improvement (tuberculosis incidence rate (SDG 3_S09) and hospital beds (SDG 3_G04)) indicators mutually support each other: universal education enhances residents’ health awareness and reduces school dropouts due to illness, while the expansion of medical resources reduces health-related constraints on education, forming a human capital synergy cycle.
The improvement of basic service indicators (access to clean fuels and cooking technologies (SDG 7_S18) and population using the internet (SDG 17_S21)) further strengthens the above interactions. For example, higher Internet penetration expands access to educational resources.
Even when the 2020 pandemic caused a decline in remittances, Tajikistan’s primary school enrollment rate remained stable at over 97.00%, and the linkage of livelihood indicators was not broken—fully confirming the high stability of transitivity.
With a long-term modularity of lower than 0.50 (Fig. 6d), there is no obvious community segmentation within the component. This is related to the “improving public financial management and optimizing the allocation of funds for people’s well-being” clearly stated in the National Development Strategy of the Republic of Tajikistan until 2030. The flat resource allocation structure avoids losses from the multi-level transmission of policies, allowing funds to be directly invested in rural preschool education facilities or the expansion of grassroots hospital beds without cross-sector coordination, further reducing barriers to indicator interaction.
The coupling efficiency of the social component is the lowest among all four components (Fig. 6e and f), constrained by two factors. First, livelihood security involves multi-dimensional needs such as poverty alleviation, education, and healthcare, and the coordination of cross-regional and cross-departmental resources leads to prolonged transmission of synergy effects. Second, more than half of Tajikistan’s population is concentrated in river valley plains, while mountainous areas face scarce resources and weak infrastructure, resulting in uneven distribution of social service resources and reduced overall coupling efficiency of the social component.

3.3.2. Environmental component

The connectivity of the environmental component has long been at a low level, and its transitivity fluctuated with changes in governance models (Fig. 6a-c). This reflected the process where early ecological governance pursued single objectives, while later, driven by policies, the coupling momentum increased.
In the early stage, ecological governance focused on single tasks. For example, indicators within the environmental component, such as SDG 6_E03 (proportion of freshwater withdrawal to available water resources), SDG 13_E09 (total greenhouse gas emissions), and SDG 15_E14 (permanent meadows and pastures), operated independently, with no cross-task linkage between them.
From 2015 to 2019, connectivity increased (Fig. 6a), which is related to the “Special Ecological Protection Initiative” proposed in the National Development Strategy of the Republic of Tajikistan until 2030. This policy promoted coordinated linkage of projects such as pollution control (total greenhouse gas emissions (SDG 13_E09)), glacier protection (proportion of freshwater withdrawal to available water resources (SDG 6_E03)), and biodiversity restoration (permanent meadows and pastures (SDG 15_E14)), shifting ecological indicators from “single-task response” to “multi-objective coordinated governance”. The interaction between indicators improved significantly, and connectivity and transitivity rose simultaneously.
However, during the same period, trade-off connectivity also increased slightly (Fig. 6b), reflecting the contradiction in fund and human resource allocation between the two types of projects under limited resources. When tasks such as pollution control and glacier protection were simultaneously advanced, funds and human resources could not be balanced, leading to enhanced synergy effects in the ecological field and intensifying trade-off relationships.
In the early stage, modularity reached 0.70, showing obvious community segmentation (Fig. 6d). This was due to the division of departmental functions in ecological management, resulting in a weak linkage between indicators of the two fields. In the later stage, the requirement of “mineral development must be accompanied by ecological restoration” was introduced—SDG 12_E07 (total natural resource rents) and SDG 11_E05 (per capita forest area) became more closely linked, as mining projects were required to invest a certain proportion of funds in afforestation or vegetation restoration. The frequency of cross-sector indicator interaction increased, and modularity dropped to 0.40.
The coupling efficiency of the environmental component ranked at a medium level among the four components (Fig. 6e and f), which was constrained by two factors: first, the inherent nature of ecological restoration—ecological effects require long-term accumulation, and the long natural cycle slows down effect transmission; second, Tajikistan’s ecological projects (such as glacier protection and grassland restoration) rely on government fiscal investment and must compete for resources with energy and agriculture sectors. Ecological projects tend to give way to other projects, further delaying the transmission of ecological effects.

3.3.3. Governance-related component

The topological properties of the governance-related component fluctuate significantly with policy cycles, reflecting the regulatory role of institutional development (including various policies and regulations) in SDG coupling momentum.
During periods of intensive policy implementation, connectivity and transitivity were both at high levels (Fig. 6a-c). From 2009 to 2015, Tajikistan intensively issued policies such as the Anti-Corruption Strategy (2013-2020) and the Concept of Social Protection of the Population of the Republic of Tajikistan. These policies strengthened the linkage between SDG indicators in the governance-related component—such as SDG 16_G17 (standard deviation of the Control of Corruption index) and SDG 3_G05 (out-of-pocket expenditure as a percentage of current health expenditure). Anti-corruption efforts optimized resource allocation, improving the implementation efficiency of policies in healthcare and education, increasing the frequency of indicator interaction, and enhancing the overall connectivity. During this period, the institutional component provided regulatory gains for coupling momentum through efficient collaboration, amplifying the synergy effects in the livelihood and economic fields.
During policy adjustment periods, connectivity declined (Fig. 6a). Taking 2016-2017 as an example: during the transition between the National Development Strategy until 2030 and earlier poverty alleviation policies, the integration of cross-sector environmental policies (Republic of Tajikistan, 2008) led to the redivision of departmental powers and responsibilities, causing short-term coordination difficulties. This weakened the linkage between indicators in the governance-related component.
The simultaneous fluctuation of synergy and trade-off connectivity reflects contradictions during policy adjustments, for instance, disputes over resource allocation caused by unclear departmental powers (weakened synergy), and short-term livelihood conflicts caused by differences between old and new standards (strengthened trade-offs). During this period, the regulatory role of the institutional component in coupling momentum was temporarily weakened, reflecting that policy consistency is a key prerequisite for institutional regulatory momentum.
During 2001-2012, modularity was higher than 0.50, showing obvious community segmentation (Fig. 6d). Before 2017, budget allocation was carried out independently by individual departments, which led to a weak linkage between economic regulation and social welfare indicators. In the later stage, request submitted all policies to include provisions on livelihood and ecology. This strengthened the linkage of modularity with SDG 16_F20 (ratio of actual government expenditure to original approved budget), SDG 3_G05 (out-of-pocket expenditure as a percentage of current health expenditure), and SDG 6_E03 (proportion of freshwater withdrawal to available water resources) dropped to 0.20.
Coupling efficiency fluctuates with implementation efficiency: it peaked at 18.00 during 2009-2013, reflecting the transmission of goal changes driven by the implementation of comprehensive policies; later, it dropped to 2.00, possibly due to reduced policy implementation efficiency caused by unclear division of powers and responsibilities (Fig. 6e).

3.3.4. Financial component

The connectivity of the financial component has long been low (synergy: 0.10-0.20; trade-off: 0.10) (Fig. 6a and b), reflecting a weak linkage between resource-based indicators (primary energy intensity (SDG 7_F05) and total natural resource rents (SDG 12_E07)) and consumption/employment indicators (household final consumption expenditure (SDG 1_F01) and total unemployed women (SDG 8_F11)). Hydropower and mineral development revenues have not been converted into residents’ consumption capacity or employment opportunities, resulting in insufficient inter-industry synergy and low economic resilience against external shocks.
In the early stage, high transitivity reflected partial stability. For example, hydropower development drove GDP growth, leading to increased foreign investment, and the linkage between indicators was close. However, from 2019 to 2023, affected by the 2020 pandemic and 2022 global energy crisis, transitivity dropped sharply to 0.60 (Fig. 6c). The pandemic caused a decline in remittances and a sharp rise in total unemployment, breaking the triangular linkage among energy, economic development, and employment—further confirming insufficient economic resilience in sustainable development.
Modularity ranks at a relatively high level among the four components in most cases (Fig. 6d), with clear boundaries between resource-based indicators (renewable energy consumption (SDG 7_F06) and total natural resource rents (SDG 12_E07)) and consumption/employment indicators (household final consumption expenditure (SDG 1_F01) and total unemployed women (SDG 8_F11)). After 2019, due to economic downturn, the government promoted cross-industry resource allocation, temporarily breaking industry barriers.
Coupling efficiency has long been lower than 3.00 (Fig. 6e and f), with contradictions hidden under normal circumstances. For example, the contradiction between increased primary energy intensity (SDG 7_F05) and total unemployed women (SDG 8_F11) can be mitigated by subsidizing social security with resource revenues, keeping trade-off coupling efficiency around 2.80 (Fig. 6f). However, when the scale of resource development expands, contradictions erupt intensively, and trade-off coupling efficiency rises sharply. For instance, during the large-scale aluminum mining in a certain period, male employment in resource-based industries accounted for a relatively high proportion, leading to a certain increase in total unemployed women (SDG 8_F11). The contradiction between resource development and gender equality in employment was transmitted rapidly, pushing trade-off efficiency to 10.00 (Fig. 6f), reflecting gender imbalance in the economic structure.

3.4. Temporal changes of network nodes

Tajikistan’s SDG coupling network exhibited distinct functional roles of hub nodes across different time windows, with each phase’s core indicators shaping the network’s synergy focus (Fig. 7).
Fig. 7. Collaborative hub of SDG coupling network during 2001-2005 (a), 2009-2013 (b), 2011-2015 (c), 2012-2016 (d), 2015-2019 (e), 2017-2021 (f), 2018-2022 (g), and 2019-2023 (h). Points of different sizes represent synergy hubs in each sliding window. Larger points indicate higher rankings, meaning that the hubs have closer relationships with other indicators and more prominent influence in the network.
During 2001-2010, hub nodes were mainly survival-related indicators, such as SDG 1_S01 (the proportion of the population below the poverty line), SDG 2_S04 (the percentage of stunted children under 5 years of age), SDG 2_S03 (prevalence of malnutrition), and SDG 2_S05 (prevalence rate of anemia among pregnant women). Their core role was to establish the “livelihood-development” foundational synergy of the early global network—by anchoring poverty reduction, food security, and basic healthcare. These indicators connected the most critical survival links, ensuring the sustainable development system’s minimum operational stability.
During 2011-2018, hub nodes shifted toward development-driven indicators, with SDG 4 series (e.g., SDG 4_S11: preschool enrollment rate), SDG 2_S02 (dietary energy supply), and SDG 1_F01 (household final consumption expenditure) becoming central. Their key function was to drive the network’s transition from survival guarantee to human capital accumulation and economic vitality—this not only upgraded basic needs (e.g., dietary energy supply (SDG 2_S02) moved food security from “sufficiency” to “nutritional adequacy”) but also echoed the strengthened collaboration between financial and social components in sliding window networks, enriching the coupling momentum of the SDG system.
During 2019-2023, basic service indicators emerged prominently, with SDG 7_G14 (electrification rate) and SDG 7_S18 (access to clean fuels and cooking technologies) ranking among the top hubs. Their critical role was to serve as core supports for sustainable development in the post-pandemic era—while maintaining uninterrupted basic livelihood services, these indicators also facilitated the rising synergy between environmental and social components in the component network, preventing external shocks from disrupting the overall coupling structure.
Notably, social component indicators consistently served as core nodes across all windows. Their role evolved from focusing on basic guarantee issues (poverty and child stunting) to sustain the system’s survival bottom line, and to shift to development indicators (education and energy accessibility) to enhance sustainable development quality—directly reflecting the phased transition of SDGs. Financial component indicators (e.g., household final consumption expenditure (SDG 1_F01)) maintained significant prominence after the mid-period; their role was to sustain the financial component’s influence as a core cluster, underscoring its long-term support for SDG coupling.

4. Discussion

4.1. Evolutionary characteristics of Tajikistan’s SDGs

The evolution of Tajikistan’s SDGs dynamically adjusts with development stages, and changes in network scale are linked to the functional performance of the four components. From an overall perspective, the SDG network has gone through three phases of change—low-interaction segmentation, high-synergy integration, and core-field contraction—with the function of each component flexibly adapting to practical needs in every phase.
In the early phase (2001-2012), during the critical post-war reconstruction period, the network was characterized by low interaction and high segmentation. At that time, the core development need was to “stabilize basic survival”. Within the social component, indicators for poverty governance and food security formed basic linkages, safeguarding the bottom line of people’s well-being by ensuring low-income groups’ access to food and basic medical services. The governance-related component showed obvious segmentation due to the “department-specific budget system”, with economic regulation and social welfare sectors advancing independently. Although cross-sector synergy was not achieved, the concentrated investment in single-sector policies laid the necessary foundation for subsequent network integration.
In the middle phase (2013-2018), entering the quality improvement stage, the network shifted toward high synergy and strong integration. This change stemmed from the practical need to “promote synergy among the economy, people’s well-being, and ecology”, with multiple components forming a joint driving force. Relying on hydropower and mineral development, the financial component channeled resource revenues into residents’ livelihoods and education, fostering positive links between people’s well-being and economic indicators. The governance-related component broke down inter-departmental barriers through institutional reforms, enabling smoother cross-sector resource allocation and reducing administrative obstacles to indicator interaction. Under the policy of “mineral development with supporting ecological restoration”, the environmental component also shifted from single ecological governance to synergy with economic development—ensuring that resource development no longer came at the expense of the ecology and forming initial “economy-ecology” linkages.
In the late phase (2019-2023), affected by external shocks, the network entered a state of contraction and focus. At this point, development needs returned to “safeguarding basic people’s well-being” and resources were concentrated in key areas. The social and environmental components jointly focused on basic services, with indicators such as electricity (electrification rate (SDG 7_G14)) and clean fuels (access to clean fuels and cooking technologies (SDG 7_S18)) becoming network hubs. These indicators were directly related to the normal operation of core well-being services like healthcare and education. Due to the decline in remittances and disrupted energy exports, the financial component’s support for overall coupling weakened, and indicator linkages in non-core fields were forced to cease due to insufficient resources.
In terms of component functions, the social component, with its stable internal linkages, provided a prerequisite for the functioning of other components. Whether in the early single-sector governance, middle cross-sector integration, or late core contraction, the social component never ceased to safeguard people’s well-being needs, serving as the “bottom line” that prevented network collapse across phases. The governance-related component played a “regulatory role”, as its institutional design and policy orientation directly influenced the degree of network segmentation and integration, early inter-departmental segmentation was a practical choice to prioritize “single tasks”, while later cross-sector synergy was a proactive adjustment to meet “sustainable development needs”. The financial and environmental components showed the characteristic of “dynamic adaptation to external conditions and policies”: the financial component drove network synergy through resource development in the middle phase but became a constraint after external shocks, and the environmental component transitioned from passive compromise to limited synergy as ecological policies improved. The functional changes of both components were always aligned with Tajikistan’s basic national conditions of “resource dependence and ecological fragility”.

4.2. Coupling drivers of Tajikistan’s SDGs

The coupling drivers of Tajikistan’s SDGs mainly come from four aspects—policy guidance, economic support, governance optimization, and external environment adaptation—and these factors worked together across different development stages.
Policies are key to promoting coupling, with policy orientations in different phases determining the development direction. From 2001 to 2012, phased poverty reduction strategies focused on single people’s well-being goals (World Bank, 2005); from 2013 to 2018, the Strategy for Improving Living Standards (2013-2015) and the “CASA-1000 Energy Programme” were used to integrate cross-sector resources, linking hydropower exports with employment and education, while the policy of “mineral development with supporting ecological restoration” promoted economy-ecology synergy (Republic of Tajikistan, 2013a); under the framework of the National Development Strategy of the Republic of Tajikistan for the Period until 2030, the government implemented emergency livelihood support measures in response to external shocks such as the 2020 pandemic and the 2022 global energy crisis.
Economic conditions provide support for coupling while also creating constraints. Hydropower and mineral development are important pillars: in the middle phase, hydropower export revenues boosted investment in education and healthcare, and mineral rents subsidized rural infrastructure—strengthening internal coupling of the social component (Republic of Tajikistan, 2013a; Ministry of Defence of the Republic of Tajikistan, 2016). However, resource dependence also brought limitations: during large-scale aluminum mining in the 2010s, the male-dominated labor structure in resource development disrupted the synergy between employment and resource utilization, exacerbating gender disparities in regional labor markets (Abdulla and Serikbayeva, 2024); in the late phase, the decline in remittances and constrained energy exports directly weakened the economy’s support for coupling (Ministry of Defence of the Republic of Tajikistan, 2016; Migranyan and Krishtal, 2024).
Governance optimization reduces coupling barriers and improves interaction efficiency. For example, the centralized fiscal payment mechanism and policy designs prioritizing people’s well-being and ecology broke down early inter-departmental segmentation, strengthening links between government expenditure (SDG 16) and indicators of healthcare and freshwater (Ministry of Defence of the Republic of Tajikistan, 2016). The Anti-Corruption Strategy enhances policy implementation efficiency by standardizing resource allocation, lowering the proportion of out-of-pocket health expenditure and reducing gaps in education investment, which in turn indirectly amplifies the synergy effect of all components.
The external environment adjusts the focus of coupling by influencing resource supply and demand. From 2014 to 2018, rising global energy prices increased hydropower export revenues, allowing more resources to be invested in promoting synergy between the environmental and social components (e.g., popularizing clean fuels); the 2020 pandemic and the 2022 global energy crisis, however, brought shocks—the decline in remittances shrank people’s well-being resources and the disrupted energy exports weakened economic support, forcing the scope of coupling to contract to core areas and only maintaining basic linkages between energy and people’s well-being.

5. Conclusions

This study takes Tajikistan as a case, based on 81 SDG indicators and a “Microscale-Mesoscale-Global” coupling network framework, to analyze the coupling dynamic patterns of the country’s SDGs from 2001 to 2023. The analysis results show that Tajikistan’s SDG network exhibits three phases—low-interaction segmentation, high-synergy integration, and core-field contraction, which align with its development processes of post-war reconstruction, quality improvement, and external shock response. The four components have differentiated functions: the social component safeguards the bottom line of people’s well-being, the governance-related component adjusts the network structure, and the financial and environmental components adapt dynamically to resource conditions and policy orientations. The coupling driving force stems from the coordinated effect of policy guidance, economic support, governance optimization, and external environment adaptation.
The results can provide the following references for similar countries. First, in the survival guarantee phase, priority can be given to safeguarding people’s well-being through social sector development. Targeted policies can focus on core needs such as poverty governance and food security, and this approach may lay a foundation for subsequent coordinated development. Second, in the quality improvement phase, attempts can be made to promote cross-sector policy integration while breaking inter-departmental barriers through governance reform. Third, when facing external shocks, consideration can be given to quickly narrowing the coupling scope and concentrating resources to safeguard core areas such as energy and basic services to minimize the risk of overall system coupling breakdown. Fourth, regarding the common contradiction between resource dependence and ecological fragility faced by similar countries, Tajikistan’s practice of “supporting ecological restoration in mineral development” in the middle phase may provide some reference for balancing economic development and ecological protection, and reducing the negative impact of resource development on sustainable development.
This study still has several key limitations that need to be addressed in future research. First, the regional scale coverage is insufficient, and sub-regional heterogeneity analysis is not included. This makes it difficult to fully reflect the impact of different geographical features (e.g., river valley plains and plateau mountains) or urban-rural differences within the country on SDG coupling, resulting in an incomplete depiction of coupling laws. Second, the identification of causal mechanisms is insufficient. Only correlation analysis was used in this study to judge the synergy and trade-off relationships between indicators, and the direction of causality was not clarified, which limits the guidance for policy prioritization. Third, the research on external shocks lacks depth. Although the impact of external shocks was identified, the specific transmission paths of shocks to the coupling system were not systematically analyzed, making it difficult to reveal the in-depth connection between external pressures and internal changes.

Authorship contribution statement

Ranna HAZIHAN: conceptualization, data curation, formal analysis, investigation, methodology, project administration, resources, validation, visualization, writing - original draft, and writing - review & editing; DU Hongru: conceptualization, funding acquisition, project administration, supervision, validation, and writing - review & editing; HE Chuanchuan: data curation and investigation; Kobiljon Khushvakht KHUSHVAKHTZODA: conceptualization; and Bobozoda KOMIL: conceptualization. All authors approved the manuscript.

Declaration of conflict of interest

Kobiljon Khushvakht KHUSHVAKHTZODA is an Editorial Board member of Regional Sustainability and a Guest Editor in Chief of the Special Issue “Green Sustainability in Tajikistan: Bridging Science, Policy, and Community Action” of Regional Sustainability, and was not involved in the editorial review or the decision to publish this article. All authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Abdulla K., Serikbayeva B., 2024. Gender gaps in labor market outcomes in a resource-dependent country. Resources Policy. 90, 104685, doi: 10.1016/j.resourpol.2024.104685.

[2]
Abdullozoda J., Yusufi S., Nandi S., et al., 2024. Informing policy with health labour market analysis to improve availability of family doctors in Tajikistan. Human Resources for Health. 22(1), 63, doi: 10.1186/s12960-024-00946-5.

PMID

[3]
Alayza N., Caldwell M., 2021. Financing climate action and the COVID-19 pandemic: An analysis of 17 developing countries. In: World Resources Institute. Working Paper.Washington, USA,36-42.

[4]
Balata E.E., Pinto H., da Silva M.M., 2022. Latent dimensions between water use and socio-economic development: A global exploratory statistical analysis. Regional Sustainability. 3(3), 269-280.

DOI

[5]
Blanc D.L., 2015. Towards integration at last? The Sustainable Development Goals as a network of targets. Sustainable Development. 23(3), 176-187.

DOI

[6]
Bongaarts J., 2021. FAO, IFAD, UNICEF, WFP and WHO the state of food security and nutrition in the world 2020. Transforming food systems for affordable healthy diets FAO, 2020, 320 p. Population and Development Review. 47, 558-558.

DOI

[7]
Cernev T., Fenner R., 2020. The importance of achieving foundational Sustainable Development Goals in reducing global risk. Futures. 115, 102492, doi: 10.1016/j.futures.2019.102492.

[8]
Cutting E., Jones M., White E., 2024. We put the “UN” in FUN: The mathematical guide to saving the world. University of Colorado Honors Journal. 1-22, doi: 10.33011/cuhj20242839.

[9]
Dai M., Huang S.Z., Huang Q., et al., 2022. Propagation characteristics and mechanism from meteorological to agricultural drought in various seasons. Journal of Hydrology. 610, 127897, doi: 10.1016/j.jhydrol.2022.127897.

[10]
Damanik F.H.S., 2025. Gender and education in the context of Sustainable Development Goals. Entita: Jurnal Pendidikan Ilmu Pengetahuan Sosial dan Ilmu-Ilmu Sosial. 1, 251-266.

[11]
Dawes J.H.P., 2022. SDG interlinkage networks: Analysis, robustness, sensitivities, and hierarchies. World Development. 149, 105693, doi: 10.1016/j.worlddev.2021.105693.

[12]
Eisenmenger N., Pichler M., Krenmayr N., et al., 2020. The Sustainable Development Goals prioritize economic growth over sustainable resource use: a critical reflection on the SDGs from a socio-ecological perspective. Sustainability Science. 15(4), 1101-1110.

DOI

[13]
Fu B.J., Wang S., Zhang J.Z., et al., 2019. Unravelling the complexity in achieving the 17 Sustainable Development Goals. National Science Review. 6(3), 386-388.

DOI

[14]
Gaforzoda B., Yuldashev R., 2023. Irrigation and drainage in the Republic of Tajikistan. Irrigation and Drainage. 72(5), 1230-1240.

DOI

[15]
Ghimire T., Harou A.P., Balasubramanya S., 2023. Migration, gender labor division and food insecurity in Tajikistan. Food Policy. 116, 102438, doi: 10.1016/j.foodpol.2023.102438.

[16]
Hagberg A., Swart P.J., Schult D.A., 2008. Exploring network structure, dynamics, and function using NetworkX. In: Los Alamos National Laboratory. No. LA-UR-08-05495. Los Alamos, the USA.

[17]
Hickel J., 2019. The contradiction of the sustainable development goals: Growth versus ecology on a finite planet. Sustainable Development. 27(5), 873-884.

DOI

[18]
Holden E., Linnerud K., Banister D., 2017. The imperatives of sustainable development. Sustainable Development. 25(3), 213-226.

DOI

[19]
Horvath S.M., Muhr M.M., Kirchner M., et al., 2022. Handling a complex agenda: A review and assessment of methods to analyse SDG entity interactions. Environmental Science & Policy. 131, 160-176.

[20]
Hu B.A., Li Z.Z., Wu H.F., et al., 2023. Coupling strength of human- natural systems mediates the response of ecosystem services to land use change. Journal of Environmental Management. 344, 118521, doi: 10.1016/J.Jenvman.2023.118521.

[21]
Kojo N.C., Sattar S., 2018. Tajikistan - Systematic Country Diagnostic: Making the National Development Strategy 2030 a Success - Building the Foundation for Shared Prosperity. [2025-05-01]. https://documents.worldbank.org/en/publication/documents-reports/documentdetail/430741528356150691.

[22]
Leal Filho W., Trevisan L.V., Rampasso I.S., et al., 2023. When the alarm bells ring: Why the UN sustainable development goals may not be achieved by 2030. Journal of Cleaner Production. 407, 137108, doi: 10.1016/j.jclepro.2023.137108.

[23]
Mackey A., Saalismaa N., Simonett O., 2024. Leave no mountains behind:the sustainable development agenda and mountain areas. In: SchneiderbauerS., PisaP.F., ShroderJ.F., (eds.). Safeguarding Mountain Social-Ecological Systems. Amsterdam: Elsevier, 133-136.

[24]
Mensah J., 2019. Sustainable development: Meaning, history, principles, pillars, and implications for human action: Literature review. Cogent Social Sciences. 5(1), 1653531, doi: 10.1080/23311886.2019.1653531.

[25]
Migranyan A.A., Krishtal I.S., 2024. Economy of Tajikistan: Growth factors against the background of law standards of living. Russian New States and Eurasia. 10(4), 120-132 (in Russian).

[26]
Ministry of Defence of the Republic of Tajikistan, 2016. National Development Strategy of the Republic of Tajikistan for the Period until 2030. [2025-06-01]. https://leap.unep.org/en/countries/tj/national-legislation/national-development-strategy-republic-tajikistan-period-until (in Russian).

[27]
Mishra M., Desul S., Santos C.A.G., et al., 2024. A bibliometric analysis of sustainable development goals (SDGs): a review of progress, challenges, and opportunities. Environmental, Development and Sustainability. 26(5), 11101-11143.

DOI

[28]
Mrowinski M.J., Orzechowski K.P., Fronczak A., et al., 2024. Interplay between tie strength and neighbourhood topology in complex networks. Scientific Reports. 14, 7811, doi: 10.1038/s41598-024-58357-4.

[29]
Naveed A., Ahmad N., Naz A., et al., 2023. Economic development through women’s economic rights: a panel data analysis. International Economics and Economic Policy. 20, 257-278.

DOI

[30]
Nilsson M., Griggs D., Visbeck M., 2016. Policy: Map the interactions between Sustainable Development Goals. Nature. 534, 320-322.

DOI

[31]
Pham-Truffert M., Metz F., Fischer M., et al., 2020. Interactions among Sustainable Development Goals: Knowledge for identifying multipliers and virtuous cycles. Sustainable Development. 28(5), 1236-1250.

DOI

[32]
Pickering J., Jotzo F., Wood P.J., 2015. Sharing the global climate finance effort fairly with limited coordination. Global Environmental Politics. 15(4), 39-62.

DOI

[33]
Pradhan P., Costa L., Rybski D., et al., 2017. A systematic study of Sustainable Development Goal (SDG) interactions. Earth’s Future. 5(11), 1169-1179.

DOI

[34]
Republic of Tajikistan, 2007. National Development Strategy of the Republic of Tajikistan for the Period to 2015. [2025-07-01]. https://leap.unep.org/en/countries/tj/national-legislation/national-development-strategy-republic-tajikistan-period-2015.

[35]
Republic of Tajikistan, 2008. Concept of Environmental Protection in the Republic of Tajikistan (Decree No. 645). [2025-04-03].http://fp7.cawater-info.net/library/rus/tj_env_concept_2008.pdf (in Russian).

[36]
Republic of Tajikistan, 2013a. Living Standards Improvement Strategy of Tajikistan for 2013-2015. [2025-04-26]. https://leap.unep.org/en/countries/tj/national-legislation/living-standards-improvement-strategy-tajikistan-2013-2015.

[37]
Republic of Tajikistan, 2013b. Strategy for Countering Corruption in the Republic of Tajikistan for the Period of 2013 - 2020. [2025-04-30]. https://www.informea.org/en/content/legislation/strategy-countering-corruption-republic-tajikistan-period-2013-2020 (in Russian).

[38]
Resolution of the Government of Tajikistan, 2007. Poverty Reduction Strategy of the Republic of Tajikistan for 2007-2009 . [2025-04-03]. https://www.informea.org/en/content/legislation/poverty-reduction-strategy-republic-tajikistan-2007-2009.

[39]
Roy A., Pramanick K., 2018. Analysing progress of Sustainable Development Goal 6 in India: Past, present, and future. Journal of Environmental Management. 232, 1049-1065.

DOI

[40]
Sachs J.D., LaFortune G., Fuller G., 2024. The SDGs and the UN Summit of the Future. Sustainable Development Report 2024. Dublin: Dublin University Press.

[41]
Sebestyén V., Bulla M., Rédey Á., et al., 2019. Network model-based analysis of the goals, targets and indicators of sustainable development for strategic environmental assessment. Journal of Environmental Management. 238, 126-135.

DOI PMID

[42]
Soto G.H., 2024. Falling behind: Evaluating projected sustainable development goals progress across varied income countries. Sustainable Development. 32(3), 2194-2207.

DOI

[43]
Strokova V., Ajwad M.I., 2017. Tajikistan Jobs Diagnostic: Strategic Framework for Jobs. [2025-05-01]. https://documents.worldbank.org/en/publication/documents-reports/documentdetail/611141486546993528.

[44]
United Nations, 2015. Transforming Our World: The 2030Agenda for Sustainable Development. New York: United Nations.

[45]
Weitz N., Carlsen H., Nilsson M., et al., 2018. Towards systemic and contextual priority setting for implementing the 2030 Agenda. Sustainability Science. 13(2), 531-548.

DOI PMID

[46]
World Bank, 2005. Republic of Tajikistan Poverty Assessment Update (Report No.30853-TJ). Washington: World Bank.

[47]
Wu X.T., Fu B.J., Wang S., et al., 2022. Decoupling of SDGs followed by re-coupling as sustainable development progresses. Nature Sustainability. 5, 452-459.

DOI

[48]
Wu X.T., Fu B.J., Wang S., et al., 2023. Three main dimensions reflected by national SDG performance. The Innovation. 4(6), 100507, doi: 10.1016/j.xinn.2023.100507.

[49]
Yuan M.M., Guo X., Wu L.W., et al., 2021. Climate warming enhances microbial network complexity and stability. Nature Climate Change. 11, 343-348.

DOI

[50]
Zelinka D., Amadei B., 2019. Systems approach for modeling interactions among the Sustainable Development Goals part 1: Cross-impact network analysis. International Journal of System Dynamics Applications. 8(1), 23-40.

DOI

[51]
Zhang J.Z., Wang S., Zhao W.W., et al., 2022. Finding pathways to synergistic development of Sustainable Development Goals in China. Humanities and Social Sciences Communications. 9, doi: 10.1057/s41599-022-01036-4.

[52]
Zhao J.Q., 2022. We should all be feminists—The impact of the protection of women’s rights on economic growth. American Journal of Industrial and Business Management. 12, 1363-1373.

DOI

Options
Outlines

/

〈 〉