Efficiency evaluation of the impact of energy poverty on sustainable development of water resources

Abstract

Water and energy resources are interdependent, yet their sustainable synergy remains understudied. Using China’s 2016–2020 provincial data, this study employs a dynamic two-stage SBM model to assess the sustainable development efficiency of water resources and regional disparities, with energy poverty incorporated as an exogenous variable. Findings reveal: (1) Water-scarce regions achieve higher efficiency due to strict policies and technological innovation, whereas water-rich areas face a “resource curse”. (2) Most provinces exhibit “high production-low sustainability”, indicating development prioritization over conservation. (3) In the east, high energy dependence causes a significant drop in TGR of water sustainability when energy poverty is factored in, widening the efficiency gap. In contrast, in central and western regions, particularly western provinces, TGR rises with the inclusion of energy poverty, supported by stringent ecological policies. Finally, targeted countermeasures are proposed: promoting water-saving technologies and new energy applications in the eastern region; strengthening green manufacturing and eco-compensation in the central region; and developing eco-industries and technological innovations in the western region.

Introduction

Water constitutes both a fundamental resource for human survival and a cornerstone of national welfare¹. Access to clean water is recognized by the United Nations as a fundamental human right, and represents an integral component of the Sustainable Development Goals (SDGs)². However, the world currently continues to face a severe water crisis³. As documented in The Sustainable Development Goals Report 2023: Special Edition by the United Nations, 2.2 billion people globally lacked access to safely managed drinking water services in 2022, while 44% of household wastewater remained untreated⁴. As population expansion and urbanization progression intensify, global water demand is projected to persist^5,6. Significantly, the water crisis resolution pathway cannot be addressed in isolation, but rather is structurally linked to energy poverty, creating a complex systemic challenge that demands integrated solutions⁷. In the context of global climate change, the relationship between energy and water becomes increasingly complex. Properly handling the interdependence of the two and promoting the synergy between water resources sustainable development and energy poverty reduction have garnered increasing attention.

Energy poverty is commonly defined as the inability of individuals in developing nations to access adequate and reliable clean energy sources⁸. In developed countries, while accessibility to modern clean energy is largely ensured, the challenge of energy affordability, commonly termed fuel poverty, continues to affect public health and livelihoods⁹. The International Energy Agency (IEA)’s report Tracking SDG7: The Energy Progress Report 2023 highlights that globally, 675 million people still lack electricity, while 2.3 billion individuals continue to rely on cooking fuels detrimental to health and the environment¹⁰. This energy access (affordability) dilemma poses a significant barrier to achieving sustainable water resources management. Every stage of water resources utilization demands energy to operate pumping, transportation, water treatment, and recycling equipment. The South-to-North Water Diversion Project and desalination projects serve as prime examples of electricity-intensive water processes¹¹. A deeper contradiction lies in the vicious cycle formed by energy poverty and water scarcity: water resources development requires energy support, while energy production depends on water inputs. This “water-energy bond” is severed in energy-poor regions, causing local development to fall into a dual trap of water and energy shortages. The deeply coupled relationship between water and energy necessitates the establishment of a systematic synergistic management framework.

China, as the world’s largest developing country, offers a distinctive case study for examining how energy poverty impacts the sustainable development of water resources. In terms of energy resources, energy poverty in China is marked by both fuel poverty and limited energy access¹². Although China achieved 100% electrification in 2015^13,14 and eradicated absolute poverty in 2020, 48.98% of households still experience energy poverty¹⁵. In terms of water resources, while China ranks sixth globally in total water resources, its per capita water consumption stands at only 2300 cubic meters—less than one-quarter of the global average¹⁶. At the same time, due to the monsoon climate, China’s water resources exhibit uneven spatial and seasonal distribution, further restricting the rational development and utilization of water resources¹⁷. Although innovative initiatives like the South-to-North Water Diversion Project and the West-to-East Gas Pipeline Project have enabled China to achieve remarkable results, challenges persist in terms of imperfect synergistic mechanisms between the energy and water sectors, inadequate policy articulation, and insufficient cross-sectoral information sharing. These issues hinder the achievement of high-quality development requirements for the synergistic safety and security of energy and water. Consequently, transcending the traditional governance paradigm and establishing a water-energy symbiotic system has emerged as a critical pathway to achieving sustainable development.

Therefore, this study explores the nexus between water resources sustainable development and energy poverty within a structured research framework. It assesses how to maximize welfare growth while minimizing water consumption, all within the constraints of water resources carrying capacity, against the backdrop of China’s energy poverty landscape. The remainder of this study is organized as follows: Chap. 2 reviews relevant literature, focusing on water resources sustainable development and energy poverty. Chapter 3 outlines the energy poverty index developed herein, along with associated models and methodologies. Chapter 4 presents and analyzes empirical results, examining energy poverty’s exogenous impact on water resources sustainable development. Chapter 5 concludes with a summary and offers policy recommendations. The primary contribution of this study is that it incorporates energy poverty as an exogenous variable into study on water resources sustainable development and delves into how the fluctuations of this variable impact the sustainable development efficiency of water resources.

Literature review

Research related to sustainable development of water resources

Sustainable development of water resources involves ensuring the long-term availability of water resources and ecosystem health by meeting current socio-economic needs without compromising the ability of future generations to meet their own needs. From this connotation, sustainable development of water resources emphasizes the synergy between ecological protection and efficient water resource allocation, considering the water needs of both present and future generations¹⁸. Accurately assessing the level of sustainable development of water resources is fundamental to alleviating escalating human-water conflicts¹⁹.

In the research on the sustainable development level of water resources in China, the existing literature primarily focuses on quantitative analysis through model construction. In terms of research objects, some studies are based on administrative regions and assess the water use efficiency of provinces and municipalities^20,21,22; some studies focus on river basins, such as the Yangtze River Basin, the Yellow River Basin, and the Ganjiang River Basin, comprehensively analyzing the representative cities of the river’s upper, middle, and lower reaches^23,24,25. There are also studies focusing on the water use efficiency of specific sectors, such as assessing industrial water use efficiency, agricultural water use efficiency, food water use efficiency, and other specific sectors^26,27,28,29. The existing studies primarily employ Data Envelopment Analysis (DEA)^30,31, Stochastic Frontier Analysis (SFA)^32,33, and Indicator System Evaluation Methods³⁴ to evaluate the efficiency of sustainable development of water resources. The DEA method has been extensively applied in the aforementioned studies. Compared to other methods, DEA is particularly suitable for scenarios involving multiple inputs and multiple outputs. It also eliminates the need to consider the correlation between input and output indicators and facilitates the incorporation of undesired outputs³⁵. Meanwhile, the multi-stage DEA model unveils the technical “black box” by modularizing or networking the production process, enabling more detailed efficiency decomposition and bottleneck identification. For example, Zhao et al. employed a two-stage SBM model to partition China’s inter-provincial water utilization system into a first-stage water use process and a second-stage pollution treatment process. They found that the efficiency of the first stage is significantly higher than that of the second stage, and that the second-stage efficiency is the key determinant of overall water resource utilization effectiveness²⁹. In terms of spatial distribution, scholars generally agree that there is significant regional heterogeneity in China’s water use efficiency³⁶. Niu et al. highlight a mismatch between China’s water resource distribution and regional socio-economic development levels. This discrepancy threatens the stability and sustainable development of water-scarce basins and necessitates spatio-temporal regulation of water resources³⁷. Furthermore, numerous scholars have employed econometric models to investigate the factors influencing water resources sustainable development. These factors encompass hydrological and climatic aspects (e.g., precipitation, groundwater content, climate change) and socio-economic elements (e.g., population density, urbanization rate, industrial structure, sewage treatment rate, inter-basin water transfer)^38,39.

Research related to energy poverty

Since the 1990s, low-income households in developed countries have frequently been unable to meet their energy needs with their current income⁴⁰. Therefore, early studies defined energy poverty as residents’ inability to access energy services. The concept of energy poverty has since been further expanded and enriched in the literature, yet no uniform definition has been established. Sefa et al. argued that energy poverty refers to individuals who cannot equitably access clean energy, equipment, or energy-related services⁴¹. Bouzarovski et al. proposed that energy poverty refers to the situation where people cannot access the energy services necessary for social activities or material needs⁴². In developed countries, energy poverty is linked to the financial burden of domestic energy costs, emphasizing energy affordability⁴³. In contrast, China’s energy poverty primarily concerns energy accessibility, characterized by the inefficient use of traditional biomass due to outdated infrastructure. This results in inadequate basic energy services that fail to meet essential survival and material needs. Alleviating energy poverty can help boost household income and promote social equity⁴⁴.

Research indicates that quantifying energy poverty empowers policymakers and researchers to more effectively address energy-related issues⁴⁵. Current methods for assessing energy poverty are categorized into single-indicator approaches, such as the LILEE method, and multi-dimensional indicators, like the MEPI method⁴⁶. The single energy poverty indicator assessment primarily estimates the number of energy-poor households by establishing a low energy income threshold⁴⁷. In contrast, the multi-dimensional energy poverty assessment method is more reliable as it emphasizes measuring energy poverty using multiple energy-related indicators. Changes in energy poverty, along with differences in energy poverty and poverty levels across various groups or regions, have been more distinctly captured⁴⁸. For example, Nussbaumer et al. constructed an evaluation index system that encompasses energy use in five areas: cooking, lighting, household appliances, education, and communication⁴⁹. Li Kang et al. developed a four-dimensional framework for assessing regional energy poverty in China. This framework clarifies the nation’s energy concept and incorporates international energy poverty assessment methods to evaluate the current state and changing trends of energy poverty in the country⁵⁰. HG et al. employed the consensus method alongside EU-SILC survey data to construct a four-dimensional index system for measuring energy poverty, with each dimension assigned a corresponding weight⁵¹. Salman Muhammad et al. developed a multidimensional energy poverty index that integrates energy affordability, cleanability, and accessibility dimensions⁵². Wang Fu et al. developed a seven-dimensional framework to assess China’s energy poverty and its dynamic changes⁵³. Xie et al. identified five key dimensions—cooking fuel, lighting, home appliance services, entertainment and education, and communication—to create a multidimensional energy poverty index and analyze its determinants⁵⁴. The multidimensional energy poverty assessment method breaks down single energy poverty factors into multiple dimensions, thereby enhancing the research’s scientific rigor and practical applicability⁵⁵. From a spatial distribution perspective, energy poverty exhibits geographical disparities. Research indicates that provinces with stronger economies and higher performance are categorized as high achievers in energy poverty alleviation. In contrast, economically underdeveloped provinces face the most significant setbacks and demonstrate the poorest performance in reducing energy poverty⁵⁶.

Research on water and energy resources

Both water and energy are universally recognized as vital resources for modern economic and social development. To address challenges such as climate change and food security, achieving synergistic sustainable development of these resources is essential. In recent years, integrating the sustainable development of water and energy resources has emerged as a hot research topic. Some studies have examined the relationship between water and energy. He et al. noted that water is essential for energy extraction, production, and transportation, while substantial energy inputs are required for the water cycle process. He et al. noted that water is essential for energy extraction, production, and transportation, while substantial energy inputs are required for the water cycle process⁵⁷. Some additional studies have also integrated water and energy into a unified framework, exploring how to achieve synergistic sustainable development of these resources. At the theoretical level, Masoud et al. contend that enhancing energy efficiency, adopting renewable energy sources, and mitigating water pollution can lower the supply costs of energy and water resources, reduce greenhouse gas emissions, and alleviate the environmental pressure associated with economic development⁵⁸. At the application level, Li et al. investigated the potential for sustainable water-energy development in the desalination process. They noted that marine energy could substitute for non-renewable sources like fossil fuels in powering desalination, potentially reducing costs, environmental impact, and supporting sustainable development⁵⁹. However, Hussey et al. contend that under the current policy framework, energy and water policies are predominantly developed independently. This degree of policy fragmentation poses a hindrance to sustainable development in both sectors⁶⁰.

While numerous studies have examined energy and water resource sustainability, few have explored how energy poverty impacts the sustainable development of water resources from an energy poverty perspective. Razmjoo et al. argued that emphasizing energy sustainability in water management enhances urban water use efficiency⁶¹. Shi et al.‘s research, on the other hand, focuses on the agricultural sector and underscoring the negative effects of energy poverty on agricultural water use efficiency through both direct and indirect means. It also emphasizes that addressing energy shortages is essential for enhancing water use efficiency and promoting sustainability⁶². Some studies have examined the dynamic linkages within the water-energy-food nexus, arguing that deficiencies in one sector can pressure the others and underscoring the need for integrated management of these sectors^63,64.

In summary, the existing studies have carried out in-depth research on the sustainable development of water resources, energy poverty, and water-energy synergy, which offer valuable reference for this study. However, these studies predominantly focus on assessing the efficiency of water production, wastewater treatment, and other stages when evaluating water resource utilization efficiency, while neglecting the sustainable development level of water resources.

Although there is a considerable amount of literature that incorporates the water and energy sectors into the same research framework and explores the synergistic management of water and energy, less literature has paid attention to the impact of energy poverty on water sustainability. As a result, the innovations of this study are as follows: First, the process of water resources sustainable development is refined into the production stage and sustainable stage, and the two-stage SBM model is applied to measure the total efficiency and the efficiency of each stage, aiming to identify the key issues in the process of water resources achieving sustainable development. Second, energy poverty is included as an exogenous variable in the model to examine its impact on the sustainable development of water resources, thus providing experience for China in realizing energy poverty reduction and accelerating the realization of SDG6 (Clean water and sanitation) and SDG7 (Affordable and clean energy) goals.

Research model and data description

Dynamic two-stage SBM model

The dynamic two-stage SBM model combines the core advantages of the super-efficient DEA model and the SBM model. On the one hand, the super-efficient SBM model can effectively overcome the limitations of traditional DEA models in distinguishing multiple effective decision units (DMUs), and can further distinguish the efficiency differences of effective decision units; On the other hand, the super-efficiency SBM model considers the issue of unexpected outputs (such as environmental pollution) when measuring efficiency. Compared to traditional DEA models that only focus on expected outputs, the super-efficiency SBM model is a more comprehensive and accurate evaluation method. Therefore, the choice of this model is suitable for topic analysis, which ensures the reliability of the conclusions.

In actual inter-provincial development, the development path of water resources obviously has stages. First, through production activities, it creates economic benefits and generates social, industrial, and agricultural value-added relative to the total water consumption. Second, industries emit pollutants, and through pollution control measures, the amount of pollutants can be reduced and environmental quality can be improved. In this way, a dynamic two-stage path for the sustainable development of water resources is formed.

According to the research purpose of this study and in order to explore the impact of energy poverty on water resources sustainable development, the energy poverty index calculated herein is treated as an exogenous variable and incorporated into the water resources sustainable development system. We can thus observe the level of change in the efficiency of water resources sustainable development with or without the exogenous variable of energy poverty. Given that the essence of water resources sustainable development lies in promoting the comprehensive and coordinated development of social life, industrial production, and the ecological environment, the dynamic two-stage SBM model is applied. This model helps us analyze and evaluate the system efficiency of water resources sustainable development, as well as the efficiency changes influenced by the exogenous variable at each stage and the overall level. The dynamic two-stage process of water resources sustainable development, including the key elements and relationships within each stage, is shown in Fig. 1.

Fig. 1

Dynamic two-stage process of sustainable development of water resources.

A decision-making unit (DMU) is composed of K sub-phases. Assume there are n DMUs, denoted as $\:{DMU}_{i}$ (i = 1,2,…, n). Any DMU contains m inputs $\:{X}_{ij}$(j = 1,2,…, m), p intermediate outputs $\:{M}_{ij}$(j = 1,2,…, p), h re-inputs $\:{H}_{ij}$(j = 1,2,…, h), and r final outputs $\:{Y}_{ij}$(j = 1,2,…, r) in T periods. Two neighboring periods are linked across periods with some of the outputs in period T-1 being some of the inputs in period T, and some of the inputs in period T + 1 being some of the outputs in period T again. There are s carry-over variables for such inter-period inputs, $\:{Z}_{ij}$(j = 1,2,…, s).

Based on the dynamic two-stage SBM model described above, the overall efficiency of the qth DMU in period T is defined as:

$$min\theta _{o} = \frac{{w_{{q1}} \left[ {1 - \frac{1}{{m + s}}\left( {\mathop \sum \nolimits_{{i = 1}}^{m} \frac{{s_{{mo}}^{ - } }}{{x_{{mo}} }} + \mathop \sum \nolimits_{{i = 1}}^{s} \frac{{s_{{so}}^{ - } }}{{z_{{so}} }}} \right)} \right] + w_{{q2}} \left[ {1 - \frac{1}{{p + h}}\left( {\mathop \sum \nolimits_{{i = 1}}^{p} \frac{{s_{{po}}^{ - } }}{{m_{{po}} }} + \mathop \sum \nolimits_{{i = 1}}^{h} \frac{{s_{{ho}}^{ - } }}{{h_{{ho}} }}} \right)} \right]}}{{w_{{q1}} \left( {1 + \frac{1}{p}\mathop \sum \nolimits_{{i = 1}}^{p} s_{{po}}^{ + } /m_{{po}} } \right) + w_{{q2}} \left[ {1 + \frac{1}{{r + s}}\left( {\mathop \sum \nolimits_{{i = 1}}^{r} \frac{{s_{{ro}}^{ + } }}{{y_{{ro}} }} + \mathop \sum \nolimits_{{i = 1}}^{s} \frac{{s_{{so}}^{ + } }}{{z_{{so}} }}} \right)} \right]}}$$

$$s.t.\left\{ {\begin{array}{*{20}l} {\mathop \sum \limits_{{j = 1}}^{n} \lambda _{j} x_{{mj}} = x_{{mo}} - s_{{mo}}^{ - } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \lambda _{j} z_{{sj}} = z_{{so}} - s_{{so}}^{ - } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \lambda _{j} m_{{pj}} = m_{{po}} + s_{{po}}^{ + } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \delta _{j} h_{{hj}} = h_{{ho}} - s_{{ho}}^{ - } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \delta _{j} m_{{pj}} = m_{{po}} - s_{{po}}^{ - } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \delta _{j} y_{{rj}} = y_{{ro}} + s_{{ro}}^{ + } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \delta _{j} z_{{sj}} = z_{{so}} + s_{{so}}^{ + } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \lambda _{j} = \mathop \sum \limits_{{j = 1}}^{n} \delta _{j} } \hfill \\ {s_{{mo}}^{ - } ,s_{{so}}^{ - } ,s_{{po}}^{ + } ,s_{{po}}^{ - } ,s_{{ho}}^{ - } ,s_{{ro}}^{ + } ,s_{{so}}^{ + } ,\lambda _{j} ,\delta _{j} \ge 0} \hfill \\ \end{array} } \right.$$

Here, $\:{\theta\:}_{o}$ is the overall efficiency of the water resources sustainable development system, which is effective if the value of efficiency is 1, and vice versa; $\:\lambda\:$ is a non-negative multiplier for integrating the production stage; $\:\delta\:$ is a non-negative multiplier for integrating the sustainability stage; $\:{s}_{mo}^{-}$ denotes the slack variables for inputs, $\:{s}_{so}^{+}\:\text{a}\text{n}\text{d}\:{s}_{so}^{-}$ denotes the slack variables for carry-over variables, $\:{s}_{po}^{+}\:\text{a}\text{n}\text{d}\:{s}_{po}^{-}$ denotes the slack variables for intermediate outputs, $s_{{ho}}^{ - }$ denotes the slack variables for re-inputs, $\:{s}_{ro}^{+}$ denotes the slack variables for final outputs. The slack variables can reflect the gap between the input and output variables and the target value of optimal allocation, which can provide a reasonable opinion for the improvement of the sustainable development of water resources in each province and the adaptation of resources in the future. Lastly, $\:{w}_{q1}$ and $\:{w}_{q2}$ denote the weights of the qth decision variable in the two sub-stages in period T, respectively.

The efficiency of the production stage subsystem is defined as:

$$\:min{\theta\:}_{1}=\frac{1-\frac{1}{m+s}({\sum\:}_{i=1}^{m}\frac{{s}_{mo}^{-}}{{x}_{mo}}+{\sum\:}_{i=1}^{s}\frac{{s}_{so}^{-}}{{z}_{so}})}{1+\frac{1}{p}{\sum\:}_{i=1}^{p}{s}_{po}^{+}/{m}_{po}}$$

$$\:s.t.\left\{\begin{array}{c}{\sum\:}_{j=1}^{n}{\lambda\:}_{j}{x}_{mj}={x}_{mo}-{s}_{mo}^{-}\\\:{\sum\:}_{j=1}^{n}{\lambda\:}_{j}{z}_{sj}={z}_{so}-{s}_{so}^{-}\\\:{\sum\:}_{j=1}^{n}{\lambda\:}_{j}{m}_{pj}={m}_{po}+{s}_{po}^{+}\\\:{\sum\:}_{j=1}^{n}{\lambda\:}_{j}=1\\\:{s}_{mo}^{-},{s}_{so}^{-},{s}_{po}^{+},{\lambda\:}_{j}\ge\:0\end{array}\right.$$

The efficiency of the sustainability stage subsystem is defined as:

$$min\theta _{2} \frac{{1 - \frac{1}{{p + h}}\left( {\mathop \sum \nolimits_{{i = 1}}^{p} \frac{{s_{{po}}^{ - } }}{{m_{{po}} }} + \mathop \sum \nolimits _{{i = 1}}^{h} \frac{{s_{{ho}}^{ - } }}{{h_{{ho}} }}} \right)}}{{1 + \frac{1}{{r + s}}\left( {\mathop \sum \nolimits _{{i = 1}}^{r} \frac{{s_{{ro}}^{ + } }}{{y_{{ro}} }} + \mathop \sum \nolimits _{{i = 1}}^{s} \frac{{s_{{so}}^{ + } }}{{z_{{so}} }}} \right)}}$$

$$\left\{ {\begin{array}{*{20}l} {\mathop \sum \limits_{{j = 1}}^{n} \delta _{j} m_{{pj}} = m_{{po}} - s_{{po}}^{ - } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \delta _{j} h_{{hj}} = h_{{ho}} - s_{{ho}}^{ - } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \delta _{j} y_{{rj}} = y_{{ro}} + s_{{ro}}^{ + } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \delta _{j} z_{{sj}} = z_{{so}} + s_{{so}}^{ + } } \hfill \\ {\mathop \sum \limits_{{j = 1}}^{n} \delta _{j} = 1} \hfill \\ {s_{{po}}^{ - } ,s_{{ho}}^{ - } ,s_{{ro}}^{ + } ,s_{{so}}^{ + } ,\delta _{j} \ge 0} \hfill \\ \end{array} } \right.$$

Here, $\:{\theta\:}_{1}$ and $\:{\theta\:}_{2}$ denote the efficiency of the assessment decision unit in the production stage subsystem and the sustainability stage subsystem, respectively.

Technology gap ratio

Technology gap ratio (TGR) is an important indicator in the framework of common frontiers. This study measures the meta-frontier technical efficiency (MTE) of the kth DMU based on the meta-frontier of the 29 DMUs according to the model construction described above. This meta-frontier is made up of the most efficient DMUs in each of the 3 groups together, and based on different groups under the boundary, the group frontier technical efficiency (GTE) can be calculated. Based on GTE and MTE under different groups, it is possible to calculate the technology gap efficiency (TGR) value, which measures the technological status of each DMU and the difference in the level of technology between the groups, reflecting the gap between groups under the meta-frontier and group frontier. The value of this ratio is between 0 and 1. The closer the value is to 1, the closer the actual technology level of the decision unit is to the optimal technology level of the common frontier, i.e., the higher the technology level. The specific calculation formula is:

$$\:0\le\:TGR=\frac{MTE}{GTE}\le\:1$$

Construction of the indicator system and data sources

Selection of research variables and construction of the indicator system

The sustainable development level of water resources is affected by social life and industrial production. This study divides the sustainable development of water resources into the production stage and the sustainable stage. Combined with the accessibility of data, the evaluation index system is constructed following the principles of systematicity, comparability, comprehensiveness and scientificity. Specific indicators are shown in Fig. 2.

• (1) The input indicators for the production stage are employment and total water supply. These input indicators are the basis for measuring the sustainable development of water resources as they affect industrial production and residential life. Fixed assets carried over from the previous period are also used as one of the inputs.

• (2) Intermediate output indicators include gross domestic product and economic efficiency of water resources. The sustainable management of water resources can promote economic activities and thus increase the GDP. Water resources economic efficiency is an output indicator in the production stage. It assesses the direct economic benefits from water resources utilization, including income from water tariffs and commercial activities related to water treatment and distribution.

• (3) The reinvestment indicator is the total investment in industrial pollution control. This indicator reflects the scale of investments made to improve the quality of water resources, including wastewater treatment facilities, pollution control technologies, etc.

• (4) The final output indicator is the final output pollution improvement index, which measures the effect of water resources management on environmental quality improvement. In addition, fixed assets generated are carried forward to the next period.

• (5) The link, i.e., waste water, exhaust gas, and waste, serves as variables linking the production and sustainable stages.

• (6) The exogenous variable is energy poverty. Energy poverty restricts access to and investment in clean energy and increases the use of environmentally harmful energy sources, further exacerbating water resource pollution and waste. It also exacerbates social inequality, restricts the ability of governments and communities to formulate policies and invest in water resources management and protection, and poses a threat to the sustainability of water resources, so energy poverty is treated as an exogenous variable. Energy poverty was assessed with reference to the methodology proposed by Li Kang (2014)⁵⁰, which provides a reliable approach for measuring energy poverty levels.

Considering the data availability and the needs of this study, we finally select 14 indicators from four aspects: accessibility of energy services, cleanliness for domestic use, completeness of energy management, and affordability and efficiency of domestic energy use, to establish an assessment index system. The specific indicators are shown in Table 1. The selected indicators are standardized and weighted using the entropy weighting method. Then, finally, the energy poverty score of each province is calculated using the weighted sum method.

Fig. 2

System of energy poverty impacts on the sustainable development of water resources.

Data sources and grouping

In view of the data availability and the fact that some data in Tibet are missing, the relevant data of 29 provinces in China from 2016 to 2020 were finally selected as the study object. The data come from China Statistical Yearbook, China Environmental Statistical Yearbook, China Energy Statistical Yearbook, China Rural Statistical Yearbook, and the statistical yearbooks of each province (city and autonomous region). In terms of grouping, due to the large regional span between provinces and the differences in geographic patterns, economic conditions, water resources utilization and industrial governance inputs, this study divides the 29 provinces into eastern, central and western regions based on their geographic locations. The details are shown in Table 1.

Table 1 Specific Groupings.

Results

Temporal analysis

As shown in Table 2, under energy poverty, most provinces exhibited low water resources sustainable development efficiency values during 2016–2020, with a fluctuating upward trend. These results indicate that China’s water resources sustainability has improved, although the overall level remains low. This improvement may result from China’s strategic water projects effectively alleviating localized water shortages, thereby maintaining water security in most regions. Furthermore, the regional disparities in water resources sustainable development efficiency further support this observation.

Table 2 Average value of provincial water resources sustainable development efficiency in China, 2016–2020.

The data clearly show that Beijing, Ningxia, Tianjin, and Qinghai demonstrate significantly higher water resources sustainable development efficiency compared to other provinces. With the exception of Qinghai, these provinces are situated in northern and northwestern China, regions characterized by comparatively limited water resources. These water-stressed conditions may compel local governments to implement stringent water resources policies and enhance water utilization technologies, leading to improved the efficiency values of water resources sustainable development. On the contrary, in areas with relatively abundant water resources such as Anhui, Guangxi, and Hubei, they may experience the “resource curse” phenomenon due to insufficient endogenous motivation, resulting in inefficiency in the sustainable development of water resources.

Beyond resource endowments, the efficiency values of water resources sustainable development is also associated with the economic structure’s path dependency. The analysis reveals that provinces including Shanxi, Shandong, and Henan exhibit significantly lower average efficiency compared to the national mean, demonstrating a strong correlation with their long-standing industrial structures. Shanxi’s economy relies heavily on resources extraction, characterized by substantial water consumption and severe pollution levels; Shandong’s industrial sector is predominantly composed of heavy chemical industries, exhibiting elevated levels of both water consumption and pollution emissions; Henan’s economy is predominantly agricultural, with comparatively extensive water utilization practices. Furthermore, its industrial sector retains water-intensive and highly polluting industries. When combined with a sizable population and substantial domestic water demand, these factors exacerbate inefficient water resources utilization. Deeply speaking, these provinces’ industrial structures remain entrenched in a water-intensive, highly polluting development model, consistent with the “revised EKC“⁶⁵. This suggests that when economies are constrained by path dependence, environmental quality may persist at deteriorated levels long after the initial industrialization phase, owing to structural-institutional dual rigidity. By comparison, Qinghai Province has achieved DEA efficiency in water resources sustainable development, having overcome path dependence through industrial restructuring. Despite possessing some of China’s most abundant water resources, Qinghai has transitioned its economic development model from traditional animal husbandry to an ecological tourism-oriented economy. Ecological tourism development boosts local economic vitality while facilitating ecological restoration and water conservation through incentive mechanisms, establishing a virtuous cycle between economic growth and environmental protection. This comparison demonstrates that industrial restructuring serves as a crucial mechanism for overcoming the water-intensive, highly polluting development path dependency. In provinces with heavy industrial or agricultural economies, advancing industrial green transformation, enhancing ecological compensation mechanisms, and achieving systemic decoupling between economic growth and resources consumption represent essential pathways toward sustainable development. Qinghai’s transformation experience offers a replicable model for water resources sustainability in other water-abundant regions.

Figure 3 presents histogram trends of water resources sustainable development efficiency, with the overall pattern classified into six distinct categories. The analysis reveals distinct patterns in water resources sustainability efficiency trends across provinces: (1) Shanghai, Jiangxi, Zhejiang, Hunan, and Guangxi exhibit consistent upward fluctuations; (2) Jiangsu, Chongqing, Hubei, and Shaanxi demonstrate persistent downward fluctuations; (3) Shanxi, Yunnan, Sichuan, Gansu, Anhui, and Fujian display initial increases followed by declines; (4) Jilin, Hebei, and Heilongjiang show initial decreases followed by increases; (5) Shandong, Tianjin, Beijing, Ningxia, Henan, Qinghai, and Hainan maintain relatively stable efficiency levels with minor fluctuations; and (6) Inner Mongolia, Guizhou, Guangdong, and Liaoning experience significant efficiency volatility. Overall, 2017 emerged as a pivotal turning point, marking substantial improvements in the efficiency of sustainable water resources across most provinces. This observed shift likely stems from China’s 2016 introduction of the National Water Resources Protection and Utilization Plan (2016–2020), which established five-year targets for: (1) enhanced water conservation, (2) improved utilization efficiency, and (3) optimized allocation mechanisms. Following this policy framework, provincial governments implemented measures that significantly enhanced water resources sustainable development efficiency. These findings demonstrate that effective policy interventions play a crucial role in advancing water resources sustainability. These results indicate that regional governments with low water resources sustainable development efficiency must prioritize this challenge, developing and enforcing targeted policies to regulate both industrial and domestic water usage, thereby enhancing local water resources conservation and sustainable management.

Fig. 3

Trend map of sustainable development efficiency of water resources.

Stage-wise efficiency analysis

The two-stage efficiency analysis differentiates performance between production and sustainability stages, while identifying the underlying drivers of variation in water resources sustainable development efficiency. Table 3 reveals a distinct pattern where production stage efficiency consistently exceeds sustainability stage efficiency in water resources utilization. Beijing, Ningxia, and Qinghai demonstrate optimal efficiency (DEA score = 1) across both stages, indicating successful coordination between water resources utilization and environmental conservation. In addition, the production stage efficiency in other provinces is significantly higher than that in the sustainable stage, indicating that the investment in pollution control in most provinces has not been effectively transformed into environmental improvement effects, which may suggest the presence of input congestion⁶⁶. These findings suggest China’s water resources management currently emphasizes development and utilization over conservation and sustainability, indicating substantial potential for improvement in sustainable water resources governance nationwide.

Table 3 Two-stage efficiency values.

Specifically, Sichuan ranks second only to Tibet in water resources endowment while maintaining the highest economic development level among western provinces. Sichuan demonstrates a production stage efficiency of 0.82 compared to a sustainability stage efficiency of merely 0.24, which indicates that the province’s water resources are more efficient in terms of development and production, but are wasted and polluted in a more serious way. On the contrary, Ningxia, as an extremely water-scarce region in China, is efficient in both the production and sustainability stages of its water resources. Despite the fact that its natural conditions are not dominant, it is possible to see that the local government attaches great importance to the utilization and protection of water resources. Further observation reveals that during the period 2016–2020, the efficiency of the production stage of water resources in each province has a clear trend of progress, but the fluctuation of the efficiency in the sustainability stage is more random. This is because the improvement of water resources production efficiency brings immediate economic and social benefits, while efficiency improvement in the sustainable stage is beneficial in the long term. Therefore, the regions are highly motivated to improve the efficiency of the production stage, while the efficiency of the sustainability stage is more closely related to the degree of policy stringency, so the motivation is not high.

In this study, to further discuss the correlation between the two-stage efficiency, we also construct a water production-sustainability efficiency mean matrix by taking the national water production efficiency mean value of 0.93 and the sustainability stage efficiency mean value of 0.38 as the dividing lines to separate the provinces and cities into four regions in Fig. 4: A (Efficient Production, Efficient Sustainability), B (Inefficient Production, Efficient Sustainability), C (Inefficient Production, Inefficient Sustainability), and D (Efficient Production, Inefficient Sustainability). The scatter points are evenly distributed in areas A, C, and D, and there is no obvious trend of centralized distribution in areas A, C, B or D. Regarding regional distribution characteristics, most provinces are concentrated on the right side of the matrix, and the efficiency of the water resources production stage is relatively high. These findings reveal that although the efficiency is high in the water resources production stage, this efficiency is not effectively transmitted to the sustainability stage, resulting in a lack of a significant positive correlation between the efficiency of the production stage and the efficiency of the sustainability stage.

There are many provinces in C (Inefficient Production, Inefficient Sustainability) and D (Efficient Production, Inefficient Sustainability). Their water resources production stage efficiency is concentrated near the national water production efficiency mean value, while the sustainability stage efficiency is at a low value level, reflecting the widespread problem of “focusing on development and neglecting protection” in water resources management. Specifically, provinces that rely on water-intensive industries, such as Hunan, Gansu, Hebei, etc., generally face the problems of high water consumption and low returns. The dependence of traditional industries such as agriculture and heavy chemical industry on water resources has prevented these regions from effectively managing water resources while improving production efficiency, resulting in low efficiency in their sustainability stage. In addition, although economically developed regions, such as Shanghai and Jiangsu, have performed excellently in the efficiency of the production stage, due to the acceleration of urbanization and the surge in water demand, water resources protection policies have not been strengthened simultaneously, resulting in the inability to improve the sustainability efficiency at the same time. While pursuing economic growth, these regions have failed to balance the protection and utilization of water resources, further exacerbating the imbalance in water resources management.

Fig. 4

Productivity Matrix, 2016–2020.

Technical efficiency analysis

Analysis of technology gap ratio

Figure 4 illustrates the differences in two-stage efficiency across various regions, revealing notable regional agglomeration characteristics. To delve deeper into the efficiency disparities among these regions, this study introduces the Technical Gap Ratio as a crucial metric for measuring the differences in water resources sustainable development efficiency between regions. The TGR value reflects the gap between the common frontier technical efficiency and the actual technical efficiency of each region. The closer the TGR value is to 1, the smaller the efficiency gap under the two frontiers, indicating that the actual technical level is closer to the optimal common frontier level. Meta-Frontier Technical Efficiency (MTE) and Group-Frontier Technical Efficiency (GTE) represent the efficiency values under different frontier conditions.

The average meta-frontier technology efficiency, average group frontier technology efficiency, and average technology gap ratio in 2016–2020 in the east, central, and west regions are shown in Table 4; Fig. 5. The specific changes are as follows.

• (1) The TGR in the eastern region is close to 1, indicating that the common frontier technical efficiency and group frontier technical efficiency in this region are highly consistent. This reflects that the water resources sustainable development level in the eastern region has approached an optimal state, characterized by high resources utilization efficiency. The results suggest that water resources sustainable development in the eastern region has reached a relatively mature stage, where external factors exert minimal influence on efficiency.

• (2) The TGR in the western region is close to the national average level. While its water resources sustainable development efficiency is slightly lower than that of the eastern region, it outperforms the central region. The small gap between the common frontier efficiency and group frontier efficiency in the west indicates relatively better water resources sustainable development efficiency in this region. Although economic development and industrial structure still require improvement, the western region boasts vast land, abundant resources, and low population density, providing significant potential to develop innovative green energy while maintaining ecological sustainability. In the future, efforts should focus on strengthening green ecological conservation and promoting the sustainable use of water resources to further enhance overall efficiency.

• (3) The TGR in the central region is below the national average and exhibits a fluctuating upward trend, indicating a significant gap between common frontier efficiency and group frontier efficiency. This implies that the central region still has substantial room for improvement. Specifically, the central region faces challenges in high-efficiency water resources development and utilization, with notable shortcomings in resources allocation, technical application, and management capacity. Thus, enhancing the water resources sustainable development efficiency in the central region and promoting the popularization of high-efficiency technologies and green development transformation have become key issues that need urgent resolution in this region.

Table 4 Average technology gap ratio, 2016–2020.

Fig. 5

Line Graph of Technology Gap Ratios for Groups and the Country, 2016–2020.

Analysis of technology gap ratio considering exogenous variable

To further quantify the effect of the exogenous variable on efficiency, this study will focus on a comparative analysis of the technical drop ratio (TGR) in two scenarios: one that considers energy poverty as the exogenous variable and another that does not consider the exogenous variable. The results are shown in Table 5. By comparing the changes in TGR with and without the exogenous variable, we can deeply reveal the specific impact of energy poverty on water management efficiency in various regions. Especially among different regions, the differences in TGR will provide us with clear evidence to analyse the effectiveness of external interventions in improving water resources sustainable development efficiency and provide theoretical basis and data support for the formulation of more precise regionalization policies.

When evaluating the TGR of water resources sustainable development efficiency with or without exogenous variables, the TGR in the eastern region decreased after introducing energy poverty factors, especially in provinces such as Shandong and Jiangsu. The introduction of the energy poverty factor has increased the efficiency gap in these regions. This shows that the high dependence on energy in the eastern region has made energy access costs and supply issues have more direct and significant impacts on water resources management efficiency. Especially in these economies with high energy demand, energy shortages not only affect industrial production efficiency, but also increase the tightness of water resources, further hindering the sustainable development of water resources.

Unlike the eastern region, in the western and central regions, the TGR of water resources sustainable development efficiency has witnessed a slight increase after the factor of energy poverty is taken into account, particularly in the western region. This shows that energy poverty factors have prompted these regions to adopt more efficient and sustainable water resources management strategies, improving the efficiency of water resources sustainable development. The performance in the western region is particularly prominent, perhaps because energy poverty has inspired local governments and enterprises to pay more attention to energy conservation and water-saving technologies, thus promoting the efficiency of water resources management. However, in the central region, although the impact of energy poverty on the TGR of water resources sustainable development efficiency is relatively limited, in provinces like Shanxi and Hunan, energy poverty factors have had a positive impact on TGR, but with a relatively small increase. These findings suggest energy poverty may incentivize improved water conservation practices and more efficient resources management in affected regions, thereby contributing to enhanced water sustainability.

Table 5 Technical efficiency in 29 provinces based on Meta-frontier.

Discussion

This study reveals the key driving mechanisms of the efficiency of sustainable development of water resources and its regional heterogeneity. The main findings have important expansion significance for the existing theoretical system.

• (1) Regions with water shortages have achieved higher efficiency through rigid policy constraints and technological intensification, while regions with abundant water resources have fallen into the “resource curse” due to the lack of endogenous motivation, confirming the “path dependence” effect in resource management.

• (2) The two-stage efficiency analysis shows that most provinces in China have a “Efficient production-Inefficient sustainability” model, reflecting the common problem of “focusing on development and neglecting protection” in the current sustainable development of water resources.

• (3) There is the regional heterogeneity of the impact of energy poverty on the efficiency of sustainable development of water resources. In the eastern region, energy poverty is manifested in unstable energy supply and high industrial energy costs. After considering energy poverty, the TGR decreases by 0.06 on average, which reduces the technical efficiency of sustainable development of water resources. Especially in major industrial provinces such as Shandong and Jiangsu, energy shortages have exacerbated the inefficient use of water-intensive industries. It is worth noting that this negative impact is more significant in late industrialized regions, indicating that the regional development stage is an important variable in regulating the effect of energy poverty. In contrast, after the introduction of energy poverty variables, the technology gap ratio in the western region increased by an average of 0.08, indicating that energy poverty factors have promoted the improvement of technical efficiency. This effect is particularly prominent in Shaanxi and Inner Mongolia, which have successfully transformed energy poverty pressure into technological innovation momentum by implementing strict ecological protection policies, such as the “ecological asset exit audit” in Inner Mongolia, and promoted the development of water-saving technologies and renewable energy, bringing the TGR value close to the optimal level. This result not only confirms the assertion of UNDP⁶⁷ that “policy intervention can transform energy poverty into innovation power”, but also reveals the synergistic mechanism between policy intensity and resources endowment. In regions with strong policy enforcement, the constraining effect of energy poverty is effectively reversed, even becoming a driving force for technological progress; whereas in areas with relatively insufficient policy intensity but similar resources endowments, the potential promoting effect of energy poverty on technological advancement fails to materialize. This differentiated effect highlights the decisive role of policy regulation in the process of resource transformation. This finding challenges the traditional view that energy poverty will inevitably reduce resource efficiency⁶⁸.

• (4) This study not only provides valuable experience for China’s water resources management, but also provides an important reference for the sustainable development of global water resources. First, water resources management in different regions faces different challenges, so regionally differentiated water resources management policies should be adopted globally. Especially in areas with scarce water resources or weak water resources management, targeted policies should be formulated according to local conditions to improve the efficiency of water resources utilization. Regions with scarce water resources can learn from the practices of the eastern region in China to improve the efficiency of water resources management through technological innovation and intelligent water systems. In regions with relatively abundant resources but relatively weak management, attention should be paid to the green economic transformation experience of the central region, and water resources waste and pollution should be reduced by promoting the widespread application of green manufacturing and water-saving technologies. Secondly, the study shows that energy poverty is not only a challenge, but can also be transformed into a driving force for promoting technological innovation and the application of water-saving technologies. This experience is particularly suitable for countries facing energy shortages, and thus can motivate local governments and enterprises to improve water resources management through technological innovation.

Conclusions

Taking energy poverty as an exogenous variable, this study divides the process of water resources sustainable development into two associated sub-system stages, production and perpetuation, and adopts a two-stage dynamic SBM model to measure the efficiency of the sample data of 29 provinces and municipalities from 2016 to 2020. To exclude the influence caused by regional differences, this study applies the common frontier model to group the 29 provinces and cities and to measure their technical efficiency values, so as to better explore the influence of energy poverty on the efficiency of water resources sustainable development.

This study reveals three key characteristics of China’s water resources sustainable development efficiency through empirical analysis. First, water-poor areas have successfully improved the efficiency of water resources utilization by implementing rigid policy constraints and technological intensification. On the contrary, in areas with relatively abundant water resources, due to the lack of endogenous motivation, they often fall into the dilemma of “resource curse”, resulting in low efficiency of water resources sustainable development. Secondly, further two-stage efficiency analysis shows that there is a widespread structural contradiction of “Efficient production-Inefficient sustainability” across the country, with the efficiency level of the production stage far exceeding that of the sustainability stage, reflecting the prominent problem of emphasizing development over protection. Finally, the impact of energy poverty shows obvious regional heterogeneity. The eastern region has reduced efficiency due to energy cost pressure, while the western region has transformed energy poverty into innovation momentum under strict ecological policies.

Based on the regional heterogeneity discovery of this study, the following policy recommendations are put forward for the characteristics of different regions in China. The economic development of the eastern region is relatively rapid, and water resources management is facing great pressure. It is recommended to strengthen water resources protection policies, especially the management of high-water consumption industries, and increase investment and application of renewable energy. The central region relies on heavy industry and agriculture, and the efficient use of water resources faces challenges. We should accelerate the promotion of green industrial transformation and improve the ecological compensation mechanism. The sustainable development of water resources in the western region is relatively efficient, but the economic structure still needs to be optimized. It is recommended to further strengthen ecological protection policies, promote the development of green ecological industries, especially ecological tourism and green agriculture, and achieve a win-win situation for water resources protection and economic development. Energy poverty has played a role in promoting technological innovation in the western region. It is recommended to further increase support for water-saving technology and renewable energy technology to promote the continuous improvement of local water resources management levels.