Assessing rural sustainability in Guoliang village, China: an expectation livelihood prophecy approach

Rural sustainability is key to maintaining prosperity in developed rural areas, yet many existing evaluation methods focus on material resources while overlooking villagers’ perceptions. This study introduces an ecological–livelihood–perception (ELP) model that extends the sustainable livelihoods framework (SLF) by refining its natural capital dimension to explicitly incorporate ecological quality and villagers’ environmental perceptions, alongside a psychological perception derived from a self-fulfilling prophecy theory. We conducted a structured survey of 100 households in Guoliang village. Then, 21 standardized indicators spanning the natural, financial, physical, social, human, and ecological dimensions were introduced to evaluate the survey data. After reliability screening, principal component analysis (PCA) was applied to construct a single composite score for each capital, capturing the dominant variance in the data without multicollinearity. PCA combined with entropy-based weighting was applied to reduce collinearity and derive indicator weights. Descriptive results show that normalized indicator means range from 0.15 to 0.86, and indicator weights are relatively balanced, ranging from 0.044 to 0.051, with the highest-weight indicator reaching 0.051. The sustainable individual perception (SIP) index was calculated from qualitative items and was linked to the six capital scores through an ordinary least squares regression. The results reveal a threshold pattern: once a basic level of natural capital is met, it leads to further gains, especially in social and ecological assets, which are strongly associated with higher psychological optimism. Households exhibiting high SIP consistently possess balanced and multidimensional capital portfolios, whereas those with low SIP show deficits across most dimensions. The ELP model, therefore, provides a more comprehensive tool for rural sustainability evaluation by integrating objective livelihood assets with subjective perceptions, offering practical insights for targeted policy interventions and development strategies.

Introduction

In the course of human development, rural areas have played a pivotal role in national progress (Long, 1997). In recent years, many villages have pursued industrial transformation in countries such as China (Zhang et al., 2024), Spain, and India (Gómez Valenzuela and Holl, 2024). These efforts are often driven by the need to address rural depopulation (Belton and Filipski, 2019), economic stagnation, and widening regional disparities (Loras-Gimeno et al., 2025). Nevertheless, rural areas in the majority of countries continue to face constraints in improving living conditions. Despite local innovations, persistent challenges directly undermine the sustainability of villages (Zhu J. et al., 2021). Accordingly, a transparent and standardized evaluation system is required to support comprehensive analysis and actionable policy recommendations.

It has been widely documented that governments and researchers have developed multiple evaluation models to assess rural development. For example, a three-dimensional framework of rural sustainability was constructed using spatial regression to reveal heterogeneous urbanization effects in the Beijing, Tianjin, and Hebei region (Hu et al., 2022). A composite SDG-based indicator system with Pareto ranking was proposed to evaluate rural sustainability and inter-dimensional trade-offs in Zhaoyuan city (Liu et al., 2024). A systems-oriented perspective on rural sustainable land-use has also been advanced (Li and Lei, 2023). However, despite this diversity, refined and standardized criteria remain limited (Valizadeh and Hayati, 2025). Moreover, most assessments insufficiently incorporate villagers’ subjective perspectives as drivers of development trajectories despite their recognized complexity.

The sustainable livelihoods framework (SLF) proposed by the Department for International Development (1999) conceptualizes livelihoods through five forms of capital—natural, financial, physical, human, and social—each of which interacts with the vulnerability context to shape the outcomes (Natarajan et al., 2022). Although widely applied in poverty and rural studies (Thomas et al., 2025), environmental degradation and ecological dynamics remain underrepresented. To address this limitation, the framework has been extended to include an explicit ecological dimension, termed the sustainable livelihoods framework with ecology (SLFE), which quantifies ecological capital (Shi and Yang, 2022).

Beyond material and structural dimensions, the role of cognition and perception has increasingly been recognized in shaping rural sustainability. The self-fulfilling prophecy (SFP) theory, first articulated by Merton (1948), explains how individual beliefs influence objective outcomes through behavioral feedback. In rural contexts, farmers’ expectations and psychological judgments regarding policy feasibility, industrial prospects, and environmental conditions have been shown to affect participation behavior and, in turn, the performance of revitalization efforts (Van Amsterdam et al., 2025). Appadurai (2004) further conceptualized the imagination of future life as a form of institutional capital that can determine whether poverty structures are overcome. Consistent with this view, perceptions of income adequacy and work conditions have been found to shape daily behavior and productivity (Maican et al., 2021). These perspectives indicate that sustainable development is not determined solely by resources or infrastructure but also by villagers’ perceived opportunities and confidence. Integrating SFP into sustainability assessment, therefore, enables capturing both objective livelihood conditions and subjective psychological dimensions, thus helping explain why similarly endowed villages may exhibit divergent development trajectories.

In addition to conceptual frameworks, a variety of analytical approaches have been applied to assess rural livelihood sustainability. Intelligent and data-driven methods have been developed to optimize complex problems in the humanities and social sciences (Xu et al., 2019; Yu et al., 2022; Zhu G. et al., 2021). In particular, optimization and statistical modeling techniques from computer science have increasingly been adapted to reduce computational complexity and improve the interpretability of multidimensional systems (Zhou et al., 2023a). Among the commonly used quantitative approaches, factor and cluster analyses play a central role in reducing indicator dimensionality and identifying household typologies, thereby clarifying structural relationships among livelihood variables (Hsu and Peng, 2023). Principal component analysis (PCA), first introduced by Pearson (1901), has become one of the most widely adopted tools for synthesizing multidimensional indicators into composite indices (Mardia, 1979; Jiang et al., 2018). Compared with traditional single-variable strategies, PCA captures interrelationships among indicators and extracts components that explain the highest variance, thereby providing objective weights and minimizing redundancy (Zhang et al., 2020; Zhan et al., 2021). Building upon these strengths, PCA is utilized within the SLFE framework in the present study to construct a comprehensive evaluation of rural sustainability, ensuring that each dimension of livelihood and ecological capital contributes proportionally to the overall assessment.

To bridge the gap between sustainability evaluation and the role of human attitudes and productivity, an expectation–livelihood–prophecy (ELP) model is proposed and applied in a Chinese village case study. The ELP model integrates objective livelihood–ecological indicators and subjective psychological perceptions into a unified assessment framework. By linking structural conditions with behavioral attitudes, a more comprehensive evaluation is advanced. As shown in Figure 1, the model retains the foundational SLF elements of the vulnerability context and livelihood capitals while explicitly integrating subjective perception, which is measured via the sustainable individual perception (SIP) index. Perceptions are treated as both an outcome and a feedback factor. In tourism-affected or transforming rural areas, vulnerability is characterized by exposure to policy shifts, environmental degradation, and economic shocks.

Figure 1

Figure 1. Conceptual framework of the ELP model.

Analytical methods and indicator system

Study area and data sources

Guoliang village, which was investigated in this case study, is situated on the east bank of the Xiangjiang River in the northern part of Wangcheng District, Changsha, China. It benefits from naturally abundant resources and has a long-standing reliance on agriculture, as shown in Figure 2. It is part of the Tongguan subdistrict, which includes Tongguan town, Shutangshan subdistrict, Guoliang village, and part of the Shenjiaqiao community, covering an area of over 90.31 km² and having a large population of 42,286, according to data from the National Bureau of Statistics of China (2017) (Tan, 2017).

Figure 2

Figure 2. Schematic diagram of the geographical location of Guoliang village.

Tongguan subdistrict is known for its famous history since the late Tang dynasty; Tongguan is renowned for its official dragon kilns that mastered underglaze polychrome ceramics, firing tens of thousands per batch at 1,200 °C. For centuries, the city of Changsha has been renowned for its poetic inscriptions, and its pottery traveled globally via river routes and the Maritime Silk Road to Southeast Asia and beyond, cementing the city’s enduring archaeological legacy as a pottery capital (Guo et al., 2021). Moreover, since the establishment of Tongguan town, this sub-district has built two national 4A-level tourist attractions and one national 3A-level tourist attraction. Among these attractions, Tongguan kiln is the most welcomed, receiving over 1.5 million tourists and achieving a comprehensive cultural tourism income of 780 million RMB. In addition, the total income of Tongguan town reached 1,839.7 million RMB by 2019 (Li, 2020).

Guoliang village covers an area of 9.1 square kilometers, consists of 47 villager groups, and has a population of 3,899 in 1,002 households. It possesses 229.6 hectares of cultivated land, with an average of approximately 0.059 hectares of arable land per person (Tan, 2017).

Methodology

Sustainability indicator selections with PCA

Many intelligent methods are used to solve different problems in human productivity (Xu et al., 2019; Yu et al., 2022; Zhu G. et al., 2021). In the field of computer science, optimization techniques are commonly used to minimize complex computational functions (Sun and Jiang, 2022; Zhou et al., 2023b). Among the widely applied statistical tools for rural livelihood sustainability assessment, factor and cluster analysis play a central role in reducing indicator dimensionality and identifying household typologies. These approaches aim to capture the underlying structures among multiple livelihood variables and have been used extensively to construct sustainability indices or classify livelihood strategies (Hsu and Peng, 2023). PCA was first proposed by Pearson (1901), and its core idea is based on the principle of spatial least-squares fitting. This method aims to compress the information structure for variables from multiple dimensions through linear transformation and identify the most representative direction of change in the data (Zhan et al., 2021). This method was chosen as the most effective approach to optimize the investigated subjects among these innovative techniques (Zhang et al., 2020; Zhou et al., 2023b).

PCA was used mainly in psychometrics and scale development in the social sciences but gradually expanded to complex fields such as economics, sociology, human geography, and agricultural systems science. It became an effective tool for processing multidimensional data and explaining variable structures. Compared to the traditional statistical strategy centered on a single variable, PCA emphasizes the overview between variables and retains the principal component with the largest amount of information through the dimensionality reduction process (Mardia, 1979).

Within the framework of SLFE in this study, PCA is employed to determine whether certain indicators retain their significance, given the complexity of measuring rural sustainability using multiple indicators. The indicators are diverse in their representative meaning, exhibiting strong correlations due to overlapping. PCA is thus applied to these standardized indicators to reduce dimensionality and extract uncorrelated principal components that best capture the variance structure across the dataset. PCA not only simplifies the indicator system but also ensures that the derived components reflect the most influential dimensions of rural livelihoods and ecological wellbeing. In this study, the resulting principal components are then used to construct composite scores for each capital, enabling an integrated and objective evaluation of rural household sustainability. To investigate the five capitals based on the SLF (Department for International Development, 1999) and previous research on ecological analysis (Shi and Yang, 2022), 21 indicators are used to establish the basic structure of the evaluation system needed in this research. Data obtained from the survey conducted are used for the quantitative analysis (Table 1). The information entropy weight method is used to quantify the indicators.

Table 1

Table 1. Indicators based on the SLFE standards.

In order to quantify each indicator, a well-structured questionnaire was designed considering the former investigation conducted by Chen and Cai (2025). Most indicators were derived from a structured household questionnaire that combined numeric responses and scored Likert-scale items, ensuring adequate coverage of both material and perceptual dimensions. In addition, two ecological indicators, air quality (E20) and water quality (E21), were obtained from official environmental monitoring datasets for the same period, standardized, and incorporated into the PCA analysis to represent the objective ecological conditions from government statistics published in the ecology department of Hunan Province. To arrive at a reliable index, a procedure of three phases is followed: (1) reliability and validity screening, (2) PCA for dimensional reduction, and (3) weight and score calculation.

PCA was utilized to reduce the complexity of the proposed 21 standardized livelihood and ecological indicators and derive objective weights for index construction. PCA is a dimensional reduction technique that linearly transforms correlated variables into a smaller set of uncorrelated principal components, each representing a direction of maximal variance in the data (Jolliffe, 2002). By summarizing most of the original information in a few components, PCA provides statistically grounded weights based on variance contribution for each component (Kaiser, 1960). To ensure that the 21 standardized indicators form a coherent scale, internal consistency reliability is assessed using Cronbach’s α, the most widely used coefficient quantifying the degree to which items intercorrelate and jointly reflect a single latent construct. α is computed as follows:

$\alpha =\frac{n}{n-1}\left(1-\frac{{\displaystyle {\sum }_{i=1}^n}{S}_i^2}{S_X^2}\right),(1)$

where $n$ is the number of sustainability indicators, $S_i^2$ is the sample variance of the ith indicator, and $S_X^2$ is the total sample variance of all indicator variables. Any indicator whose deletion would increase $\alpha$ to ≥0.70 was deleted (Tavakol and Dennick, 2011).

After validating the reliability of these indicators, the standardization process was carried out using SPSS, and the specific equation is as follows:

$z_{k,i}=\frac{x_{k,i}-{\overline{x}}_i}{s_i},(2)$

where $x_{k,i}$ is the original value of indicator i for the household unit k, $z_{k,i}$ is the standardized value of indicator i for the household unit k, $s_i$ is standard deviation of the indicator, and ${\overline{x}}_i$ refers to the sample mean.

Then, PCA is used to select the most important indicators that can be considered in this investigation (Pearson, 1901):

$X_{n\times p}=\left[\begin{array}{cccc}{x}_{11}& x_{12}& \cdots & x_{1p}\\ x_{21}& x_{22}& \cdots & x_{2p}\\ ⋮& ⋮& ⋮& ⋮\\ x_{n1}& x_{n2}& \cdots & x_{np}\end{array}\right],(4)$

where each row of the $X_{n\times p}$ matrix corresponds to one of $n$ survey units (households or sites) and each column corresponds to one of the $p$ standardized sustainability indicators. The procedure then proceeds as follows:

$\left(\begin{array}{c}{Y}_1\\ Y_2\\ ⋮\\ Y_m\end{array}\right)=\underset{E_{m\times p}}{\underbrace{\left(\begin{array}{cccc}{e}_{11}& e_{12}& \cdots & e_{1p}\\ e_{21}& e_{22}& \cdots & e_{2p}\\ ⋮& ⋮& \ddots & ⋮\\ e_{m1}& e_{m2}& \cdots & e_{mp}\end{array}\right)}}\left(\begin{array}{c}{x}_1\\ x_2\\ ⋮\\ x_p\end{array}\right),(5)$

where $E_{m\times p}$ is the matrix of eigenvectors of the indicators’ covariance matrix and $Y_m$ represents the vector of principal component scores.

It can be inferred that the new uncorrelated variables are the principal components (PCs). While the component scores are the result of PCA, they also contribute to the PCA criterion, with the loadings playing a similar role in this procedure (Jolliffe and Cadima, 2016). In this study, the procedure is conducted following the standard PCA computation steps to extract the most influential indicators that are needed to obtain a comprehensive perspective on sustainability in Guoliang village.

In the PCA process, each principal component is constructed as a linear combination of the original standardized variables. The first principal component (PC1), which captures the greatest variance across all indicators, is taken as a value of the construct validation result and formally expressed as follows (Jolliffe and Cadima, 2016):

${\text{PC}}_1=a_{11}{w}_1+a_{12}{w}_2+\cdots +a_{1p}{w}_p,(6)$

where $w_j$ denotes the jth standardized indicator and $a_{1j}$ represents its loading coefficient on PC1. These coefficients are derived from the eigenvector corresponding to the largest eigenvalue of the correlation matrix. More precisely, the loading of indicator $X_j$ on the principal component PC_i is calculated as follows (Jolliffe and Cadima, 2016):

$l\left({\text{PC}}_i,X_j\right)=\sqrt{{\lambda }_i}·e_{ij}\left(i=1,2,...,m;j=1,2,...,p\right),(7)$

where ${\lambda }_i$ is the eigenvalue of the ith component and $e_{ij}$ is the jth entry of its eigenvector. The loadings represent the contribution and direction of each indicator to the overall variance captured by the component.

The PCA procedure makes each capital a score system, ignoring the specific indicators, which is designed to synthesize the multi-dimensional livelihood dimensions into a unified principal component. After standardizing the raw data matrix $X_{n\times p}$, PCA transforms the correlated indicators into a set of uncorrelated principal components, with the first component $F_1$ capturing the maximal variance. For instance, the first principal component score $F_{1k}$ for unit k is expressed as follows (Jiang et al., 2018):

$F_{1k}=e_{11}{z}_{k1}+e_{12}{z}_{k2}+\cdots +e_{1p}{z}_{kp},(8)$

where $F_{1k}$ refers to the first principal component score of unit k.

To ensure comparability across multiple dimensions and eliminate scale differences, all principal component scores were normalized to a [0, 1] interval using min–max normalization. Specifically, according to Equation 8 each unit $k$, the normalized score $C_i\left(k\right)$ is calculated as follows (Jiang et al., 2018):

$C_i\left(k\right)=\frac{F_{1k}-PC{A}_{\mathit{min}}}{PC{A}_{\mathit{max}}-PC{A}_{\mathit{min}}},(9)$

where $PC{A}_{\mathit{min}}$ and $PC{A}_{\mathit{max}}$ represent the minimum and maximum values across all units, respectively. This normalization method preserves the original ranking while scaling all the scores to the same range, which facilitates further integration and comparison across capital dimensions.

Subsequently, according to Equation 9 performing score can be calculated as follows (Jiang et al., 2018):

$S\left(k\right)=\frac{1}{6}\left(\overline{C_n\left(k\right)}+\overline{C_h\left(k\right)}+\overline{C_s\left(k\right)}+\overline{C_f\left(k\right)}+\overline{C_p\left(k\right)}+\overline{C_e\left(k\right)}\right).(10)$

Before applying PCA, indicator reliability and content validity are assessed following the procedures according to Equations 1–7. As shown in Figure 3, internal consistency was evaluated using Cronbach’s α, and structural contribution was assessed based on the loading of each indicator on PC1. Among the 21 indicators derived from the SLFE, most exhibited satisfactory performance, with α values exceeding 0.7 and PC1 loadings above the 0.5 threshold. These results confirm the statistical adequacy of the selected indicators and support the integration of ecological capital into the SLF system.

Figure 3

Figure 3. Indicator reliability (α) and factor loading (PC1) analysis for SLFE items with the validation of the indicators.

However, five indicators, namely, forest ownership (N3), household income (F5), social connectedness (S10), off-farm labor (H17), and vocational skills (H18), failed to meet either the reliability or loading criteria and were, therefore, excluded from subsequent analysis. Their removal strengthens the coherence of the final indicator set while preserving its theoretical comprehensiveness.

We retain only PC1 as the composite index for each indicator. This approach not only simplifies the dimensional structure but also ensures a statistical basis without compromising the majority of the information embedded in the original indicator set. Figure 3 visualizes the loading values of all indicators on PC1, illustrating their contribution to the sustainability dimension.

Through screening of the indicators, five indicators were deleted to strengthen the statistical accuracy in constructing a valid evaluation framework, with each removal based on methodological criteria. Specifically, N3 exhibits low PC1 loading, suggesting its small contribution to the overall variance of capitals. This is consistent with the land-use characteristics of Guoliang village, where forest resources are scarce and unevenly distributed; forest ownership is largely homogeneous and, thus, statistically uninformative. F5 failed both the internal consistency and factor loading thresholds due to the high similarity in income sources across households because many of them rely on agriculture and local tourism, resulting in limited variance. Moreover, possible seasonal fluctuations and income, which is not reported, may further affect its reliability. S10 demonstrated its poor correlation with other social capital indicators, which is because of the informal and kinship nature of social networks in the village that are difficult to quantify through generalized surveys. Therefore, it is necessary to reduce its structural coherence within the social capital construct. H17 and H18 also failed the screening, implying that neither labor migration nor technical certification is strongly aligned with other human capital indicators such as education and health. This is because many residents engage in informal and simple work without formal training, leading to inconsistent patterns that undermine their validity within the composite human capital dimension. Therefore, the removal of these five indicators improves the statistical structure under PCA and enhances its contextual relevance by retaining indicators that meaningfully capture the structural variation in sustainable capacities among Guoliang households (Juracka and Valaskova, 2025; Domagalska-Grędys et al., 2025).

Application of self-fulfilling philosophy

In addition to normal investigation into the production materials of the villagers, mental interaction is also vital in pending the sustainable development of the village, normally.

In this study, the self-fulfilling philosophy concept is used to investigate individual psychology, determining whether it has a positive or negative impact on villagers’ motivation for production. Data from the qualitative questions in the questionnaire are fully analyzed using MAXQDA software. All transcripts were extracted to obtain an overall understanding of the data and facilitate analysis; a structured, multi-cycle coding procedure was then conducted in line with the guidelines proposed by Saldaña (Lungu, 2022). The core perceptual and behavioral feedback loops are summarized in Table 2, which is based on themes derived from qualitative analysis of open-ended responses and semi-structured interviews (Chen and Cai, 2025). Materials were coded in MAXQDA using a multi-cycle approach. Items listed here are synthesized manifestations rather than direct questionnaire questions.

Table 2

Table 2. Core mechanisms of the self-fulfilling prophecy identified from qualitative materials in Guoliang village.

Table 2 does not reproduce individual questionnaire items; rather, it presents synthesized qualitative findings identified through analysis of open-ended survey responses and semi-structured interviews with purposively selected households. Transcripts and textual materials were imported into MAXQDA and analyzed using multi-cycle coding following established procedures. Codes were iteratively refined through open and axial coding, with constant comparison until thematic saturation was reached. Where applicable, dual coding on a subsample was used to check coding consistency. The resulting themes summarize the beliefs, expectations, behaviors, and outcomes that characterize self-reinforcing perception loops in the study setting.

The interviewees who recognize themselves as optimistic or negative are sorted based on the SIP, which ranges from 0 to 1, as specifically defined in Table 3.

Table 3

Table 3. Classification criteria for the SIP levels based on the range values.

To examine whether villager' perceptions are shaped by their livelihood foundations, according to Equation 9, an OLS regression model is established:

${SIP}_k={\beta }_0+{\beta }_1{C}_{\mathrm{n}}\left(k\right)+{\beta }_2{C}_{\mathrm{h}}\left(k\right)+{\beta }_3{C}_{\mathrm{s}}\left(k\right)+{\beta }_4{C}_{\mathrm{f}}\left(k\right)+{\beta }_5{C}_{\mathrm{p}}\left(k\right)+{\beta }_6{C}_{\mathrm{e}}\left(k\right)+{\epsilon }_i,(11)$

where $C_{\mathrm{n}}\left(k\right)$, $C_{\mathrm{h}}\left(k\right)$, $C_{\mathrm{s}}\left(k\right)$, $C_{\mathrm{f}}\left(k\right)$, $C_{\mathrm{p}}\left(k\right)$, and $C_{\mathrm{e}}\left(k\right)$ are the first components of PCA scores for the natural, human, social, financial, physical, and ecological capital (all rescaled to 0–1); ${\beta }_i$ represents the marginal effects of each capital on psychological perception; and ${\epsilon }_i$ refers to the residual between the actual SIP score from the villagers and the predicted value of the model.

Results and discussion

This section presents the descriptive statistics of the variables and survey data used in the ELP model. The dataset includes responses from 100 households in Guoliang village, covering 21 indicators categorized into six capitals: natural, financial, physical, social, human, and ecological.

According to Equation 10, the capital score of each household is computed by projecting its standardized indicators onto the first principal component and applying min–max normalization. Figure 4 explicitly illustrates the dataset of the score obtained under PCA by ranking households according to their capital scores. The natural capital curve declines steadily, reflecting generally similar cropland quality and land tenure security across Guoliang village. In contrast, social and financial curves show sudden decreases, indicating that few households maintain strong networks and diversified income sources before those capacities collapse. Physical capital fluctuates only among the leading households and then sharply declines, revealing gaps in infrastructure access. Ecological capital starts at a low level and declines gradually, underscoring significant environmental concerns within the region. Human capital follows a mild slope, suggesting relatively uniform access to education and health resources in Guoliang village. These patterns demonstrate the role of livelihood capacities; once natural capital falls below a critical level, households experience cascading losses in social, financial, and physical assets (Das et al., 2025). This phenomenon explains why some families think differently and are likely to have a higher risk of falling back into poverty (Hu and Xu, 2025). Empirical evidence from rural Europe and South Asia further suggests that once land quality or basic resource security falls below a critical level, households are more likely to withdraw from collective activities and productive investments, accelerating multidimensional vulnerability (Bohnet et al., 2025; Saha, 2025). The pattern observed in Guoliang village, therefore, aligns with emerging international findings that rural sustainability is governed by non-linear capital coupling rather than linear accumulation.

Figure 4

Figure 4. Distribution of PCA-derived normalized capital scores across 100 households in Guoliang village in 2025.

According to Equation 11, each household’s SIP index is calculated by aggregating normalized scores from qualitative survey responses and mapping them onto a continuous [0, 1] scale. A three-dimensional scatter visualization is illustrated in Figure 5, in which the x-axis represents the comprehensive capital score of the households (ordered by natural capital), the y-axis indicates the SIP value, and the z-axis reflects the sample rank across all 100 surveyed households.

Figure 5

Figure 5. Three-dimensional visualization of livelihood capital and subjective perception ordered by natural capital.

The overall spatial distribution follows a clear pattern: households with low SIP scores cluster in the front-left corner, corresponding to minimal capital endowments across all dimensions. In contrast, those with high SIP values occupy the rear right corner, demonstrating a convergence of strong livelihood assets and high psychological optimism. Households with moderate SIP scores form a transitional diagonal band, rising from the lower-left to upper-right, indicating a graded and positive relationship between material conditions and psychological outlooks.

These results align with the logic of the SFP, suggesting that after a household reaches a certain level of comprehensive score, increases in other capital tend to enhance their optimism about future development (Balboni et al., 2022). Conversely, households falling below these thresholds face multiple disadvantages, both materially and psychologically, and may be more vulnerable to cyclical poverty risks in the future (Jin and Liu, 2025).

A few households are notably different from the general pattern. For example, a relatively high comprehensive score exhibits low SIP values, which suggests that objective assets may not fully support positive expectations if other psychosocial or contextual factors, such as prior negative experiences, intergenerational pessimism, or lack of community cohesion, undermine the belief in future improvement (Bozzato et al., 2025; Lin and Luo, 2025). Conversely, a small number of households with short capital endowments nonetheless demonstrate relatively high SIP scores, reflecting cases where strong personal agency, external support, or perceived opportunity foster optimistic outlooks despite structural limitations. These exceptions highlight the heterogeneity of psychological responses under similar material conditions and reinforce the necessity of integrating both objective and subjective dimensions into sustainability assessments of rural areas.

During the regression, the presence of multicollinearity among the six PCA-derived capital scores is assessed. As illustrated in Figure 6, the variance inflation factor (VIF) values for all variables are well below the conventional threshold of 4, with the highest value of 1.93 for financial capital and the lowest value of 1.39 for ecological capital. The average value of VIF is 1.58, indicating a low degree of collinearity among the explanatory variables. These results explain that the capital dimensions used in the regression model are sufficiently independent, thereby ensuring the statistical reliability of subsequent estimations and interpretations. While the PCA-derived capital scores are theoretically orthogonal at the indicator level, practical overlap across livelihood dimensions may still occur because the capitals are interconnected through shared socio-economic mechanisms. Reporting VIF, therefore, serves as a transparency check that the estimated regression effects are not dominated by inflated standard errors due to harmful collinearity; however, commonly used VIF cutoffs should be treated as heuristics rather than universal rules and interpreted alongside the model context and specification quality (Kalnins et al., 2025). In addition, because VIF diagnostics do not fully address other threats to inference, complementary diagnostics such as condition indices and influence analysis can further strengthen the robustness of the regression-based interpretation (Osman, 2025).

Figure 6

Figure 6. Variance inflation factors of each capital.

The regression results are displayed in Figure 7, which shows a radar visualization of the cumulative regression coefficients across six livelihood capital dimensions. The results reveal that social capital has the strongest negative impact on subjective perception, with a coefficient of −0.119, implying that deficiencies in social connectivity limit access to information, and poor institutional support significantly weakens villagers’ psychological wellbeing. Physical capital ranks second in negative influence, with a cumulative effect of −0.085, reflecting the limited contribution of infrastructure assets, such as housing and transportation equipment, to mental optimism under current conditions (Chen and Cai, 2025).

Figure 7

Figure 7. Radar visualization of the cumulative effects of ELP on subjective individual perception (SIP) based on OLS coefficients.

Financial capital also demonstrates a negative effect of −0.044, suggesting that economic assets such as deposits or income are insufficient to boost residents’ perception. It is possibly due to low variability or perceived insecurity in income sources. Human capital shows a slight negative effect of −0.026, implying that although education and health are essential, their influence on psychological expectations remains limited in the short term (Redwan and Shabur, 2025).

In contrast, ecological capital displays the strongest positive influence, with a coefficient of 0.052, underscoring the importance of environmental quality and green space in shaping confidence about the future. Natural capital shows a weakly positive effect of 0.003, indicating that basic assets have limited but foundational value in sustaining positive outlooks.

It is suggested that the influence of livelihood capitals on subjective perception and enhancing social networks and ecological environments may be more effective for boosting rural psychological sustainability than merely increasing economic or physical inputs (Cui et al., 2025; Song, 2025).

The findings confirm a structured relationship between household capital conditions and subjective development attitudes. PCA exhibits a good result in summarizing the different capital scores. The multicollinearity test ensures the independence of model inputs, providing statistical support to the analysis. These combined results demonstrate that the interaction between objective capital and personal outlook is not uniform and helps explain why households with similar conditions may still show different responses in terms of future confidence. This forms a practical basis for the following discussion on targeted policy directions.

Conclusion

This study proposes the ELP model to evaluate rural sustainability by integrating objective assessment of materials with villagers’ subjective psychological perceptions and applies it to a household survey in Guoliang village. Methodologically, PCA was used to reduce 21 indicators into six uncorrelated livelihood capital dimensions, enabling the construction of a standardized household sustainability score with low multicollinearity (average VIF = 1.58). Based on this framework, regression analysis incorporating the logic of the self-fulfilling prophecy examined how livelihood capitals shape SIP. The results indicate that stronger overall capital endowments are associated with higher optimism, while the effects of different capitals are asymmetric: ecological and natural capitals exert weakly positive influences, whereas social and human capitals show the strongest negative associations with psychological pessimism. The model explains a moderate share of variation in subjective perception (R² = 0.281), suggesting that improvements in social networks and ecological environments may be more effective for enhancing rural psychological sustainability than solely increasing economic or physical assets. Despite these insights, the study is limited by its cross-sectional design and single-village sample. Future research should extend the ELP framework to longitudinal and comparative settings to further explore dynamic feedback between livelihood conditions and subjective perception in rural sustainability transitions.

This study presents that sustainability is not only about what resources a village possesses but also about how local villagers see and react to their circumstances. Including personal perceptions fills an important gap left by many existing assessment models, which often focus heavily on tangible assets and overlook community agency. From a policy perspective, these findings offer practical ideas for rural renewal efforts in line with the Chinese policy of rural revitalization strategy. The observed patterns suggest that simply improving natural resources or financial support may not be sufficient. Effective policies should also address social ties and ecological health while encouraging positive development through comprehensive governance, transparent planning, and engaging tourism options.

In this research, the level of regional development is not taken into consideration as Guoliang village is in the provincial capital city of Changsha. Due to the siphon effect, it will receive a higher level of consumption, more people flow, more convenient transportation, and more financial support from the local government than several villages whose positions are remote. Therefore, this study has certain limitations as the evaluation system may only be applicable to Guoliang village and similar locations with favorable prerequisites. Although the ELP model offers a new perspective for assessing rural sustainability, it also has several shortcomings that indicate directions toward future research. Because the analysis focuses on Guoliang village, which is an area with advantages provided by the close distance to Changsha, strong tourism flows, and targeted policy support, the generalizability of our findings to more isolated or resource-scarce communities remains uncertain. In addition, relying solely on linear PCA might overlook some of the more non-linear relationships between different aspects of livelihood capital.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent from the patients/ participants or the patients/participants’ legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.

Author contributions

HL: Conceptualization, Formal Analysis, Validation, Writing – original draft, Writing – review and editing. JL: Investigation, Methodology, Writing – review and editing. YT: Investigation, Supervision, Writing – review and editing.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fenvs.2025.1704173/full#supplementary-material

References

Appadurai, A. (2004). “The capacity to aspire: culture and the terms of recognition,” in Culture and public action. Editors V. Rao, and M. Walton (Stanford University Press), 59–84.

Google Scholar

Balboni, C., Bandiera, O., Burgess, R., Ghatak, M., and Heil, A. (2022). Why do people stay poor? Q. J. Econ. 137 (2), 785–844. doi:10.1093/qje/qjab045

CrossRef Full Text | Google Scholar

Belton, B., and Filipski, M. (2019). Rural transformation in central Myanmar: by how much, and for whom? J. Rural Stud. 67, 166–176. doi:10.1016/j.jrurstud.2019.02.012

CrossRef Full Text | Google Scholar

Bohnet, I. C., Bryce, R., Måren, I. E., Barraclough, A. D., Malcolm, Z., Külm, S., et al. (2025). Co-creating cultural narratives for sustainable rural development: a transdisciplinary learning framework for guiding place-based social-ecological research. Curr. Opin. Environ. Sustain. 73, 101506. doi:10.1016/j.cosust.2024.101506

CrossRef Full Text | Google Scholar

Bozzato, P., Corradi, E., Crudo, M., and Pelizzoni, I. (2025). The perception of educational barriers, their sociodemographic correlates, and their relationship with future orientation in Italian adolescents. Educ. Sci. 15 (9), 1208. doi:10.3390/educsci15091208

CrossRef Full Text | Google Scholar

Chen, Q., and Cai, L. A. (2025). Self-fulfilling prophecy in livelihood sustainability: rural tourism in pingqian Village, China. Tour. Manag. 106, 104988. doi:10.1016/j.tourman.2024.104988

CrossRef Full Text | Google Scholar

Cui, H., Wang, Y., and Zheng, L. (2025). Livelihood sustainability of rural households in response to external shocks, internal stressors and geographical disadvantages: empirical evidence from rural China. Environ. Dev. Sustain. 27 (8), 18221–18250. doi:10.1007/s10668-024-04666-7

CrossRef Full Text | Google Scholar

Das, S., Sharma, K. K., Majumder, S., and Chowdhury, I. R. (2025). Evaluating sustainable agricultural livelihood security in West Bengal, India: a principal component analysis approach. Environ. Dev. Sustain. 27 (2), 4769–4816. doi:10.1007/s10668-023-04097-w

CrossRef Full Text | Google Scholar

Department for International Development (DFID) (1999). Sustainable livelihoods guidance sheets. London, UK: DFID.

Google Scholar

Domagalska-Grędys, M., Niewiadomski, M., and Piecuch, K. (2025). Aspects of support and types of work–life balance among employees from rural areas in Poland. Sustainability 17 (18), 8313. doi:10.3390/su17188313

CrossRef Full Text | Google Scholar

Gómez Valenzuela, V., and Holl, A. (2024). Growth and decline in rural Spain: an exploratory analysis. Eur. Plan. Stud. 32 (2), 430–453. doi:10.1080/09654313.2023.2179390

CrossRef Full Text | Google Scholar

Guo, Y., Mao, L., Mo, D., Shu, J., and Guo, A. (2021). Vegetation dynamics and human activities over the past 1300 years revealed by pollen record at the Tongguan kilns, lower xiangjiang river, China. Quat. Int. 577, 139–146. doi:10.1016/j.quaint.2020.09.023

CrossRef Full Text | Google Scholar

Hsu, K., and Peng, L.-P. (2023). Understanding vulnerability and sustainable livelihood factors from coastal residents in Taiwan. Underst. Vulnerability Sustainable Livelihood Factors Coastal Residents Taiwan. Mar. Policy 155, 105793. doi:10.1016/j.marpol.2023.105793

CrossRef Full Text | Google Scholar

Hu, Z., and Xu, J. (2025). How does the risk of returning to poverty emerge among poverty-alleviated populations in the post-poverty era? A livelihood space perspective. Sustainability 17 (11), 5079. doi:10.3390/su17115079

CrossRef Full Text | Google Scholar

Hu, S., Yang, Y., Zheng, H., Mi, C., Ma, T., and Shi, R. (2022). A framework for assessing sustainable agriculture and rural development: a case study of the beijing-tianjin-hebei region, China. Environ. Impact Assess. Rev. 97, 106861. doi:10.1016/j.eiar.2022.106861

CrossRef Full Text | Google Scholar

Jiang, Q., Liu, Z., Liu, W., Li, T., Cong, W., Zhang, H., et al. (2018). A principal component analysis based three-dimensional sustainability assessment model to evaluate corporate sustainable performance. J. Clean. Prod. 187, 625–637. doi:10.1016/j.jclepro.2018.03.255

CrossRef Full Text | Google Scholar

Jin, H., and Liu, H. (2025). How does livelihood capital alleviate psychological anxiety among immigrants? Front. Public Health 13, 1700964. doi:10.3389/fpubh.2025.1700964

PubMed Abstract | CrossRef Full Text | Google Scholar

Jolliffe, I. T. (2002). Principal component analysis. 2nd ed. Springer. doi:10.1007/b98835

CrossRef Full Text | Google Scholar

Jolliffe, I. T., and Cadima, J. (2016). Principal component analysis: a review and recent developments. Philosophical Trans. R. Soc. A Math. Phys. Eng. Sci. 374 (2065), 20150202. doi:10.1098/rsta.2015.0202Shi

PubMed Abstract | CrossRef Full Text | Google Scholar

Juracka, D., and Valaskova, K. (2025). Progress towards sustainable activities: principal component analysis (PCA) of SMEs in the european union. J. Int. Stud. (2071-8330) 18 (2), 9–26. doi:10.14254/2071-8330.2025/18-2/1

CrossRef Full Text | Google Scholar

Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educ. Psychol. Meas. 20 (1), 141–151. doi:10.1177/001316446002000116

CrossRef Full Text | Google Scholar

Kalnins, A., and Praitis Hill, K. (2025). Additional caution regarding rules of thumb for variance inflation factors: extending o'Brien to the context of specification error: additional caution regarding rules. Qual. and Quantity 59. doi:10.1007/s11135-024-01980-0

CrossRef Full Text | Google Scholar

Li, Z. (2020). Research on the problems and countermeasures in the transition between the old and new government accounting systems in tongguan subdistrict. School of Business Administration. [Master’s thesis]. China: China National Knowledge Infrastructure.

Google Scholar

Li, X., and Lei, L. (2023). Evaluating rural sustainable land use from a system perspective based on the ecosystem service value. Reg. Sustain. 4 (1), 96–114. doi:10.1016/j.regsus.2023.03.002

CrossRef Full Text | Google Scholar

Lin, S., and Luo, N. (2025). Widening income inequality and the decline in fertility intentions: micro evidence from Chinese households. Int. Rev. Econ. and Finance 104, 104723. doi:10.1016/j.iref.2025.104723

CrossRef Full Text | Google Scholar

Liu, D., Li, F., Qiu, M., Zhang, Y., Zhao, X., and He, J. (2024). An integrated framework for measuring sustainable rural development towards the SDGs. Land Use Policy 147, 107339. doi:10.1016/j.landusepol.2024.107339

CrossRef Full Text | Google Scholar

Long, N. (1997). An introduction to the sociology of rural development. Routledge: Taylor and Francis.

CrossRef Full Text | Google Scholar

Loras-Gimeno, D., Díaz-Lanchas, J., and Gómez-Bengoechea, G. (2025). Rural depopulation in the 21st century: a systematic review of policy assessments. Regional Sci. Policy and Pract. 17 (5), 100176. doi:10.1016/j.rspp.2025.100176

CrossRef Full Text | Google Scholar

Lungu, M. (2022). The coding manual for qualitative researchers. Am. J. Qual. Res. 6 (1), 232–237. doi:10.29333/ajqr/12085

CrossRef Full Text | Google Scholar

Maican, S. Ș., Muntean, A. C., Paştiu, C. A., Stępień, S., Polcyn, J., Dobra, I. B., et al. (2021). Motivational factors, job satisfaction, and economic performance in Romanian small farms. Sustainability 13 (11), 5832. doi:10.3390/su13115832

CrossRef Full Text | Google Scholar

Mardia, K. V. (1979). Multivariate analysis. London ; New York: Academic Press. Available online at: http://archive.org/details/multivariateanal0000mard.

Google Scholar

Merton, R. K. (1948). The self-fulfilling prophecy. Antioch Rev. 8 (2), 193–210. doi:10.2307/4609267

CrossRef Full Text | Google Scholar

Natarajan, N., Newsham, A., Rigg, J., and Suhardiman, D. (2022). A sustainable livelihoods framework for the 21st century. World Dev. 155, 105898. doi:10.1016/j.worlddev.2022.105898

CrossRef Full Text | Google Scholar

Osman, E. G. A. (2025). Entropy based solutions for detecting and treating multicollinearity in econometrics. Discov. Anal. 3 (1), 18. doi:10.1007/s44257-025-00047-0

CrossRef Full Text | Google Scholar

Pearson, K. (1901). LIII. On lines and planes of closest fit to systems of points in space. Lond. Edinb. Dublin Philosophical Mag. J. Sci. 2 (11), 559–572. doi:10.1080/14786440109462720

CrossRef Full Text | Google Scholar

Redwan, T., and Shabur, M. A. (2025). Analyzing the socio-economic factors of adolescent malnutrition in Bangladesh using best worst method. Discov. Sustain. 6 (1), 1–16. doi:10.1007/s43621-025-01121-z

CrossRef Full Text | Google Scholar

Saha, S. K. (2025). Empowering rural south Asia: off-grid solar PV, electricity accessibility, and sustainable agriculture. Appl. Energy 377, 124639. doi:10.1016/j.apenergy.2024.124639

CrossRef Full Text | Google Scholar

Shi, J., and Yang, X. (2022). Sustainable development levels and influence factors in rural China based on rural revitalization strategy. Sustainability 14 (14), 8908. doi:10.3390/su14148908

CrossRef Full Text | Google Scholar

Song, N. (2025). Social network, sustainable livelihood capital and risk response: an empirical analysis of rural China. Local Environ. 30 (6), 815–829. doi:10.1080/13549839.2024.2407605

CrossRef Full Text | Google Scholar

Sun, Y., and Jiang, W. (2022). Human behavior recognition method based on edge intelligence. Discrete Dyn. Nat. Soc. 2022 (1), 3955218. doi:10.1155/2022/3955218

CrossRef Full Text | Google Scholar

Tan, C. Y. (2017). Study on environmental restoration of gaoling community in Changsha tongguan street. Hunan Agricultural University. [Master’s thesis]. China: China National Knowledge Infrastructure.

Google Scholar

Tavakol, M., and Dennick, R. (2011). Making sense of Cronbach’s alpha. Int. J. Med. Educ. 2, 53–55. doi:10.5116/ijme.4dfb.8dfd

PubMed Abstract | CrossRef Full Text | Google Scholar

Thomas, S. J., Sahoo, S. S., Thomas, S., G, A. K., and Awad, M. M. (2025). Indian rural livelihoods and renewable energy interventions – a critical analysis for a bottom-up approach for sustainability from an energy-water-food nexus context. Energy Nexus 18, 100421. doi:10.1016/j.nexus.2025.100421

CrossRef Full Text | Google Scholar

Valizadeh, N., and Hayati, D. (2025). A systematic review on selection and comparison of holistic agricultural sustainability assessment approaches. Front. Sustain. Food Syst. 9, 1559503. doi:10.3389/fsufs.2025.1559503

CrossRef Full Text | Google Scholar

Van Amsterdam, W. A. C., Van Geloven, N., Krijthe, J. H., Ranganath, R., and Cinà, G. (2025). When accurate prediction models yield harmful self-fulfilling prophecies. Patterns 6 (4), 101229. doi:10.1016/j.patter.2025.101229

PubMed Abstract | CrossRef Full Text | Google Scholar

Wang, J., Xiao, H., and Liu, X. (2022). The impact of social capital on multidimensional poverty of rural households in China. Int. J. Environ. Res. Public Health 20 (1), 217. doi:10.3390/ijerph20010217

PubMed Abstract | CrossRef Full Text | Google Scholar

Xu, X., Du, Z., Chen, X., and Cai, C. (2019). Confidence consensus-based model for large-scale group decision making: a novel approach to managing non-cooperative behaviors. Inf. Sci. 477, 410–427. doi:10.1016/j.ins.2018.10.058

CrossRef Full Text | Google Scholar

Yu, M., Chen, Y., Yi, Z., Wang, Q., and Zhang, Z. (2022). Benefits of market information and professional advice in a vertical agricultural supply chain: the role of government provision. Int. J. Prod. Res. 60 (11), 3461–3475. doi:10.1080/00207543.2021.1924409

CrossRef Full Text | Google Scholar

Zhan, M., Liang, H., Zhu, C., and Dong, Y. (2021). Opinions and actions dynamics under bounded confidence. Int. J. Inf. Technol. and Decis. Mak. 20 (01), 321–340. doi:10.1142/S0219622021500012

CrossRef Full Text | Google Scholar

Zhang, H., Cai, G., and Yang, D. (2020). The impact of oil price shocks on clean energy stocks: fresh evidence from multi-scale perspective. Energy 196, 117099. doi:10.1016/j.energy.2020.117099

CrossRef Full Text | Google Scholar

Zhang, D., Lin, Q., and Mao, S. (2024). Addressing rural decline: china’s practices in rural transformation and farmers’ income growth. Agriculture 14 (9), 1654. doi:10.3390/agriculture14091654

CrossRef Full Text | Google Scholar

Zhou, X., Hu, Y., Wu, J., Liang, W., Ma, J., and Jin, Q. (2023a). Distribution bias aware collaborative generative adversarial network for imbalanced deep learning in industrial IoT. IEEE Trans. Industrial Inf. 19 (1), 570–580. doi:10.1109/TII.2022.3170149

CrossRef Full Text | Google Scholar

Zhou, X., Liang, W., Yan, K., Li, W., Wang, K. I.-K., Ma, J., et al. (2023b). Edge-enabled two-stage scheduling based on deep reinforcement learning for internet of everything. IEEE Internet Things J. 10 (4), 3295–3304. doi:10.1109/JIOT.2022.3179231

CrossRef Full Text | Google Scholar

Zhu, G., Cai, C., Pan, B., and Wang, P. (2021). A multi-agent linguistic-style Large Group decision-making method considering public expectations. Int. J. Comput. Intell. Syst. 14 (1), 188. doi:10.1007/s44196-021-00037-6

CrossRef Full Text | Google Scholar

Zhu, J., Yuan, X., Yuan, X., Liu, S., Guan, B., Sun, J., et al. (2021). Evaluating the sustainability of rural complex ecosystems during the development of traditional farming villages into tourism destinations: a diachronic emergy approach. J. Rural Stud. 86, 473–484. doi:10.1016/j.jrurstud.2021.07.010

CrossRef Full Text | Google Scholar