Rural revitalization is a global challenge, particularly pertinent in developing nations like China. China's approach prioritizes poverty eradication and socialist principles, employing various measures including financial subsidies and strategic plans since 2004. While progress has been made in farmers' income and rural environment, rural revitalization still largely relies on external factors like urban capital and technology. This study focuses on the endogenous driving force of higher vocational education (HVE), exploring its impact on rural revitalization. HVE, specializing in occupational fields, provides technical skills and knowledge, offering potential endogenous drive across education, culture, and economy. Unlike existing studies focusing on theoretical perspectives, this research empirically examines the mechanism of HVE's contribution to rural revitalization in China, considering both the scale and quality of HVE and employing spatial econometric models to capture spatial spillover effects and a panel threshold regression model to analyze non-linear relationships.

Existing literature identifies three key factors influencing rural revitalization: scale, technology, and structure. Scale effects involve the quantity of factors like population, capital, and resources, often with dual impacts (e.g., population growth can stimulate innovation but also lead to resource depletion). Technology effects center on the quality of technology, innovation, and knowledge, generally exhibiting positive impacts (e.g., technological upgrades enhance productivity). Structural effects focus on the organization of factors like industry, ecology, and governance. Previous research, while acknowledging HVE's potential role in rural revitalization, lacks comprehensive empirical analysis, spatial effect examination, and consideration of non-linear relationships. This study aims to address these gaps by empirically analyzing the relationship between HVE and rural revitalization.

This study uses panel data from 30 Chinese provinces (excluding Tibet, Hong Kong, Macau, and Taiwan) from 2007 to 2020. A double fixed-effects model is employed as the benchmark model to control for individual and time-fixed effects after a Hausman test confirmed its superiority over the random effects model. Rural revitalization is measured using a composite indicator based on several sub-indicators (e.g., per capita output value, crop damage area, rural latrine penetration rate, rural residents' expenditure on education) weighted using the entropy weight method. Higher vocational education is assessed through two dimensions: scale (number of graduates, students, and schools) and quality (per capita education expenditure, operating expenses, and number of full-time teachers), also weighted using the entropy weight method. Control variables include per capita GDP, rural fiscal expenditure, rural fixed assets, and transportation infrastructure (highway mileage). To analyze spatial effects, Moran's I test confirmed spatial autocorrelation, leading to the use of a spatial Durbin model (SDM) with double fixed effects. LeSage and Pace's (2009) spatial effect decomposition method was employed to disentangle direct, indirect, and total effects. Finally, a panel threshold regression model was applied to investigate non-linear relationships between HVE and rural revitalization, using per capita GDP and the urban-rural income gap as threshold variables.

The analysis reveals that both the scale and quality of higher vocational education significantly and positively impact rural revitalization at the national level (using double fixed effects model). However, regional heterogeneity exists: in the eastern and central regions, the scale of higher vocational education significantly contributes to rural revitalization; in the western region, the quality of higher vocational education plays the more significant role. Spatial analysis using the SDM shows a significant positive spatial spillover effect of the scale of higher vocational education, implying that its expansion in one region benefits neighboring regions. In contrast, the quality of higher vocational education only demonstrates a significant positive direct effect on the local rural revitalization level. The threshold regression analysis indicates that the impact of higher vocational education on rural revitalization is non-linear. Using per capita GDP as a threshold variable, an accelerating threshold effect is observed for both scale and quality: as the economic level increases, the positive impact of HVE on rural revitalization strengthens. However, using the urban-rural income gap as a threshold variable, the scale effect shows a decelerating threshold effect: while a widening income gap lessens the positive effect of the scale of HVE, the quality effect demonstrates an accelerating effect, highlighting its increasing importance in bridging the income gap. Robustness checks using alternative weighting methods (PCA), lagged dependent variables, and different spatial weight matrices confirmed the consistency and reliability of the findings.

The findings demonstrate the crucial role of higher vocational education in rural revitalization in China, emphasizing the importance of considering both scale and quality, and acknowledging regional heterogeneity. The positive spatial spillover effect of HVE scale suggests the importance of regional cooperation and knowledge sharing. The nonlinear relationship highlights the need for context-specific policy interventions. The accelerating effect of economic development suggests that investing in HVE is more effective in economically advanced regions, while the urban-rural income gap's influence underlines the need for targeted policies to enhance HVE quality in regions with high inequality, mitigating the negative impact of rural-urban migration.

This study provides compelling empirical evidence of the significant contribution of higher vocational education to rural revitalization in China. The findings highlight the importance of a nuanced approach, considering both scale and quality, regional differences, spatial spillovers, and non-linear relationships. Future research could investigate causal mechanisms more rigorously using instrumental variables or difference-in-differences methods, analyze data at finer geographic levels (county or village), and explore the long-term impacts of HVE on rural sustainability.

This study is limited by the use of provincial-level data, potentially obscuring within-province variations. The chosen indicators for rural revitalization and HVE quality might not fully capture the complexity of these concepts. Further research using more granular data and a broader range of indicators is needed to enhance the generalizability of the findings. The study primarily focuses on the correlation between HVE and rural revitalization, not causal effects; more rigorous methods are needed to establish causality.