Does higher vocational education matter for rural revitalization? Evidence from China

The study examines whether and how higher vocational education (HVE) contributes to China’s rural revitalization, a national strategy aimed at industrial prosperity, ecological livability, cultural vitality, effective governance, and affluent living. Despite progress in rural income and living conditions, rural development remains reliant on external inputs and faces an endogenous drive deficit. Given HVE’s role in cultivating technical and skilled talent aligned with regional industry and social needs, the authors ask if the scale (quantity) and quality of HVE promote rural revitalization, whether effects vary spatially across regions, whether there are spillovers across provinces, and whether relationships are nonlinear depending on development conditions. The purpose is to provide empirical evidence to inform education and rural policy in China and similar contexts.

Prior research identifies multiple determinants of rural revitalization grouped into scale effects (population, economic growth, urbanization), technology effects (technological upgrading, R&D, digital finance), and structure effects (industrial and spatial structure, transportation infrastructure). While these factors can support rural development, few studies directly analyze HVE’s role. Theoretically, HVE can foster industrial upgrading, ecological protection, cultural and governance capacity, and higher incomes through skills formation. Empirically, micro data suggest vocational education raises income and status; macro studies estimate positive contributions but seldom clarify mechanisms, spatial spillovers, nonlinearities, or jointly consider HVE scale and quality. Gaps include limited empirical evidence, lack of spatial analysis despite mobile human capital, neglect of nonlinear threshold relationships by development level or inequality, and insufficient attention to HVE quality alongside scale.

Data: Balanced panel of 30 mainland Chinese provinces (excluding Tibet, Hong Kong, Macau, Taiwan) from 2007–2020. Sources include national and provincial statistical yearbooks covering economy, rural statistics, civil affairs, population and employment, and education expenditure. Measures: Dependent variable is a composite provincial rural revitalization index (2007–2020) constructed via an entropy-weighted system reflecting five dimensions (industrial prosperity, ecological livability, rural civilization, effective governance, affluent life), with component indicators (e.g., agricultural output per capita, mechanization, health and sanitation metrics, education/culture spending, governance attributes, rural Engel coefficient, per capita disposable income). Log transformations applied to stabilize variance. Key independent variables: HVE scale and HVE quality indices. Scale measured by number of HVE graduates, enrolled students, and institutions. Quality measured by inputs: full-time teachers, per-student education funding, and operating education expenditures for HVE institutions. Weights derived via entropy method; indices log-transformed. Controls: Economic development (real per capita GDP, deflated to 2000), transportation infrastructure (highway mileage, log), rural fixed asset investment share, and rural fiscal expenditure share for agriculture/forestry/water (rm). Descriptive statistics reported. Models: (1) Benchmark double fixed effects panel model (province and year fixed effects) chosen via Hausman test after comparing province-fixed, time-fixed, and two-way FE. (2) Spatial Durbin model (SDM) to account for spatial dependence, using a geographic inverse-distance spatial weight matrix (robustness checks with squared-distance and economic-geographic matrices). Global Moran’s I tests confirm spatial autocorrelation. Effects decomposed into direct, indirect (spillover), and total effects following LeSage and Pace (2009). (3) Panel threshold regression (Hansen 1999) to test nonlinear (threshold) effects using economic development (rgdp) and urban–rural income gap (ti) as threshold variables, allowing for single or multiple thresholds. Robustness checks include alternative weighting (PCA) for indices and lagged dependent variable specifications.

• National average effects: In two-way fixed effects regressions, both HVE scale and HVE quality significantly and positively associate with rural revitalization. Illustratively, the benchmark models report ln(scale) coefficient ≈ 0.066 (t ≈ 3.09) and ln(quality) ≈ 0.090 (t ≈ 2.97). • Regional heterogeneity: In eastern and central regions, the scale of HVE significantly promotes rural revitalization, while HVE quality is not significant. In the western region, HVE scale is not significant, but HVE quality shows a significant positive association. • Spatial effects (SDM): The HVE scale exhibits significant positive direct, indirect (spillover), and total effects on rural revitalization (e.g., direct ≈ 0.081, indirect ≈ 0.231, total ≈ 0.312). HVE quality shows a significant positive direct effect (≈ 0.064) but no significant indirect effect, indicating local benefits without detectable spillovers to neighbors. • Nonlinear threshold effects: Using rgdp (economic development) as threshold, there is an accelerating effect: for HVE scale, the coefficient rises from ≈ 0.202 to ≈ 0.370 when rgdp exceeds 5.799; for HVE quality, coefficients increase as rgdp moves past 2.554 and 5.799 (e.g., from ≈ 0.115 up to ≈ 0.302). Using the urban–rural income gap (ti) as threshold, HVE scale shows a decelerating effect (coefficient drops from ≈ 0.298 to ≈ 0.113 when ti > 0.043), while HVE quality shifts from insignificant to significantly positive (≈ 0.088) when ti > 0.034. • Robustness: Results remain consistent when (a) replacing entropy weights with PCA for index construction, (b) introducing first- and second-order lags of the dependent variable, and (c) substituting alternative spatial weight matrices (squared-distance, economic-geographic).

Findings demonstrate that both the expansion of HVE (scale) and improvements in its quality contribute to higher levels of rural revitalization, supporting the hypothesis that vocationally oriented human capital fosters economic, social, and governance outcomes in rural China. The differentiated regional patterns reflect varying development stages and labor market demands: in more developed eastern and central regions, expanding HVE scale aligns quickly with industrial demand and labor absorption, while in the less developed west, quality improvements that better match emerging industry requirements matter more than sheer expansion. Spatial analyses reveal that enlarging HVE scale benefits neighboring regions through talent mobility, demonstration effects, and technology diffusion, whereas quality improvements primarily generate local gains, likely requiring complementary regional conditions to diffuse. Threshold analyses show that stronger economic development amplifies the contributions of both scale and quality, while widening urban–rural gaps dampen the returns to scale but heighten the importance of quality, highlighting the need to tailor HVE strategies to local development and inequality contexts. Collectively, the results inform policies on aligning HVE investments with regional economic structures, leveraging spillovers via spatial coordination, and emphasizing quality where inequality constraints limit scale returns.

Using provincial panel data (2007–2020) and two-way fixed effects, spatial Durbin, and panel threshold models, the study shows: (1) HVE scale and quality are positively associated with rural revitalization; (2) effects differ by region—scale is key in the east/center, quality in the west; (3) scale exhibits positive spatial spillovers, while quality’s benefits are local; (4) economic development strengthens both scale and quality effects, whereas a wider urban–rural income gap reduces the payoff to scale but elevates the importance of quality. Policy suggestions include: balancing expansion with quality assurance and supervision; tailoring HVE programs to regional needs and groups; optimizing the spatial layout and coordination of HVE to enhance spillovers; and adjusting threshold conditions and incentive mechanisms (e.g., flexible admissions, targeted financial aid, performance incentives) to improve HVE’s effectiveness for rural revitalization. Future research should use finer-grained and longer time-span data, apply causal identification strategies (e.g., instrumental variables, difference-in-differences), and incorporate additional factors that may mediate or moderate HVE–revitalization linkages, including cross-country comparisons.

The analysis relies on provincial-level data from 2007–2020, limiting temporal and spatial granularity and potentially obscuring local heterogeneity. The study focuses on associations and does not establish causality due to the lack of suitable instruments or exogenous policy shocks. Variable selection follows prior literature but cannot capture all relevant determinants of rural revitalization, leaving room for omitted factors and further model enrichment.