A Dynamic Analysis of the Asymmetric Effects of the Vocational Education and Training on Economic Growth, Evidence From China.

Since 2010, China's economic growth has stagnated due to an unbalanced regional industrial structure and lack of sufficient qualified technical personnel. A nonlinear autoregressive distributed lag (NARDL) model has been used in this study to examine the asymmetric effects of secondary vocational education and training (SVET) and higher vocational education and training (HVET) and their interaction with high-tech industries on economic growth over the period 1980-2020. The findings show that an increase in secondary vocational education and training (SVET) significantly boosts long-term economic growth, while a decrease in secondary vocational education and training (SVET) insignificantly reduces long-term China economic growth. Likewise, the upward change in higher vocational education and training (HVET) promotes and the downward fluctuation in higher vocational education and training (HVET) significantly reduces China's long-term economic growth. The moderating role of secondary vocational education in the impact of high-tech industries on China's economic growth is positive, but not significant. However, higher vocational education plays a significant positive moderating role in high technology industries impact on economic growth. Strategically, the study analysis suggests that economic transition prosperity can be achieved by encouraging higher vocational education and the equal development of high-tech industries in all regions. In addition, this study also proposes to cultivate high-quality talents related to high-tech development and modern industrial innovation and upgrading through higher vocational education, improve productivity, and promote the country's intensive development.

Introduction

China's economy has entered a new phase after a period of economic deterioration and a declining labor force participation rate. Judging from the gross national income (GNI) data, China entered a lower-middle-income economy in 2001 and a middle-high-income country in 2010, but since 2010, China's economic growth has been stagnant (World Bank, 2021). The recent recessionary discussions due to the “high-middle-income trap” have been very prominent and diverse (Bulman et al., 2017). Thus, there is a “trap,” and cracking the mystery of the economic recession black box has become an economic problem that needs to be addressed urgently. Contemporary Chinese economic practice shows that middle-income economies urgently need to discover the driving force and mechanism of development beyond the low-income stage. Thus, in order to ensure higher sustained economic and social development, economic transformation has become a significant notion (Erumban et al., 2019). Disparities exist across the country in this regard. Southeast coastal provinces such as Jiangsu, Zhejiang, Fujian, and Guangdong are important manufacturing hubs, contribute to China, with the opening of international trade and the rapid growth of national income. Per capita incomes in coastal areas and other cities are growing more rapidly than in other provinces as economic activity slumps into these dynamic regions. However, areas beyond the coastal are old industrial and energy bases in China's northeastern provinces. The manufacturing industries established in these areas are relatively backward, the economic development is facing major social structural problems, and the energy development has entered a period of recession (Chen et al., 2021). Many of such regions are even located in hilly or desert regions, these low-income provinces are naturally inland, adjacent to low-income peripheral economies, which are not conducive to manufacturing activities and urbanization. By contrast, coastal cities, including seaports, are close to wealthier neighbors, favoring urbanization and infrastructure development, and thus more manufacturing and trade opportunities (Poonpatpibul and Foo, 2019). Once regional disparities start to widen, more different paths emerge, with the rumbling of coastal areas, the establishment of special economic zones within these provinces, and further development. An increasing number of trained and unskilled workforces and businesses will flow to these regions in search of opportunities, and with it the stagnation of the overall economic structure. Hence, in order to achieve economic transformation, it is necessary to narrow the gap between regional industries and technologies, make full use of modern technology to strengthen traditional industries, develop high-tech industries on an equal footing in all regions, and optimize the mode of economic growth. However, in order to develop a balanced regional industrial structure and continue to innovate and create, the lack of a sufficient number of competent technical personnel is the primary constraint. Cultivate high-quality talents related to high-tech development and modern industrial innovation and upgrading, improve productivity, and promote intensive development. Later, this higher productivity enhances the country's financial and economic resources, which can be reinvested in less developed regions to narrow regional development gaps and thereby achieve the goal of economic transformation. Follow General Secretary Xi Jinping's instructions to cultivate high-quality technical endowments with an advanced contemporary vocational education system, aiming at the forefront of technological change and industrial upgrading, and keeping pace with economic transformation or growth (Xinhua News Agency, 2021). The vocational education system for high-level technical majors focuses on national modernization and high-quality development (Ren and Jiang, 2021). Undoubtedly, in the structural reform of the economic supply side and the long-term development of politics, economy and culture, higher vocational education and training (VET) and its reform should not be underestimated (Yuan and Wang, 2021). China with the continuous transformation of economic development mode and the imbalance of regional industrial structure, the state should develop a modern vocational education and training system linked to lifelong higher vocational education and training, and cultivate high-quality workers and comprehensive skilled talents that meet the needs of industrial upgrading and enterprise development (Ren et al., 2021b). China current population size and vocational education and training (VET) scale are relatively large, the management system is imperfect, the management efficiency of schools and colleges is low, the development of training programs is not ideal, and vocational education institutions lack cooperation with the industry. Considering the huge economic and social regional differences, China needs to simultaneously establish a balanced high-level vocational education and training (VET) system in all regions, and on the basis of regional equality, give full play to the potential of talents and coordinate and cooperate with upgraded industries open up development paths, and ultimately achieve the goal of economic transformation (Zhao and Liu, 2019). According to the World Bank (2020), the proportion of students enrolled in China—Secondary and Higher Vocational Education and Training is highlighted in Figure 1 below, reflecting that the enrolment rate in secondary vocational education and training (SVET) is lower than the higher vocational education and training (HVET) during the specified period. The latest enrollment rates for secondary vocational education and training (SVET) and higher vocational education and training (HVET) in China are 18.7 and 40%, respectively. In addition, enrolment rates at both vocational education and training levels are on a downward trend.

Figure 1

China's secondary and higher vocational education and training enrollment rate. Source: World Bank (2021).

The main determination of this study is to empirically detect the asymmetric effects of secondary and higher vocational education and training on China's growth, using non-linear autoregressive distributed lag (NARDL) model, ignoring the ARDL model commonly used in other studies. Besides, the study also estimates the impact of high-tech industry in the link between secondary vocational education and China economic growth, and also impact of high-tech industry in the link between higher vocational education and China's economic growth using annual secondary data from 1980 to 2020.

Literature Review

Schultz's (1960) innovative research on educational investment and return has gone through two stages. The first stage is related to the growth effect of quality education supported by Solmon (1985), Doménech-Betoret (2018), Shu et al. (2020), and Chen et al. (2022). The second stage is linked to the growth effect of the scale of education, which is prioritized (Romer, 1986; Lucas, 1988; Fassbender and Leyendecker, 2018; Navarro-Carrillo et al., 2020). However, the connection between higher education and economic growth has been explored extensively (Fassbender and Leyendecker, 2018; Li et al., 2020; Zhang, 2020) and the growth effect of different level of vocational education analyzed by Loyalka et al. (2016), Wang and Sun (2017), Asadullah and Ullah (2018), Chinedu et al. (2018), and Zhibin and Weiping (2017).

Mankiw et al. (1992) revealed that the development or formation of human and physical capital opens the door to possible economic growth. Several theoretical and empirical studies have shown that the development of human capital in vocational education and training (VET) can contribute to growth significantly. The importance of vocational education and training (VET) for human capital development has grown exponentially as technological dynamism stimulates the skill set required in the workforce. With support from UNESCO and the World Bank, vocational education and training (VET) can be used as a tool to alleviate poverty and maintain economic prosperity, in addition, VET has the potential to improve the livelihoods, productivity, and employability of young working age population in developing economies (Loyalka et al., 2016; World Bank, 2021). Lee et al. (2020) selected 92 countries and used a regression panel model to reveal the impact of lower and higher vocational education on growth over the period 2000–2016. The results of the analysis show that the impact of high school vocational education on growth is gradually significant, while vocational education in junior high school has a weaker gradual influence on growth. In addition, this study further examines the impact of junior high school and high school vocational education on the respective economic output indicators below and above the median. The results show that when the economic output indicator is lower than the median group, the economic growth effect of high school vocational education is stronger than that of the entire sample. Similarly, another study by Pan et al. (2020) used multiple explanatory designs to explore the development of vocational education and training in Singapore and Hong Kong as two Asian metropolises. The findings suggest that the combination of characteristics of freedom, development and collective skill formation has resulted in a hybrid construction of vocational education and training systems in both metropolises, increasing productivity and developing a high-skilled society.

Regarding the impact of industrial structure in the link between vocational education and economic growth, most scholars focus on qualitative research on the interaction between development of vocational education, industrial structure adjustment and growth (Meng et al., 2015; Li et al., 2016; Liu et al., 2017; Li and Wang, 2018). Su (2017) argues that vocational education, in terms of scale, level and training, can adjust the pattern of technical talents required for industrial structural upgrading, possibly stimulating higher productivity and growth. Similarly, Shi and Xia (2016) found a significant progressive link between human capital in vocational education and China economic growth, through the provision of skilled labor to upgrading industries.

A study by Carruthers and Jepsen (2020) identified three themes in the rapidly expanding literature on the social and economic outcomes of VET, first vocational education will develop a flatter worker-age employment profile than academic educators, secondly those with secondary vocational education may have earlier careers and better achievements, third vocational post-secondary education is proportional to improvements in labor market outcomes compared with no or incomplete post-secondary education. Wiemann and Fuchs (2018) enlighten that vocational education is considered one of Germany's growth-stimulating export initiatives. The findings suggest that VET's revolutionary development of Mexico's skill-forming system by offering firms with a medium-capacity workforce will make Mexico a knowledgeable producer. The contribution of this study is to summarize the innovation gap in manufacturing skills related to rehearsal, thereby showing that vocational education is under-considered in economic geography and regional studies. Chinedu et al. (2018), carefully reviewed the extensive literature on ESD and anticipates a merger of the literature to determine the mutual information pursued by SD and ESD models. The findings highlight the elite and important role that vocational and technical teachers can play in accelerating the transition to sustainable growth. The study then analyzed various ways to conclude that VET can help promote social wellbeing and community development if properly incorporated into VET teacher training programs. Shu et al. (2020) established research on creativity and innovation from a vocational education and training (VET) perspective. The study proposes that designing creative and innovative curriculum based on the skills industry needs for different levels of vocational education and training (VET) may lead to sustainable economic and social development. Nilsson (2010) highlighted two main reasons for the resurgence of vocational education and training (VET). First, VET is seen as an appropriate means of fostering economic growth, and second, it is seen as a potential impact tool for social increase. The study analysis illuminates that appropriate integration of sustainable development with VET teacher training programs can contribute to social and community wellbeing, as well as economic development. Likewise, Kareem et al. (2015) concluded that vocational education and training (VET) plays an important role in developing human capital by generating a skilled workforce, improving quality of life and promoting industrial growth. VET contributes significantly to Nigeria's national industrialization and economic growth through appropriate human development, in full accordance with the aspirations of the industry and the country. Another study by Asadullah and Ullah (2018) examined the growth impact of the vocational education and training (VET) spending in 31 OECD country groups using 15 years of secondary data. Statistical findings from the study support the positive impact on growth of strengthening vocational education and training (VET) through social inclusion. Zhibin and Weiping (2017) also studied the synergistic revolution of industrial growth and higher vocational education. The results show that, based on theory and case studies, the cooperation of industrial growth and higher vocational education must emphasize the formation of a common vision of different industries and educational subjects, establish the common curiosity foundation of school-enterprise micro-entities, and carry out parameter sorting. The research also focuses on the collaborative framework of vocational education and industry, and designs a long-term process of multi-school-enterprise discipline collaboration, value integration, benefit assimilation, resource mixing, and collaborative execution to achieve higher growth and development. Similarly, another study by Ren et al. (2021a) analyzed the conjugate link between industrial economy and vocational education. The results show that, based on the conjugation theory, the conjugation of vocational education and industrial economy is generally good, but not structurally enough.

In a nutshell, the literature reviewed so far clearly shows that there is very little empirical research on this topic, so this study aims to empirically explore the growth effects of vocational education and training (VET) at secondary and higher education levels in China. This study also aims to reveal the influence of high-tech industry in the link between secondary vocational education and China economic growth, and also impact of high-tech industry in the link between higher vocational education and China's economic growth. Short and long-term parameter estimates for the above relationships can be analyzed using non-linear autoregressive distributed lag (NARDL).

Model Specifications, Data Sources, and Estimation Methods

Model Specification

The human capital accumulation of a well-trained and educated workforce capable of taking on its protagonist role in the production line. Countries promote economic growth by having more competent human resources than foreign direct investment. Most empirical studies clearly show that higher enrolment in skills development institutions is associated with higher productivity (Cao and Jariyapan, 2012). Romer (1990) claims that human capital is closely connected to productivity due to the ability of countries to adopt and absorb new technologies, implying that a knowledgeable workforce is an important determinant of constructing, implementing and utilizing new knowledge, resulting in productivity growth. More importantly, Mankiw et al. (1992) pointed out that growth determinants are closely related to economic growth theory, especially their augmented Neoclassical (Solow) model, which divides capital into physical and human capital. As assumed by Mankiw et al. (1992) that the total production (Yi,t) in country i and time t is determined by three pillars: labor (Li,t), capital (Ki,t), and human capital (Hi,t), while linking growth theory to empirical growth regressions. In particular, they follow constant productivity to extend the standard Cobb-Douglas production function:

where Yi,t is the total economic output, (Ai,t) represents technical efficiency, multiplied by labor (L), which means the labor efficiency unit, 0e denotes total years of education and ni represents health. Both labor supply and technical change have exogenous growth:

Instead, the growth rates of physical and human capital, denoted by Ht,i and Kt,i, depend entirely on the country-specific savings rates represented by sKi,t and sHi,t after deducting their respective depreciation rates.

The depreciation rate (α) is expected to remain constant and not fluctuate by country or capital type. However, it can be seen that the savings rate varies by country or capital type, but is fixed over time.

Following the Mankiw et al. (1992) growth empirical model described above, we derive the following three econometric models.

GDP is used as an alternative for economic growth and represents gross domestic product, SVET indicates secondary vocational education and training, HVET denotes higher vocational education and training, LF stands for employed labor force, CF shows gross fixed capital formation, GSE is used for government spending on education, HTI represents high-tech industry. HTIt and HTIt, respectively, illustrate the interaction between secondary vocational education and training and high-tech industry and the interaction between higher vocational education and training and high-tech industry.

Data Sources

Variable data sources, measurements and their definitions are explicitly mentioned in Appendix A below. This study used annual secondary data through the date range 1980–2020.

Estimation Method

This study used a relatively new asymmetric or (NARDL) techniques developed by Shin et al. (2014) for detecting long-term and short-term asymmetries between variables. As suggested by Katrakilidis and Trachanas (2012), NARDL models outperform traditional ARDL techniques when examining small samples for cointegration. This approach has been adopted by various studies to explore whether a surge or contraction of the regressor has different effects on regressand (Bahmani-Oskooee and Ghodsi, 2017; Fassbender and Leyendecker, 2018).

Considering asymmetric cointegration regression, while following Shin et al. (2014),

where α+ and α− are the long-term optimistic and pessimistic parameters, and yt is a k × 1 regressor vector decomposed as

By substituting the partial positive and negative sums of equations (8) and (9), the main model [i.e., equation (5)] is decomposed into asymmetric equation (10).

In Equation (10), the motion of SVET and HVET is decomposed into their increasing and decreasing parts Sum, namely SVET = SVET+ + SVET−, HVET = HVET+ + HVET−, where the sign indicates the increase and decrease of SVET and HVET, respectively. Generate partial sums of individual SVET and HVET increase and decrease changes using formulas [Equations (11–14)].

The various steps involved in the asymmetric ARDL cointegration method are shown in Equation (14).

Initially, after evaluating the null hypothesis Ho = κ+ = κ− and its alternative hypothesis H1 = κ+ ≠ κ−, this study uses the Wald test to examine long-term non-linear effects. Rejecting the null hypothesis describes asymmetric or non-linear effects of SVET and HVET on growth. Stimuli and adverse long-term effects are represented by Ω+ and Ω−, while = or υ+i = υ−I, respectively, short-term asymmetric effects of positive and negative changes in SVET and HVET.

The asymmetric dynamic multiplier effect with a unit change in , and expressed as follows: κ

Note: m → ∞, mp+ → β+ and mp− → β−, where β+ and β− as described above, signify long-term asymmetry coefficients, respectively. This study uses Brown's et al. (1975) recursive cumulative sum of residuals (CUSUM) and recursive cumulative sum of squares of residuals (CUSUMSQ) to evaluate model stability.

Results and Discussion

Kwiatkowski-Phillips-Schmidt-Shin (KPSS) (Kwiatkowski et al., 1992) and Augmented Dickey-Fuller (ADF) (Dickey and Fuller, 1981) are two different unit root tests used in empirical analysis to test the integration order for each model variable. Consideration of unit root tests describing the variable stationarity level is a necessary step in verifying cointegration among variables. Such as, if any variable found integrated at order I (2), the ARDL technique cannot be used (Dickey and Fuller, 1981; Kwiatkowski et al., 1992). Table 1 below highlights the results of the KPSS and ADF tests, confirming that none of the variables are integrated in the second order; therefore, we can proceed with ARDL's approach and conduct further analysis. The findings in Table 1 highlight that entire variables are integrated as hybrids, such as, GDP, LF, HVET, HTI are integrated at level I(1) and CF, GSE, and SVET are integrated at level I(0).

Table 1

Results of unit root test.

	ADF		KPSS
Variables	At level	First difference	At level	First difference
GDP	−2.435	−3.921***	0.324	0.524***
SVET	−1.213***	−2.145***	0.832*	0.921***
HVET	−3.921	−4.214***	0.216	0.421***
LF	0.213	1.923***	0.313	0.415***
CF	−2.14***	−3.921***	0.241*	0.234***
GSE	2.345***	4.312***	0.523*	0.713***
HTI	3.214	3.910***	0.632	0.813***

represent the significant levels of 10, 5, and 1%, respectively.

Next, the highlighted F-statistics are 4.872, 5.898, and 5.901 obtained from the results of the Bound test method shown in Table 2, exceeding the upper critical value at the 1% level of significance. Hence, the results illustrate that the variables have stable long-term equilibrium relationships.

Table 2

Results of bound testing approach of cointegration.

	Model 1	Model 2	Model 3
F-statistics	4.872***	5.898***	5.901***
Lower-upper bound (1%)	3.41–4.59	3.18–4.36	3.16–4.48
Lower-upper bound (5%)	2.64–3.89	2.48–3.66	2.46–3.76
Lower-upper bound (10%)	2.35–3.47	2.44–3.45	2.06–3.45 K
K	6	5	5

indicates statistical significance at 1%. K represents the number of regressors in the three specified models.

The long-term elasticities of the NARDL model is shown in Table 3. Asymmetric effects can be detected from the sum of the positive and negative parts of SVETt−1 and HVETt−1, namely , and and (Bahmani-Oskooee and Ghodsi, 2017). Both the positive and negative components of SVETt−1 and HVETt−1 appear to be highly significant in our analysis, but there are conflicting signs that expansion and contraction of SVETt−1 and HVETt−1 have different effects on economic growth. The increases in the SVETt−1 () and HVETt−1 () coefficients indicate that both SVETt−1 and HVETt−1 have a significant long-term positive effect on the growth of the specified main model. These findings are very consistent with the new growth theories that the impact of secondary and higher vocational education and training in China on growth is gradual and significant (Shi and Xia, 2016; Su, 2017; Zhibin and Weiping, 2017; Carruthers and Jepsen, 2020). The results of the analysis clearly show that for every 1% expansion of SVETt−1, economic growth increases significantly by 0.309%. A negative coefficient for SVETt−1 () indicates that a decrease in SVETt−1 leads to a but insignificant decline in economic growth. The long-run negative lag one coefficient of SVETt−1 () is −0.215, which means that a contraction of SVETt−1 shrinks economic growth by 0.215%, but it is not statistically significant. These results of the asymmetric effects of SVET clearly show that economic growth parallels the expansion or contraction of SVET. Likewise, the positive lag-one coefficient of higher vocational education and training () is 0.325, reflecting that the expansion of higher vocational education has significantly boosted economic growth by 0.325 percentage points. The negative lag-one coefficient of higher vocational education () is −0.261, indicating that the reduction of higher vocational education () significantly contracted economic growth by 0.261 percentage points. Hence, empirical analysis clearly proves that higher vocational education (HVETt−1) plays a vital and decisive role in stimulating China's growth by cultivating high-quality human resources and later with employing them in productive sectors. This finding is consistent with research by Vu et al. (2012), Zhibin and Weiping (2017), and Wang et al. (2021). The positive significant coefficients of high-tech industry (HTIt−1) and government education expenditure (GSEt−1) are 0.173 and 0.132, respectively, indicating that for every 1 percentage point increase in high-tech industry and government education expenditure, economic growth will expand significantly by 0.173 and 0.132 percentage points, respectively. Similarly, the coefficients of capital formation (CF) and employed labor force (LF) are 0.251 and 0.334, both positive and significant, indicating that every 1% increase in LF and CF significantly expands economic growth by 0.251 and 0.334 percentage points, respectively.

Table 3

Long-Term asymmetric ARDL model elasticities.

Variables	Model 1	Model 2	Model 3
InSVETt-1+	0.309** (0.014)	0.532** (0.021)	0.213** (0.021)
InSVETt-1-	−0.215 (−0.221)	–	–
InHVETt-1+	0.325*** (0.005)	0.621*** (0.002)	0.419*** (0.007)
InHVETt-1-	−0.261*** (0.002)	–	–
InGSEt-1	0.132*** (0.005)	0.514** (0.033)	0.058*** (0.005)
InLFt-1	0.334*** (0.007)	0.464*** (0.005)	–
InCFt-1	0.251*** (0.004)	0.813** (0.031)	0.261** (0.027)
InHTIt-1	0.173*** (0.002)	0.251** (0.021)	0.132* (0.061)
InSVETt-1*HTIt−1	–	0.071(0.525)	–
InHVETt-1*HTIt−1	–	–	0.314*** (0.009)
Constant	−13.612** (0.021)	−8.271*** (0.005)	−19.216** (0.031)
R2	0.97	0.88	0.98
Adj R2	0.73	0.64	0.63
F-statistic	952.62	963.63	936.26

represent the significance levels of 1, 5, and 10%, respectively, and the numbers in parentheses are the probabilities of each coefficient.

The interaction coefficient between secondary vocational education and high-tech industry (SVETt-1*HTIt-1) is 0.071, which is positive and insignificant, indicating that expansion in the interaction term (SVETt-1*HTIt-1) by one percentage point boost economic growth insignificantly by 0.071 percentage. This result clearly reflects the mismatch between human development in secondary vocational education and high-tech industries in promoting high-level economic growth in China. High-tech industries require people to master modern high-level knowledge and skills to expand product excellence and capabilities. Similarly, the interaction coefficient of higher vocational education and high-tech industry (HVETt-1*HTIt-1) is 0.314, which is also positive and significant, reflecting that the interaction variable (HVETt-1*HTIt-1) improved by 1 percentage point, the economic growth increased significantly by 0.314 percentage points. It can be understood that the development of human capital in higher vocational education can well meet the requirements of industrial progress and enterprise development, thereby making a significant contribution to China economic growth.

The short-run elasticities of parameters in the analysis are reported in Table 4, and has been observed that the coefficient magnitude of the progressive and adverse partial sum of SVET moves in the same direction as economic growth. For example, the rise of secondary vocational education () has a significant gradual influence on growth. However, the decline of secondary vocational education () has insignificant stimulating influence on growth. Similarly, the magnitude coefficient of the sum of the positive and negative parts of HVETt-1 also fluctuates in direct proportion to economic growth, such as the rise in higher vocational education (HVETt−1) and the decline in higher vocational education () both significantly contributed to the short-term economic growth. The interaction between secondary vocational education and high-tech industry (SVETt-1*HTIt-1) insignificantly contribute to growth in the short term. However, the interaction of higher vocational education with the high-tech industry has significantly boosted growth in the short term. The error correction mechanism (ECTt-1), represented by the velocity adjustment coefficient, tends to balance after short-term shocks due to negative and significant statistics. The ECTt−1 coefficients are −0.69, −0.87, and −0.73, respectively, indicating that short-term imbalances can be adjusted to long-term balances in the range of 69–87%.

Table 4

Short-run elasticities of asymmetric NARDL models.

Variables	Model 1	Model 2	Model 3
InSVETt-1+	0.215** (0.021)	0.312** (0.013)	0.215** (0.021)
InSVETt-1-	−0.315 (−0.721)	–	–
InHVETt-1+	0.362*** (0.001)	0.321*** (0.007)	0.621*** (0.002)
InHVETt-1-	0.215*** (0.002)	–	–
InGSEt-1	0.181* (0.052)	0.314* (0.051)	0.031* (0.071)
InLFt-1	0.031*** (0.002)	0.171*** (0.001)	0.214** (0.021)
InCFt-1	0.243*** (0.001)	0.314** (0.031)	0.219** (0.021)
InHTIt-1	0.125*** (0.006)	0.616* (0.082)	0.215* (0.071)
InSVETt-1*HTIt−1	–	0.0503 (0.524)	–
InHVETt-1*HTIt−1	–	–	0.139*** (0.011)
ECTt−1	−0.695** (0.001)	−0.875*** (0.001)	−0.735*** (0.009)

represent the significance levels of 1, 5, and 10%, respectively, and the numbers in parentheses are the probabilities of each coefficient.

Diagnostic tests such as RESET, RAMSEY, ARCH, LM, and JB have been used to check for heteroskedasticity and autocorrelation. As clearly highlighted in Table 5, the selected variables in the model are free of autocorrelation and heteroscedasticity. The Breusch-Godfrey LM test clearly demonstrated the non-significance of the F statistic, indicating that the variables included in each model had no serial correlations.

Table 5

Diagnostic test results.

Test	Model 1	Model 2	Model 3
RAMSEY	1.721 (−0.732)	1.821 (−0.321)	0.382 (−0.981)
JB	4.216 (−0.327)	7.842 (−0.744)	12.865 (−0.643)
ARCH	1.432(−0.643)	1.325 (−0.436)	7.764 (−0.765)
RESET	3.658 (−0.763)	7.642 (−0.867)	6.986 (−0.653)
LM	1.246 (−0.765)	0.763 (0.357)	1.875 (0.432)

Inside in the parenthesis are the probability values.

Model stability has been tested using CUSUM and CUSUMSQ highlighted in Appendicies B–D, which validates the stability of all estimated models as the graphs lie within the critical boundaries at the 5% significance level.

Finally, we examine dynamic multipliers to establish the final growth-order dynamics, while accommodating the context of initial imbalances and short-term dynamics due to unparalleled shocks to secondary vocational education (SVET) and higher vocational education (HVET). The existence of this initial equilibrium is supported by the rejection of the null hypothesis in Appendix E, so the investigation of these two figures provides insight into the asymmetric validity highlighted above in Table 3. It is clear that growth has responded gradually and significantly to the rise in HVET outflows. Regarding Appendix E, positive HVET shocks and negative SVET shocks are prominent and most overbearing. However, the short-term dynamics are characterized by shocks to both types of growth, with the exception of SVET where negative shocks stimulate growth, but in all cases the imbalances are corrected after about 6 years.

Concluding Remarks

This study explores the dynamic relationship between secondary vocational education (SVET), higher secondary vocational education (HVET), the effect of high-tech industry in the link between secondary vocational education and high-tech industry interaction (SVET*HTI), higher vocational education and high-tech industry interaction (HVET*HTI) and economic growth, using annual secondary data for China from 1980 to 2020. This study uses a non-linear ARDL method to examine the asymmetric effects of retention in secondary vocational education (SVET) and higher vocational education (HVET) on growth, and to reveal the impact of high-tech industry in the link between secondary vocational education and China's economic growth. Also this study examined the influence of high-tech industry in the link between higher vocational education and china economic growth. As asserted based on theoretical and empirical backgrounds, high-quality human resources generated by secondary and higher vocational education and training can significantly contribute to economic growth by capitalizing them into productive sectors or upgrading industries.

This study empirically proposed an asymmetric link between SVET, HVET, and growth, whereas the existing literature mainly focused on symmetric links for selected variables. The analysis shows that an increase in SVET significantly boosts long-term economic growth, while a decrease in SVET reduces long-term economic growth, but not significantly in China. Likewise, the upward change in HVET promotes and the downward fluctuation in HVET significantly reduces China's long-term growth. The analysis shows that, compared with secondary vocational education, higher vocational education plays a pivotal role in China's economic transformation. The contribution of secondary vocational education and high-tech industry interaction (SVET*HTI) to China's economic growth is not significant. However, the interaction between higher vocational education and high-tech industries (HVET*HTI) has significantly boosted China's economic growth. It is understandable that the development of human capital in higher vocational education can well meet the requirements of industrial upgrading and enterprise development, thus making a significant contribution to China economic transformation. Other control variables in the models, such as capital formation, employed labor, and government spending on education also contributed significantly to China's economic growth.

In terms of policy implications, the analysis of this study suggests that economic transition prosperity and associated productivity gains, in turn, can be achieved by encouraging higher vocational education and the equal development of high-tech industries in all regions. In addition, this study also proposes to cultivate high-quality talents related to high-tech development and modern industrial innovation and upgrading through higher vocational education, improve productivity, and promote the intensive development of the country.

In terms of study limitations, since this is an analysis of individual country, we suggest that in order to obtain broader conclusions, future studies can apply the NARDL method to countries with a higher sample.

Data Availability Statement

Data supporting the findings of this study are publicly available by visiting the website https://databank.worldbank.org/source/world-development-indicators.

Author Contributions

XL, AA, and QD designed and conducted the whole research, AA did write up and analysis, HW and XW both did editing and scientific revision, all authors contributed to the article and approved the submitted version.

Funding

This research was supported by the Collaborative Education Project of Industry University Cooperation of the Ministry of Education of China 2021 (Project Nos. 202102171008 and 202101201002) and the Practice Teaching Special Reform Project of Jiaxing University in 2020 (Project No. SJZY20072307-014).

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

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.