The Relationship Between the Human Capital and Economic Growth: A Case of Vietnam

Human capital are not only the engine of economic growth but also increase the global competitiveness for countries. Improving labor quality helps countries improve long-term economic efficiency. This article uses annual data in the period 1990 - 2017 in Vietnam, which attempts to explore the relationship between total capital formation, the labor quantity, education levels and life expectancy with economic growth. By using OLS regression, the analysis results shown that the gross capital formation, the labor quantity, education levels and life expectancy are positive and have a significant impact on GDP in Vietnam. Furthermore, the Granger causality test indicates that there is a two-way causal relationship between labor and economic growth (GDP) in Vietnam.


Introduction
Labor is the growth motivation for the economic development of countries in the world, including Vietnam. In the era of knowledge economy, the world economy moved to the stage of industrial revolution 4.0, the role of the labor force became increasingly important. Human capital have a direct relationship with labor productivity, thereby contributing positively to economic growth (Becker, 1994;Schultz, 1961;Mincer, 1958;Solow, 1956;Romer, 1987 andBarro, 1991). Modern growth theory has demonstrated the factors contributing to economic development not only in terms of material but also human capital and arguing that the main driving force of economic growth is the accumulation of human capital and the main difference in living standards among countries is the difference in education and skill level (Amjad, 2005). Therefore, improving the quality of labor is an urgent requirement, requiring governments to make significant reforms in all key areas of the economy, to increase productivity, skills and quality of labor to maintain competitiveness.
There are many studies in the world that have focused on the role of human resources in explaining the level and change in production and growth. The results have shown that the long-term growth and sustainable development of countries is driven to a great extent by productivity growth (Easterly and Levine, 2001). There is also growing evidence that the education and skills of the workforce are important determinants for economic growth and productivity. Romer (1990) and Lucas (1988) in their endogenous growth models have played a central role in education in the course of economic growth. Renelt and Levine (1992) found that education seems to have a high positive impact on economic growth. Therefore, in the last century, the focus of researchers is still the impact of human capital on economic growth by increasing educational and health facilities. Several empirical studies have noted a strong and positive relationship between human capital (education and health) and economic growth (Akram et al., 2008, Kakar et al., 2011. Education and health are two important aspects in improving the quality of human capital (Becker, 1964;Schultz, 1961). High quality labor increases labor productivity. Improving productivity when workers have high skills and knowledge, along with their physical and mental health can perform their tasks with higher efficiency (Bong, 2009). Workers with higher education levels can also adapt to new technology faster than workers with low education levels.
In Vietnam, there are also a number of studies on the above topic. However, most are qualitative studies that explore factors affecting employment in Vietnam (Dang 2002) or use econometric models to test the elasticity of employment growth (Pham H.M & Nguyen V.N, 2014). The observed labor market conditions in Vietnam pose an appropriate question: does the quality of human capital affect labor productivity and economic growth in Vietnam? This study has the main objective of estimating the contribution of human capital quality to economic growth in Vietnam. We consider labor with education levels and life expectancy representing the quality of human resources. While most of the previous studies based on Vietnam, the aspect of education is considered a measure of the quality of human capital.
The following section provides an overview of the quality and productivity of human capital. Part 3 presents research methods based on model specifications related to productivity with capital, number of employees and labor quality, and discusses data and methodological approaches. use. Part 4 analyzes the results, while part 5 draws conclusions and some policy recommendations. This study is unique because of the measure of human capital quality taking into account both education and health components and it is also based on Vietnam's updated data set. health on labor productivity (Umoru and Yaqub, 2013). In macro-level analysis, both educational and health variables are often included as representative of the quality of human capital. The variables used to represent education are the average number of years of schooling, education level, enrollment rate, government spending on education and literacy. Health variables are measured by life expectancy, government spending on health and adult survival. Several studies combining other variables can enhance human capital such as capital market improvement, foreign policy and trade policy (Lee and Barro, 1998;Sacerdoti et al., 1998).
Studies using company data (micro level) to examine the impact of human capital quality on labor productivity based on a single national case made using the company data or industry data. Jajri and Ismail (2007) investigated the effects of educational attainment of human capital on productivity and labor productivity of Malaysian companies based on the Cobb-Douglas function. Data were collected from 574 Malaysian companies surveyed in 2001 and 2002. They analyzed the effects of education (average school year) on labor productivity. Their findings suggest that education has a significant positive impact on labor productivity in only a few sectors. Secondary education has contributed positively to labor productivity only in the textile industry. They also found that in metal products, electricity and the electronics and food industries, the development of labor productivity is marginal due to the major contribution from the development of capital intensive production. Their research also found that in the service industry, variables such as the average school year and workers with primary, secondary and tertiary education are statistically significant in explaining labor productivity. In another sectorial study, Afrooz et al. (2010) estimated the effect of human capital on labor productivity in Iran's food industry based on the Cobb Douglas production function. The authors used panel data of 22 food production companies between 1995 and 2006. By fixed effects method, workers have skill and qualified was found to have a significant impact on labor productivity. The coefficients indicate that when the percentage of workers with education and skilled labor increases by 1%, the value added per worker in the Iranian food industry will increase by 0.14 and 0.41%, respectively (Afrooz et al., 2010;Qu andCai, 2011, Fleisher et al., 2011).
Some empirical results from studies in this area show unconvincing relationships between human capital and economic growth. While some studies show positive relationships, other studies conclude the opposite. Among the empirical results shows the negative relationship between human capital and economic growth, including studies by Sacerdoti et al. (1998), Knowles and Owen (1997). There are also studies showing that the unstable relationship between these two variables indicates a positive relationship in the early stages of development but negative relationships in the later stages (Iyigun and Owen, 1996).

Methodology and data
In order to estimate human capital effects on labor productivity, we employ a Cobb-Douglas production function in this study. This functional form is flexible and results obtained can be interpreted in a straightforward manner. The functional form also has commonly been employed in many previous studies such as Afrooz et al. (2010), Jajri and Ismail (2010) and Bloom, Canning and Sevilla (2003). A simple Cobb-Douglas production function can be expressed as: where refers to the output, is physical capital stock, is quantity of labor assumed to be homogeneous, + = 1 for constant return to scale assumption, is the efficiency parameter and is time trend. Lucas (1988) however, argues that labor is different based on his accumulated human capital. A production function that takes into account the quality of labor, therefore, can be written as: where is time allocated for producing output, (1 -) denotes time allocated for human capital investment, ℎ is human capital stock.
The term ℎ = * , constitutes effective labor. Production function based on effective labor can thus be written as: (3) In order to analyze how accumulated human capital is related to the production function, effective labor, L * refers to the labor with three levels of education and healthy mental and physical conditions, or simply expressed as: where is the proportion of labor with different ith level of education (i = 1, 2 and 3), where 1 = primary, 2 = secondary, and 3 = tertiary level at time and is the proportion of labor with good health status at time period. By substituting (4) into (2), we obtain: ln GDP t L t = α 0 +α 1 ln K t L t +α 2 lnL t +α 3 lnPE t +α 4 lnSE t +α 5 lnTE t +α 6 lnLE t +ε t (5) where / is Gross Domestic Product (GDP) per worker; / is gross capital formation per worker; is This article uses OLS multivariable regression to determine the effect of independent variables on dependent variables. The choice of the OLS method gives the least squares the least squares and has some advantages such as zero deviation, consistency, minimal variance and minimum efficiency; It is widely used based on BLUE (Best, Linear, Unbias, Estimate) rules, simple and straightforward (Gujarati 2004). The Stata econometric software 14.0 was used for this analysis. Statistical testing of parametric estimators was conducted using standard errors, t-test, F-test, R, and R 2 . Economic criteria show that the coefficients of the variable are consistent with predicted economic expectations, while the statistical criteria test is used to assess the magnitude of the overall regression. This study using annual data for the period 1990 -2017. The data were obtained from World Development Indicators published by the World Bank for Vietnam.

Ordinary Least Square Regression
The key idea of the Ordinary Least Square regression is that employing this model in order to estimate the coefficients and intercept through minimizing the sum of squared estimate errors in the multiple regression models. In the estimated regression line above, the value of (the constant term) is 0.9624027, which means that holding the value of all other variables used in this regression constant, the value of GDP will be about 0.9624027. The regression coefficient of GCF/L in the estimated regression line is 0.3735296 which implies that which shows that 1% rise in GCF/L would result in 0.3735% increase in GDP of Vietnam. The calculated t-statistics for the parameter estimates of foreign direct investment is 11.03 which is greater than the value of the tabulated t-statistics illustrates that the relationship between GDP and GCF/L is positive and statistically significant for the period under review.
Additionally, the regression coefficient of L in the estimate regression lines is 4.153785, which means that a 1% rise in GFCF would result 4.153785% increase in GDP within the period under study was accounted for by changes in labor. The calculated t-statistics for L is 4.35 which is greater than the value of the tabulated t-statistics indicates that the relationship between GDP and labor is positive and statistically significant.
In the estimated regression line above, the regression coefficient of SE is 2.203762 which implies that a 1% rise in SE may result 2.203762% of the increase in GDP within the period under study was accounted for by the SE. The calculated t-statistics for SE is 13.25 which is greater than the value of the tabulated t-statistics implies that the relationship between GDP and SE is positive and statistically significant. The regression coefficient of PE is 0.9599227 which implies that a 1% rise in PE may result 0.9599227% of the increase in GDP within the period under study was accounted for by the SE. The calculated t-statistics for PE is 7.17 which is greater than the value of the tabulated t-statistics implies that the relationship between GDP and PE is positive and statistically significant.
Similar, The regression coefficient of TE is -0.1335156 which implies that a 1% rise in TE may result -0.1335156% of the decrease in GDP within the period under study was accounted for by the TE. The calculated tstatistics for TE is -4.18 which is smaller than the value of the tabulated t-statistics implies that the relationship between GDP and TE is negative and statistically significant. The quality of highly qualified workers is based on a number of "pillars" such as civil servants, scientific and technological officials, university lecturers, high-level businessmen, technical workers .... still has not been able to fulfill the mission of "the pull of development".
On the contrary, the LE has a negative and statistically significant influence on the economic growth. Particularly, a 1% increase in the rate of LE will lead to around -19.30229 decreases in GDP. Life expectancy has Industrial Engineering Letters www.iiste.org ISSN 2224-6096 (Paper) ISSN 2225-0581 (online) DOI: 10.7176/IEL Vol.9, No.5, 2019 an immediate impact on economic growth. However, life expectancy has the opposite effect on GDP growth, indicating that if the economic development is not commensurate with the amount of labor, the advanced life expectancy will also be a pressure on the economy when the dependency ratio also from that increase.

Unit Root Test
Since most of the economic time series data are unstable, the prerequisite of conducting regression approach is to ensure that the objective time series data is stabilized; otherwise, the obtained regression results would be susceptible. ADF test and PP test are used in order to test non-stationary and stationary for all variables, which are gross domestic product per labor (GDP/L), gross capital formation per labor (K/L), PE, SE, TE and LE, and to examine the variables stationary at I(0) or I(1). The results given in Table 2 show the results with intercept and trend, and no lag for each of the four variables included in this study. The test is based on the null hypothesis that the variable contains a unit root, and the alternative is that the variable was generated by a stationary process. If the calculated test statistics are less than the critical value of the test statistics, then the null hypothesis will be rejected. The unit root tests using intercept and trend suggests that all series are non-stationary in level and becomes stationary after differencing. Thus the variables becomes integrated of order one, I(1).

The Granger Causality Test
The Granger causality test is conducted to check the existence of causality between explanatory variables and dependent variable. This model is in line with Engle and Granger (1987), Khan (2007) and Egbo (2010).
The Granger causality test was used to explore the existence of a bi-directional causality between GDP and labor for Vietnam for the proposed study period. If labor can help to forecast GDP, then we can say that labor Granger-causes GDP. However, if labor causes GDP and not versa vice, then we say there is unidirectional causality exists from labor and GDP. The Granger approach answers the question whether GDP causes labor by finding how much of the current value of GDP can be explained by past values of GDP and values of labor. Thus, to test for causality between GDP and FDI, we shall estimate the following regression equations: Where GDPt and lnLt are stationary time series sequences, and are the respective intercepts, and are white noise error terms, and k is the maximum lag length used in each time series (decided by Akaike Information Criterion (AIC) or Bayesian Information Criterion (BIC)). Labor is said to Granger cause GDP if the coefficients are jointly significantly different from zero. Similarly, GDP is said to Granger cause Lt if the coefficients are jointly significantly different from zero. According to our results we reject the null hypothesis and accept the alternative hypothesis is that GDP can cause L, GCF, PE, TE and LE. In the case of L, we reject the null hypothesis that means L can cause GDP, GCF, SE, TE and LE. In contrast, GDP and L do not have a causal relationship with SE and PE, respectively.

Conclusion and Policy implications
This paper has attempted to explore a relationship between gross capital formation, labor quantity, labor with different level of education and life expectancy with economic growth (GDP). It has employed annual data over the period of 1990 -2017. By using OLS regression in terms of level form of series variables, the result of the analysis shows that gross capital formation, labor quantity, labor with different level of education and life expectancy positively and significantly impact on GDP in Vietnam for the period under review. Besides that, the test result shown that all variable in this paper has a unit root problem in terms of level by using Augmented Dickey Fuller (ADF) test. But, when the first difference is considered, all the series become stationary at 5 percent confidence levels. Furthermore, Granger causality testing indicates that there is a two-way causal relationship between the amount of labor and economic growth (GDP) in Vietnam. Current research results show the fact that Granger labor is caused by GDP because it can reject the hypothesis at 5% significance level and vice versa. Based on the results of empirical research, we conclude that the labor quality, education and heath contributed to speed up the GDP growth into the Viet Nam economy for the period under consideration.
Improving the quality of high-quality labor in Vietnam must become the most important factor in competition and development. How to have a highly qualified workforce sufficient in size, reasonable in structure and improved in quality; how do they become "development tractors" and to connect training with use.

Policy implications:
-Improving the quality of high-qualified labor in terms of scale, rational structure and quality improvement; Create an environment and position for high-level workers to work so that they become "development tractors" and to connect training with use.
-Innovating education and training in the direction of standardization and modernization to meet the needs of the labor market and link training with enterprises; -Life expectancy is a factor representing the quality of life as well as the health care system of the society, so it is necessary to implement well the pension regime, the health care system needs to be further enhanced. in order to bring good health to the people to work in the most optimal way for the country. Promote jobs for people after retirement but still need to contribute to society to help reduce the burden of social insurance fund.