Resources
CAPK Contribution 2022-03-08
[CAPK Contribution #5] Analysis of the Effect of Education on Economic Growth in South Korea and ASEAN Countries (1990-2019)
*Disclaimer: The opinions and views expressed in this article are solely the authors’ own and do not reflect the opinions or views of the ASEAN-Korea Centre or Council of ASEAN Professors in Korea.
*이 글에 표현된 의견 및 견해는 저자 개인의 의견으로 한-아세안센터 또는 주한 아세안 교수협의회의 공식의견과 무관함을 밝힙니다.
Analysis of the Effect of Education on Economic Growth
in South Korea and ASEAN countries (1990-2019)
Theingi Aung Khin
Abstract
The purpose of this research is to compare how education impacts economic growth in South Korea and ASEAN countries: Brunei, Indonesia, Laos, Malaysia, Cambodia, Myanmar, Philippines, Thailand, Singapore and Vietnam from 1990 to 2019. Data was collected from the World Bank and United Nations Development Programme and analyzed by using VAR (vector autoregressive) and ARDL (autoregressive distributed lag) techniques. This research is tested by determining the stationarity that may occur in the spurious regression output by using the unit root test and used heteroscedasticity test to determine if the variance of the residual was constant, unbiased and had no outliers for the time series and panel data for South Korea and ASEAN countries. The result of this paper is that education index has a positive effect on economic growth in ASEAN countries but a negative effect in South Korea for only short run period.
Keywords: South Korea, ASEAN countries, Human capital, economic growth, VAR and ARDL technique
1.Introduction
Human capital is one of the most important indicators for economic growth of a nation. While the education?growth relationship is well-known, several current findings have focused on higher education and attempted to study its effect on economic growth. This could be because higher education is regarded as one of the most important aspects in a country's economic growth and competitiveness. Many observers have emphasized the crucial importance of human capital, particularly as attained through education, to economic progress (Lucas,1988 and Mankiw et al., 1990).
Kim (2001) stated that South Korea is frequently mentioned as an example of economic development achievement. Korea's social infrastructure, including school facilities, was destroyed in the Korean War just four decades ago, however Korea has achieved nearly one hundred percent coverage for basic and secondary education in under three decades. Furthermore, Korea currently boasts a tertiary education system that rivals that of affluent nations and Korea’s education has been an active cause of economic and social progress (Kim, 2001).
Afzal et al. (2010) pointed out that as education contributes to increased social returns, it leads to higher production. Education's importance in a country's development cannot be emphasized enough, as worker productivity is dependent on it, and a person's educational options and fulfilment affect household income and economic growth over time. Furthermore, boosting education spending can result in global economic benefits, according to Tarabini (2010).
With globalization and ASEAN cooperation, determining the role of education in achieving economic progress is crucial. However, enhancing education as a human capital investment is viewed as a critical basis for achieving a significant degree of economic growth as well as boosting education as a human capital investment is viewed as a basic cornerstone in obtaining substantial growth of the economy and can also invite and attract foreign direct investment (Afzal et al., 2010).
This study focuses on human capital as a determinant of economic growth. Although education and health are all included as part of human capital, this study focuses only on the education aspect, in particular by looking at the quantity of education ? measured by an expected year of schooling (of children), and the average of mean years of schooling (of adults).
This paper will therefore look at how education impacts economic growth in South Korea and 10 ASEAN countries: Brunei, Indonesia, Laos, Malaysia, Cambodia, Myanmar, Philippines, Thailand, Singapore and Vietnam, from 1990 to 2019. To achieve this aim, VAR and ARDL techniques were used.
The structure of this study is described as follows. Section one is the introduction, section two provides a review of the literature, and section three describes the methodology. Section four shows the results and section five provides conclusions and policy recommendations.
2. Literature Review
The relationship between human capital and economic growth has been studied by many researchers all over the world (Liao et al., 2019). Using many diverse methods to study the link between human capital and GDP, researchers are studying not only within one nation but also across different regions. Most of the studies point out that human capital has a positive and significant effect on economic growth. Human capital is an important element in the economic growth process (Breton, 2012).
2.1. South Korea’ s Human Capital and Economic Growth
Various studies on Korea’s economic success can be seen in the progress in human resources, investment ratios and higher savings, progresses in rule of law and larger trade openness, as important causes for economic growth (Lee, 2016).
Han and Lee (2019) mentioned that Korea has amassed an unprecedented quantity of educated workers, owing to the significant public investment in the education field and high household demands for higher education. Korea has improved the competitiveness of its sectors thanks to a plentiful supply of well-educated personnel, changing the economy into one of the world's highest exporters.
Korea is known for its fast advancements in the quality of education among its people. The number of workers aged 15 and over who have completed at least some secondary schooling has increased from 37% in 1970 to 87 % in 2010. Over the same time span, the percentage of people with a college education has risen from 6% to 42% (Barro and & Lee, 2013).
According to analysis by Han and Lee (2019), the influence of human capital to GDP growth will continue to be positive and significant. Korea can endure substantial growth in human capital over the following two decades owing to the continuous rise of educational attainment. According to the trend of South Korea’s per capita GDP and education index as shown in Figure 1 and 2, the per capita GDP decreased in 1997 and 2009 because of the Asian financial crisis and global financial crisis. Except these two periods, all of the trends have increased continuously. As shown in Figure 3, linear prediction of per capita GDP and education index have a positive relationship. If education index is higher, per capita GDP will be higher.
Figure 1
South Korea’s per capita GDP
Note. Calculation for each variable was based on a real data source using Stata software15
Figure 2
South Korea’s Education Index
Note. Calculation for each variable was based on a real data source using Stata software15
Figure 3
Linear prediction of per capita GDP and Education Index
Note. Calculation for each variable was based on a real data source using Stata software15
2.2. ASEAN countries’ Human Capital and Economic Growth
Most countries have highlighted the significance of education in succeeding economic progress, particularly in this internationally competitive economic situation, and notably with ASEAN cooperation. This emphasizes the significance of education in economic progress. If countries devote their efforts to education, human capital will improve, resulting in increased productivity (Manlagnit, 2011).
Acemoglu (2012) explained that the function of organizations in economic growth is a new topic of inquiry within the theory of economic growth. Organizations are established in order to coordinate economic activities and make existing technologies more affordable. Acemoglu (2012) thought that organizational work is more effective than individual. ASEAN is an attractive institution among the other organizations. This work not only highlights the function of organizations, but it also adds to notions about uniformity as a growth engine and a growth barrier.
According to the trend of ASEAN countries’ per capita GDP and education index as shown in Figure 4 and 5, all of the trends are increasing continuously. In a linear relationship with per capita GDP and education index in ASEAN countries as shown in Figure 6, the relation is positive. If education index is higher, per capita GDP will be higher.
Figure 4
ASEAN countries’ per capita GDP
Note. Calculation for each variable was based on a real data source using Stata software15
Figure 5
ASEAN countries’ Education Index
Note. Calculation for each variable was based on a real data source using Stata software15
Figure 6
Linear prediction of per capita GDP and Education Index
Note. Calculation for each variable was based on a real data source using Stata software15
2.3.GDP per capita (constant 2010 US dollars)
Some research has found that economic growth has an impact on education, and that education, in turn, has an influence on economic growth. With growth of the economy, the government will be able to invest more and expand education spending to meet the population's demand for human capital.
Gross domestic product divided by midyear population is called GDP per capita. GDP is the totality of gross value added by all local producers in the economy plus any product taxes and minus any supports excluded in the value of the products. It is considered without making deductions for depletion and degradation of natural resources.
2.4. Human Capital (Education index)
Human capital is measured as an education index in expected years of schooling (of children) and an average of mean years of schooling (of adults), both expressed as an index attained by scaling with the corresponding highest. These calculations were founded on expected and mean years of schooling from the UNESCO Institute for Statistics (2020) and other sources. This variable is taken from the United Nations Development Programme.
Barro and Lee (2013) analyzed the link between education and income using the latest schooling data and confirmed that worker education has a large positive impact on the level of income in a country.
Human capital is one of the important factors for a country, particularly for developing countries. This investing also contributes significantly to development by transferring assets, improving management, and transferring technologies in order to boost a country's economic growth (Liao et al., 2019).
2.5. Foreign direct investment (FDI) (% of GDP)
FDI is defined as the net inflow of funds used to obtain a long-term management stake (10% or more of voting shares) in a company that operates in a country other than the investor's. This data is segmented by GDP and presents net inflow from foreign investors in reporting economy.
Borensztein and De Gregorio (1995) claimed that foreign direct investment is an essential mode of knowledge transfer and contributes more to economic growth than local investment. Iamsiraroj (2016) observed that the overall impact of FDI is positively connected with economic growth and vice versa, and pointed out that economic freedom, labor force, and trade openness are the most important determinants of FDI to a country. FDI can attract trade openness and, as a result, an open economy.
2.6. Capital formation on a gross basis (% of GDP)
Expenditures on accompaniments to the economy's fixed assets create gross capital formation (previously gross domestic investment) by World Bank. Land improvements (fences, waterways, drainage system, and so on); factory, equipment, and equipment purchases; and road construction, railway lines, and other similar structures, such as schools, offices, hospitals, private residential dwellings, and commercial and industrial buildings are all examples of fixed assets. Businesses keep stock of items to accommodate unanticipated fluctuations in production or sales. Net acquisitions of assets are considered formation of capital, according to the 1993 System of National Account(SNA).
Feldstein (1994) performed a cross-sectional examination of OECD nations and discovered a significant negative association between domestic and foreign direct investments. According to this study, a one dollar increase in an overseas subsidiary's assets results in a 0.2~ 0.4 dollar decrease in capital stock in the United States. However, in a time series analysis of aggregate data consisting of foreign and domestic investments made by US multinational enterprises as a whole, Desai, Foley, and Hines (2005a) discovered a positive relationship between foreign and domestic investments and suggested that greater foreign investment was related to the increase of domestic investment.
2.7. Trade (% of GDP)
The sum of commodities and services exported and imported as a percentage of GDP is known as trade, as defined by the World Bank. According to the findings of Blomstrom et al. (2000), a considerable amount of FDI alone is insufficient to promote economic growth and prosperity in a host country. As a result, we have included both FDI and trade in the production function to examine their impact on economic growth. FDI and trade are frequently viewed as significant accelerators for economic progress in developing countries. FDI is an essential means of transferring technology from industrialized to poor countries. Frankel and Romer (1999) considered international trade to be a tool for economic progress.
As the rate of economic growth plays a vital role in economic development, it has become very important to investigate its pattern and responses to macroeconomic conversion in the country. A fall in economic growth could delay investments in productive sectors and also affect the economic stability of the nation. The main task for each government, no matter developed or developing, is to develop the economy of a country and improve the lives of the people.
3.Methodology
3.1. Types of Data, Variables
The dataset used in this paper has been taken from the World Bank Indicators and United Nations Development Programme. The period used in this paper is limited to 30 years from 1990 to 2019. The data is recorded annually (longitudinal) and the investigation is done on South Korea and ASEAN countries.
This study employs a quantitative design to construct a research model. The GDP per capita will be the dependent variable, and the following independent variables will be investigated: FDI for foreign direct investment (% of GDP), K for gross capital formation (formerly known as gross domestic investment) (% of GDP), T for imports and exports of goods and services (% of GDP), and HC for education index.
3.2. Factors of Economic Growth
The most essential determinants of economic growth and their indexes have been selected based on the previous studies mentioned above in the literature review.
Table 1
List of Variables
| Variable | Description | Predicted Effect |
|---|---|---|
| LnGDP | GDP per capita (constant 2010 US$) | Dependent Variable |
| lnHC | Education index | Independent Variables (+) |
| LnFDI | Foreign direct investment (%of GDP) | Independent Variables (+) |
| LnK | Gross capital formation (%of GDP) | Independent Variables (+) |
| LnT | Trade (%of GDP) | Independent Variables (+) |
3.3. Descriptive Statistics
To present a clear description of practical proof for this paper, the data is clarified with the tables shown below.
The statistical information related to the variables used in this study is reported in table 2 for South Korea. South Korea’s result shows that the per capita GDP has the highest value of?10.26 and the lowest value of 9.04 with a standard deviation of 0.365 and a mean value of 9.777. Education index as a human capital has the highest value of 0.867 and the lowest value of 0.676 with a standard deviation of 0.062 and a mean value of 0.805. Similarly, the foreign direct investment (FDI) has the highest value of 0.768 (%of GDP) and the lowest value of -1.551 (%of GDP) with a standard deviation of 0.607 and a mean value of -0.310. Moreover, gross Capital formation(K) (%of GDP) has the highest value of 3.719 and the lowest value of 3.324 with a standard deviation of?0.105 and a mean value of 3.505, while Trade (T) (% of GDP) has the highest value of 4.659 and the lowest value of 3.848 with a standard deviation of 0.242 and a mean value of 4.232.
Table 2
Descriptive Statistics for South Korea
| Variable | Observations | Mean | Std. Dev | Min | Max |
|---|---|---|---|---|---|
| LnGDP | 30 | 9.777858 | .3658456 | 9.047301 | 10.26378 |
| lnHC | 30 | -.21993 | .07996 | -.39156 | -.14271 |
| LnFDI | 30 | -.3102822 | .6070955 | -1.55135 | .7682479 |
| lnK | 30 | 3.504846 | .1056163 | 3.324282 | 3.719264 |
| LnT | 30 | 4.23172 | .2429286 | 3.848416 | 4.659339 |
Note. Calculation for each variable were based on a real data source using Stata software15
The statistical information related to the variables used in this study is reported in table 3 for ASEAN countries. The result shows that the per capita GDP (constant 2010 US$) has the highest value of?10.99 and the lowest value of 5.28 with a standard deviation of 1.53 and a mean value of 8.02. Education index as a human capital has the highest value of?0.844 and the lowest value of 0.251 with a standard deviation of 0.138 and a mean value of 0.530. Similarly, the foreign direct investment (FDI) has the highest value of 32.16 (%of GDP) and the lowest value of -2.75 (%of GDP) with a standard deviation of 5.50 and a mean value of 5.18. Moreover, gross capital formation(K) (%of GDP) has the highest value of 43.63 and the lowest value of 10.43 with a standard deviation of?7.71 and a mean value of 25.30 and Trade (T) (% of GDP) has the highest value of 437.32 and the lowest value of 0.167 with a standard deviation of 91.63 and a mean value of 119.12.
Table 3
Descriptive Statistics for ASEAN countries
| Variable | Observations | Mean | Std. Dev | Min | Max |
|---|---|---|---|---|---|
| LnGDP | 300 | 8.027249 | 1.530489 | 5.286244 | 10.99162 |
| lnHC | 300 | -.6703911 | .2798343 | -1.382302 | -.1696028 |
| LnFDI | 300 | 5.18627 | 5.492332 | -2.75744 | 32.16984 |
| lnK | 300 | 25.30688 | 7.715668 | 10.4374 | 43.6399 |
| LnT | 300 | 119.1245 | 91.63539 | .1674176 | 437.3267 |
Note. Calculation for each variable were based on a real data source using Stata software15
3.4. Methodology
In this section, the impact of education on economic progress based on existing theories will be presented in terms of a series of equations. Firstly, the classical theory of production function can be used to discuss the impact of education on economic growth. Now we may talk about the production function below, in which output is a function of labor and capital. Mallick (2016) used this strategy, and it was adapted to suit this study’s purpose. It is written as follows: where L denotes the quantity of labor required and K denotes the amount of capital required to produce a given level of production in the economy.
Y = f(L, K) ….(1)
We can incorporate education as an indispensable variable in the production function to test the effect of education on economic growth. This is consistent with Mallick(2016). Thus:
GDP = f(HC) … (2)
here, GDP represents the GDP per capita and HC denotes education. Education promotes the development of human capital, which can result in a skilled labor force. This trained worker force can increase the productivity of both physical and human resources, resulting in increased economic growth. It should be emphasized that different researches used different proxies to measure schooling variables. For example, education quantity can be measured by schooling enrolment ratios defined by Mankiw et al., (1990), Robert Barro(1989),and Levine and David Renelt (1992); adult literacy rate by Durlauf and Johnson (1994) and Romer (1989); and the mean years of schooling such as used by Hanushek and Woessmann (2008) and Krueger and Lindahl (2000).
The influence of education on economic growth is then estimated using Equation (2). This equation can be formulated as follows:
GDPit=α0 +α1 HCit +eit----(3)
Equation (3) was transformed into a linear panel model with education (HC) as the independent variable and per capita GDP as the dependent variable in order to empirically evaluate the relationship between education and GDP. Z denotes a vector of conditioning information that accounts for various economic growth-related aspects, whereas ε is the error term. This is stated as follows:
GDPit=α0 +α1 HCit +α2 Zit +eit----(4)
Barro (1996) and Anaman (2004)mentioned that in addition to human capital accumulation, the endogenous growth model also permits for the addition of other variables such as government expenditure, exchange rate, inflation, labor, consumption expenditure, foreign aid, FDI, corruption, institutional quality, education, financial development, population growth and life expectancy. Therefore, we can replace in Z which represents a vector of all the variables that determine growth rate such as FDI, domestic investment and trade and so on.
We then apply the OLS estimators created for dynamic models of panel data and the resulting equation is as follows:
GDPit=β0+β1(HCit)+β2(FDIit)+β3(Kit)+B4(Tit)+αi +eit --(5)
We modified the above equation number (5) and extended one more equation so we can observe if these variables have an effect on growth and on their own value.
GDPit=β0+β1(HCit)+β2(FDIit)+β3(Kit)+eit---(6)
i=1, 2...10, t=1,2,….30
Where GDP is per capita GDP, HC is education index, FDI is foreign direct investment (percentage of GDP), K is gross capital formation (percentage of GDP), T is trade of goods and services (percentage of GDP), and αi ??i is the countries variables and eit eit is the error term.
We used the Stata 15 statistical software tool to estimate our model using the VAR and ARDL method.
4. Results
This study’s goal was to observe the influence of education on economic growth in ASEAN countries and South Korea and used a unit root test for both time series and panel data before estimating our equation to remove potential spurious estimation results.
4.1. Test of Unit Root for South Korea
This test was used as the initial investigation and accomplished starting the data investigation further. These test results verify that the stationary series does not include problems with the unit root. Augmented Dickey Fuller tests are useful in time series data for South Korea and the outcomes are reported in table 4. There have been three Augmented Dickey Fuller methods such as intercept only, intercept and trend, and no intercept and no trend. In this test statistic, there are three values, 1%, 5% and 10%, and we choose the 5% value normally. If the test statistic is greater than the 5% value, we can reject the null hypothesis and accept the alternative hypothesis and ignore the minus sign. Firstly, regarding the intercept only test, the test statistic is greater than the 5% value, we can reject the null hypothesis. The coefficient of L1 is negative, therefore we can accept this model. Therefore, variables are stationary. In the second model, the intercept and trend test, the test statistic is less than the 5% value, we cannot reject the null hypothesis. Therefore, variables are not stationary. In the third model, the no intercept and no trend test, the test statistic is greater than the 5% value, we can reject the null hypothesis Therefore, variables are stationary. The outcome indicates that at a certain level, non-stationary variables, and at the first difference, all of the variables in I (I) come to be stationary.
Null hypothesis: variable is not stationary and has a unit root.
Alternative hypothesis: variables are stationary and there is no unit root.
Therefore, the series does not express a unit root and can be applied for additional analysis.
Table 4
Test of Unit Root for South Korea
| Series | Level I(0) | First Difference I(1) | ||||||
|---|---|---|---|---|---|---|---|---|
| No intercept | Intercept and Trend | No Intercept and No Trend | No intercept | Intercept and Trend | No Intercept and No Trend | |||
| LnGDP | Test statistic | -3.213 | -1.712 | 6.852 | -5.222 | -6.438 | -2.776 | |
| 5%critical value | -2.989 | -3.584 | -1.950 | -2.992 | -3.588 | -1.950 | ||
| LnFDI | Test statistic | -2.136 | -1.965 | -2.020 | -3.942 | -3.918 | -4.018 | |
| 5%critical value | -2.989 | -3.584 | -1.950 | -2.992 | -3.588 | -1.950 | ||
| LnK | Test statistic | -2.327 | -2.924 | -0.647 | -5.981 | -6.031 | -5.951 | |
| 5%critical value | -2.989 | -3.584 | -1.950 | -2.992 | -3.588 | -1.950 | ||
| LnT | Test statistic | -1.362 | -1.467 | 0.753 | -5.255 | -5.259 | -5.258 | |
| 5%critical value | -2.989 | -3.584 | -1.950 | -2.992 | -3.588 | -1.950 | ||
| LnHC | Test statistic | -3.687 | -0.427 | 4.301 | -4.332 | -6.671 | -3.009 | |
| 5%critical value | -2.989 | -3.584 | -1.950 | -2.992 | -3.588 | -1.950 | ||
Note. Calculation each variable based on real data source using Stata software15
4.2. Test of Unit Root for ASEAN countries
This test is the initial investigation and accomplished starting the data investigation further. These test results verify that the stationary series does not include problems with a unit root. This experiment guarantees the stationary nature of the variables used in the study. The Levin-Lin-Chu and Hadri LM tests are useful and the outcomes are reported in table 5. The outcome indicates that at a certain level, non-stationary variables, and at the first difference, I (I) come to be stationary. Therefore, the series does not express a unit root and can be applied for additional analysis.
Table 5
Test of Unit Root for ASEAN countries
| Variables | Levin, Lin and Chu(H0: Unit Root) | Hadri LM(H0: No Unit Root) | ||||
|---|---|---|---|---|---|---|
| I(0) | I(1) | I(0) | I(1) | |||
| LnGDP | -1.709 | -1.273 | -9.1624 | -4.195* | 56.781 | 4.056 |
| LnFDI | -6.572 | -2.591* | -14.999 | -8.990* | 21.495 | -2.444 |
| LnK | -4.730 | -1.264 | -11.045 | -5.699* | 26.857 | -0.455 |
| LnT | -3.395 | -0.210 | -12.510 | -6.187* | 34.878 | 0.302 |
| LnHC | -3.454 | -2.795* | -10.620 | -5.550* | 57.052 | 2.331 |
Probability Note*, **, *** show level of significance at 10%, 5%, and 1%.
4.3. Test of Multicollinearity
Table 6 indicates the multi-collinearity test in our model. In these two tables, all of the values of the variables are less than ten. If the value of these variables are less than ten, there will not be a multi-collinearity problem in our study.
Table 6
Multicollinearity test for South Korea
| Variable | HC | FDI | K | T |
|---|---|---|---|---|
| Vif | 5.97 | 1.56 | 3.11 | 4.49 |
Multicollinearity test for ASEAN countries
| Variable | HC | FDI | K | T |
|---|---|---|---|---|
| Vif | 1.64 | 1.13 | 1.30 | 1.55 |
Note. Calculation for each variable was based on a real data source using Stata software15
4.4. Test of Heteroscedasticity
We test for heteroscedasticity in table 7, if the p value is greater than 5%, they have a heteroscedasticity problem in ASEAN countries, but South Korea is not. If there is heteroscedasticity, we cannot use an OLS method and another method to solve this problem is needed.
Table 7
Heteroscedasticity test for South Korea
| Test | Null | Hypothesis | Chi-square | Statistic |
|---|---|---|---|---|
| Breusch-Pagan/ Cook-Weisberg | test no | heteroscedasticity | 1.2 | 0.2728 |
Heteroscedasticity test for ASEAN countries
| Test | Null | Hypothesis | Chi-square | Statistic |
|---|---|---|---|---|
| Breusch-Pagan/ Cook-Weisberg | test no | heteroscedasticity | 6.03 | 0.0141 |
Note. Calculation for each variable was based on a real data source using Stata software15
4.5. Test of Co-integration for South Korea and ASEAN countries
4.5.1. Test of Co-integration for South Korea
Given the data is stationary at first difference, we test for the presence of co-integration as the next step. Variables are tested from unit root, the following stage that follows assesses the long-term relationship between foreign direct investment, capital formation, and education. As presented in table 8, the co-integration test results show that we can accept the null hypothesis and that there is no co-integration, Thus, we cannot obtain the long-term relationship among the variables from 1990 to 2019. If variables are co-integrated and I(1) or stationary, this study can guess a regression between the levels of variables without distress of facing a spurious regression.
Co-integration rank is estimated by using the Johansen test of co-integration methodology. There are two likelihood estimators for the rank of co-integration. The results are presented in table 8.
To consider the hypothesis that if trace statistic is more than critical value, we can reject null hypothesis. Here, trace statistic is less than critical value, we cannot reject null hypothesis. Four variables are not co-integrated and do not have long-run association ship. If Max-eigen statistic is more than critical value, we cannot accept the null hypothesis. Here, Max-eigen statistic is less than critical value, we cannot reject null hypothesis. Four variables are not cointegrated and do not have long-run association ship.
Therefore, our four variables are not co-integrated model in this system. If the variables are co-integrated or have long run association-ship, we can run restricted VAR, that is VECM model. But if the variables are not co-integrated, we cannot run VECM model and we shall run unrestricted VAR. Therefore, we can estimate by using the unrestricted VAR Model. And there is no long run causality and there is short run causality running from independent variable to dependent variable.
Table 8
Test of Co-integration for South Korea
| Unrestricted Cointegration Rank Test (Trace) | ||||
|---|---|---|---|---|
| Hypothesized No. of CE(s) | Eigenvalue | Trace Statistic | 0.05 Critical Value | Prob.** |
| None | 86.00 | 68.52 | ||
| At most 1 | 0.80350 | 40.542 | 47.21 | |
| At most 2 | 0.55430 | 17.915 | 29.68 | |
| At most 3 | 0.34029 | 5.7631 | 15.41 | |
| At most 4 | 0.16012 | 1.3775 | 3.76 | |
Trace test indicates no cointegration at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
| Unrestricted Cointegration Rank Test (Maximum Eigenvalue) | ||||
|---|---|---|---|---|
| Hypothesized No. of CE(s) | Eigenvalue | Maximum Eigenvalue | 0.05 Critical Value | Prob.** |
| None | 45.5593 | 33.46 | ||
| At most 1 | 0.80350 | 22.626 | 27.07 | |
| At most 2 | 0.55450 | 11.652 | 20.97 | |
| At most 3 | 0.34042 | 4.885 | 14.07 | |
| At most 4 | 0.16012 | 1.3775 | 3.76 | |
4.5.1.1. Vector Autoregressive Estimation
In this model, each variable is a linear function of past lags of itself and past lags of the other variables. In the analysis, when the maximum value of lag is 2, the SC value (Schwartz criterion) arrives at a minimum (-14.552). While at the same time, the AIC (Akaike information criterion) is - 17.169. The simple mathematical expression of the VAR model is as follows:
Lngdp= 2.218 + 0.526 lngdpt-1 -1.581 lnedut-1
Lnedu= 0.374 -0.040lngdpt-1 + 0.798 lnedut-1
Using the VAR model, it worthy to note that edu is negatively affected by the previous period, and LnGDP is also negatively influenced by the previous period.
4.5.1.2. Granger Causality Wald test
Null hypothesis: All the education lag variable does not cause per capita gdp.
Alternative: All the education lag variable does cause per capita gdp.
There is short run causality education to lngdp as a whole. But lngdp to education , there is no short run causality.
Table 9
Granger Causality Wald test
| Dependent variable - Lngdp | Dependent variable -Education index | ||||||
|---|---|---|---|---|---|---|---|
| Excluded | Chi2 | df | prob | Excluded | Chi2 | df | prob |
| LnEducation index | 4.366 | 3 | 0.113 | Lngdp | .705 | 3 | 0.703 |
| All | 18.187 | 12 | 0.02 | All | 5.151 | 12 | 0.741 |
4.5.1.3. Serial Autocorrelation test
If p-value is less than 5%, we can reject Ho and accept the alternative hypothesis. Here, p value is greater than 5%, we cannot reject null hypothesis and our variables do not have serial correlation problem.
Table 10
Serial Autocorrelation test
| Lagrange multiplier test | Chi2 | df | Prob>chi2 |
|---|---|---|---|
| Lag 1 | 14.218 | 25 | 0.9577 |
Note. H0: no autocorrelation at lag order
4.5.1.4. Normal Distribution Test
In table 11, as a whole, p value is less than 5%, all of the variables are normally distributed.
Table 11
Normal Distribution test
| Jarque bera test | Chi2 | df | Prob>chi2 |
|---|---|---|---|
| lngdp | 11.316 | 2 | 0.00349 |
| LnEducation index | 0.198 | 2 | 0.9056 |
| lnFDI | 17.320 | 2 | 0.00049 |
| lnK | 1.769 | 2 | 0.4129 |
| lnT | 0.385 | 2 | 0.824 |
| All | 30.989 | 10 | 0.00059 |
4.5.1.5. Empirical Results for South Korea
This paper examines an empirical investigation of the relationship between education and economic growth in both the short and long run, as influenced by dynamic variables and their inter-related impact. Using the most updated available data and advanced econometrics approach, we exemplified our analysis by employing the VAR model to time-series data from 1990 to 2019. Findings revealed that the Granger causality test displays a one-sided causal relationship from economic growth to education index, which implies that the economic growth influences the education index. In contrast, education index does not play a significant role in promoting economic growth. There is only a short run causality with these variables.
4.5.2. Test of Co-integration for ASEAN countries
Given the data as stationary at first difference, we test for the presence of co-integration as the next step for ASEAN countries.
Variable are tested from unit root, the following stage that follows assesses the long-term relationship between foreign direct investment, capital formation, and education. As presented in tables 12,13 and 14 the Kao, Pedroni, and Westerlund co-integration test results show that we can accept the null hypothesis and there is no co-integration, Thus, we cannot obtain the long-term relationship among the variables from 1990 to 2019. If variables are co-integrated and I(1) or stationary, this study can guess a regression between the levels of variables without distress of facing a spurious regression. If variables are not co-integrated, they do not exhibit a long run relationship. We have to estimate only Autoregressive distributed lag model.
Table 12
Co-integration Kao test for ASEAN countries
| Ho: do not have co-integrationHa: All panels are co-integrated Estimates | Stats. | Prob. |
|---|---|---|
| Modified Dickey-Fuller t | 0.8937 | 0.1857 |
| Dickey-Fuller t | 0.5463 | 0.2927 |
| Augmented Dickey-Fuller t | 0.5271 | 0.2991 |
| Unadjusted modified Dickey-Fuller t | 1.3675 | 0.0857 |
| Unadjusted Dickey-Fuller t | 1.0270 | 0.1522 |
Table 13
Co-integration Pedroni test for ASEAN countries
| Ho: do not have co-integrationHa: All panels are co-integrated Estimates | Stats. | Prob. |
|---|---|---|
| Modified Phillips-Perron t | 1.7990 | 0.0360 |
| Phillips-Perron t | 0.0352 | 0.4859 |
| Augmented Dickey-Fuller t | -0.2622 | 0.3966 |
Probability Note*, **, *** show level of significance at 10%, 5%, and 1%.
Table 14
Co-integration westerlund test for ASEAN countries
| Ho: do not have co-integrationHa: Some panels are co-integrated Estimates | Stats. | Prob. |
|---|---|---|
| Variance ratio | -1.1802 | 0.1190 |
Probability Note*, **, *** show level of significance at 10%, 5%, and 1%.
4.5.3. Empirical Results for ASEAN countries
The panel ARDL models have the advantage of reducing endogeneity. The PMG estimator allows for different responses across countries in the short-run non-heterogeneity in the long-run. The principal advantage of PMG estimators is the good results in case of small number of countries (only ten countries in this study), simultaneous correction of autocorrelation and the minimum sensitivity in case of outliers.
The results of the pooled mean group (PMG) and mean group (MG) are reported in Table 6. According to PMG and MG estimators, HC has a positive and significant impact on LnGDP in the long run, whereas MG estimator suggests significant impact of HC on LnGDP both in the long-run and short-run. The PMG estimator do not support short run causality between lnHC and LnGDP variables. However, the MG estimator shows the existence of short-run causality between the two variables at 5% level of significance. However, in order to measure efficiency and consistency among the estimators (PMG and MG), the Hausman test has been applied. The validity of long-run homogeneity restrictions across ASEAN countries, and hence efficiency of PMG estimator over the MG estimators, is examined by Hausman test. The Hausman test results accept the null hypothesis of homogeneity restrictions on the long-run regressors, which indicates that PMG is a more efficient estimator than MG. From the overall panel ARDL model, we found that education index led growth hypothesis is valid in ASEAN countries.
Table 15 indicates that education has a favorable impact on economic growth. As a result, the coefficient of education shows a positive relationship and is significant. This means that as the education index increases, per capita GDP increases. If education increases by 1%, per capita GDP increases by 22 % in ASEAN countries. Moreover, FDI and domestic investment is a positive relationship, but trade has a negative relationship on economic growth in ASEAN countries. If FDI increases by 1%, per capita GDP increases by 7 % in ASEAN countries. However, there is only a short run relationship.
Table 15
Panel ARDL Model Results (Pooled Mean Group and Mean Group Estimates) in ASEAN countries (Dependent Variable: Lngdp)
| Variable | Pooled Mean Group | Mean Group | |
|---|---|---|---|
| Long run relationship | LnHC | 12.792* | 14.002* |
| Lnfdi | 0.1311 | 0.154 | |
| LnK | 1.054* | 0.076 | |
| LnT | -1.138 | -1.868* | |
| Error Correction Term | 0.000 | 0.101* | |
| Short run relationship | LnHC | 0.220 | .777* |
| Lnfdi | 0.007* | 0.008 | |
| LnK | 0.002 | 0.030 | |
| LnT | -0.002 | 0.015 | |
| Observations | 330 | 330 | |
| Hausman test | Chi square | 0.09 | |
| p-value | -0.991 |
Probability note. *, **, *** show significance level at 10%, 5%, and 1%.
The objective of our research was to analyze the effects of education on economic growth. The key outcomes point out that education index, FDI and trade have a positive effect on economic growth in ASEAN countries. The main outcomes are strong on the inclusion of these other determinants of growth such as FDI, trade and capital formation.
5. Conclusions and Policy Recommendations
The purpose of this study was to observe the effects of education on economic growth in South Korea and ASEAN countries for the 1990?2019 period. For ASEAN countries, we have applied the panel ARDL model (PMG and MG) to verify the short-run and long-run effects between education index and per-capita gross domestic product. Empirical results revealed that education index positively affects the GDP per capita in ASEAN countries. It expresses that economic growth is reliant on the effort in education and devoting more in education is an important provider to the country’s economic growth. This study highlights the economic status of each country within the ASEAN countries and their economic performance with FDI, trade, and capital formation. Foreign direct investment and domestic investment have a positive impact on ASEAN’s GDP per capita but trade show a negative impact on GDP per capita in ASEAN countries.
For South Korea, we demonstrated our analysis by using the VAR model to time-series data from 1990 to 2019. Findings revealed that the Granger causality test shows a one-sided causal relationship from economic growth to education index, which implies that the economic growth influences the education index. In contrast, education index does not play a significant role in promoting economic growth. There is only a short run causality with these variables in South Korea.
The more human capital enables, the more the absorption of higher technologies from other countries. This channel is expected to be mainly essential for schooling. Cabauatan and Manalo (2018) described that most of the research has indicated that education has a positive impact on economic growth. However, some research has found that economic expansion has an impact on schooling. Together, with the growth of the economy, the government can devote more and expand education expenditure to meet the population's claim for human capital. To improve the effectiveness of ASEAN countries’ education, ASEAN countries should invest in the education sector and strive to improve the quality of their human resources.
According to this research, the government should consider enhancing the quality of education as human capital in order to boost economic growth. Furthermore, the findings of this study are useful to policymakers who may create efficient government programs to boost the country's per capita GDP growth.
Author’s Biography
Mrs. Khin Theingi Aung is of Myanmar nationality and is currently a PhD candidate majoring in Economics at Pusan National University, Busan. Her specialty is in ASEAN study, economic growth, and foreign direct investment. Mrs. Theingi graduated with a Master of Development Policy from the Korea Development Institute in Sejong City in 2018 and also holds a Master’s and Bachelor’s degree in Economics from Meiktila University of Economics in Myanmar. She served as an urban planning officer in Ministry of Planning and Finance in Myanmar for about 10 years from 2009 to 2019.
저자소개
미얀마 국적의 Khin Theingi Aung 은 현제 부산대학교에서 경제학 박사과정생 입니다. 저자는 아세안학, 경제발전, 그리고 해외직접투자를 전공하고 있습니다. 또, 지난 2018년 한국개발연구원 (KDI)에서 개발정책 석사학위를 수료하였고 미얀마의 Meiktila University of Economics에서 경제학 학사를 수료했습니다. 지난 2009년부터 2019년까지 10년간 미얀마 기획재정부에서 근무한 경험이 있습니다.
References
Acemoglu, D. (2012). Introduction to Economic Growth. Journal of Economic Theory, 147(2), 545?550. https://doi.org/10.1016/j.jet.2012.01.023
Afzal, M., Farooq, M., & Ahmad, H. (2010). Relationship between School Education and Economic Growth in Pakistan: ARDL Bounds Testing Approach to Cointegration. Pakistan Economic and Social Review, 48(1), 39?60. http://www.jstor.org/stable/41762413
Anaman, K. A. (2004). Determinants of Economic Growth in Brunei Darussalam. Journal of Asian Economics, 15(4), 777?796. https://doi.org/10.1016/j.asieco.2004.05.019
Barro, R. J. (1996). Determinants of Economic Growth: A Cross-Country Empirical Study. In National Bureau of Economic Research (5698 NATIONAL; Vol. 66, Issue December).
Blomstrom, M., Globerman, S., Kokko, A., Blomstrom, M., Kokko, A., & Globerman, S. (2000). The Determinants of Host Country Spillovers from Foreign Direct Investment. CEPR Discussion Paper, 2350, 44. http://search.ebscohost.com/login.aspx?direct=true&db=ecn&AN=0693618&site=ehost-live&scope=site
Breton, T. R. (2012). The Role of Education in Economic Growth: Theory, History, and Current Returns. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2184492
Cabauatan, R. R., & Manalo, R. A. (2018). A Comparative Analysis on Selected Issues on Economics of Education in ASEAN Countries. Review of Integrative Business and Economics Research, Suppl. Supplementary Issue, 7(1), 68?78. http://buscompress.com/journal-home.html
Desai, M. A., Foley, C. F., & Hines, J. R. (2005). Foreign Direct Investment and the Domestic Capital Stock. In NBER Working Paper Series, January.
Durlauf, S. N., & Johnson, P. A. (1994). Multiple Regimes and Cross-Country Growth Behavior. UNIVERSITV OF WISCONSIN- MADISON.
Borensztein, E., Gregorio, J.D. & Lee, J.W. (1995). How does Foreign Direct Investment Affect Economic Growth? In NBER Working Paper Series 5057.
Feldstein, M. (1994). Tax Policy and International Capital Flows. NBER Working Paper Series, 4851.
Frankel, J. A., & Romer, D. (1999). Does trade cause growth? American Economic Review, 89(3), 379?399. https://doi.org/10.1257/aer.89.3.379
Han, J.S. & Lee, J.W. (2019). Demographic change,Human Capital,and Economic Growth in Korea. CAMA Working Paper.
Hanushek, E. A., & Woessmann, L. (2008). The Role of Cognitive Skills in Economic Development. Journal of Economic Literature, 607?668.
Hill, R. C., William E, G., & C.Lim, G. (2011). Principles of Econometrics.
Iamsiraroj, S. (2016). The foreign direct investment-economic growth nexus. International Review of Economics and Finance, 42, 116?133. https://doi.org/10.1016/j.iref.2015.10.044
Kim, G.J. (2001). Education Policies and Reform in South Korea. Working Paper Series, December, 1?92.
Krueger, A. B., & Lindahl, M. (2000). Education for Growth: Why and For Whom? NBER Working Paper Series.
Lee, J.W. (2016). The Republic of Korea’s Economic Growth and Catch-Up: Implications for the People’s Republic of China. In SSRN Electronic Journal (Issue 571). https://doi.org/10.2139/ssrn.2779486
Levine, R., & Renelt, D. (1992). A Sensitivity Analysis of Cross Country Growth Regressions. American Economic Review.
Liao, L., Du, M., Wang, B., & Yu, Y. (2019). The impact of educational investment on sustainable economic growth in Guangdong, China: A cointegration and causality analysis. Sustainability (Switzerland), 11(3). https://doi.org/10.3390/su11030766
Lucas, R. E. (1988). On the mechanics of economic development. Journal of Monetary Economics, 22(1), 3?42. https://doi.org/10.1016/0304-3932(88)90168-7
Mallick, L. (2016). Impact of educational expenditure on economic growth in major Asian countries?: Evidence from econometric analysis. Theoretical and Applied Economics, XXIII(2), 173?186.
Mankiw, N. G., And, D. R., & Weil, D. N. (1990). A Contribution to the Empirics of Economic Growth. In NBER Working Paper Series.
Manlagnit, M. C. V. (2011). The Economic Effects of Foreign Bank Presence: Evidence from the Philippines. Journal of International Money and Finance, 30(6), 1180?1194. https://doi.org/10.1016/j.jimonfin.2011.06.015
Barro, R.J. (1989). Economic Growth in a Cross Section of Countries. In NBER Working Paper Series (Issue 3120).
Barro, R.J., & Lee, J.W. (2013). A New Data Set of Educational Attainment in the World, 1950?2010. NBER Working Papers, 92(90), i. https://doi.org/10.5089/9781451851335.001
Romer, P. M. (1989). Human Capital and Growth. NBER Working Paper Series, 3173.
Tarabini, A. (2010). Education and Poverty in the Global Development Agenda: Emergence, Evolution and Consolidation. International Journal of Educational Development, 30(2), 204?212. https://doi.org/10.1016/j.ijedudev.2009.04.009