Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.

20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

The Linkage Among Economic Growth, Education and Health:

Empirical Study in Java Island

Sang Aji Kharismarizky Susilo, Banatul Hayati, Amin Pujiati

Faculty of Economics and Business, Diponegoro University
Corresponding Author: sangajikharismarizky@gmail.com

Received: November 2022 | Revised: January 2023 | Accepted: March 2023

Abstract
Java Island accounts for more than half of Indonesia’s gross domestic product (GDP), hence it is
where the majority of economic activities are concentrated, even though mostly of those major
contributions are from certain cities. However, the quality of education and health shown by average
years of schooling and life expectancy at birth in Java Island is unevenly distributed when compared
with economic growth. This research aims to analyze the causalities and effects of economic growth,
education, and health using simultaneous equations models on panel data that consists of 34 cities
in Java Island spanning from 2015 to 2019, which are decomposed into three income classifications.
Results indicate positive and significant effect from education toward economic growth on every
income level. Health instead only brings positive and significant effect toward economic growth in
high-income cities. Economic growth affects education positively in upper-middle and lower-middle
income cities, and affects health positively only in upper-middle income. Ultimately, the relationship
between education and health shows that they have positive and significant effects on each other and
are consistent across all income groups.
Keywords: Economic growth; Human capital; Simultaneous equation model
JEL classification: O15; O47; I15; I25; C30
How to Cite: Susilo S. A. K., Hayati B., Pujiati A. (2023). The Linkage Among Economic Growth,
Education and Health: Empirical Study in Java Island, 24(1), 24-39. doi:https://doi.org/10.23917/jep.
v24i1.20194
DOI: https://doi.org/10.23917/jep.v24i1.20194

1. Introduction et al., 2021). Human capital, as defined by the

Economic growth is a continuous expansion modern approach, has a broader definition as
of various production possibilities as measured now encompasses health factors rather than
by the increase in the GDP over a certain period just the traditional education or skills (Hongyi &
(Parkin, 2011). Todaro & Smith (2015) believed Huang, 2009). Eggoh et al. (2015) and Ogundari
that economic growth also raises a nation’s & Awokuse (2018) supported this by pointing
standards of living and brings great benefits in out that investments in both education and
the form of increased consumption in the future. health create human capital, which is a crucial
Generally, labor and investment are considered component of economic growth and development.
the main contributors to this growth (Bjork, Moreover, the concept of human capital should
1999). However, as time went on, more and be viewed broadly, taking into account aspects
more researchers came to the conclusion that, in related to health and education.
addition to labor and investment, human capital As of 2019, a significant portion of Indonesia’s
is a significant driver of economic growth (Mankiw GDP, 59.03% to be precise, was attributed by
et al., 1992; Barro, 2001; Hanushek, 2013; Dhrifi Java Island (BPS, 2020). The advantages of

24 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

economic growth in Java Island are not without portion of the city is concentrated below the per
explanation, as it is being the center of industries, capita GRDP figure of IDR 123 billion (average
business, and trade in Indonesia (Silalahi, 2019). value), so mostly falls into the Quadrant 3, and
Java Island also has the advantage in terms of Quadrant 4, whereas Quadrant 2 has the fewest
demographics, as it is home to more than half of city, with only one on Figure 1 and two on Figure
the country’s population (BPS, 2021). As a result, 2. The horizontal and vertical lines (separators
it is not surprising that development in Indonesia between quadrants) are drawn from the average
is heavily concentrated on the island of Java. value of each indicator. The diagonal line is for
Although Java Island appears to be superior to the linearity reference between variables.
other islands, the high contribution is only due to The discussion begins with an indication
a few urban areas on the island of Java, especially of the unequal distribution between the quality
the metropolitan area and the provincial capital of education and economic growth (Figure 1).
(Maryaningsih et al., 2007). Generally, there will be a positive correlation
When looking at the distribution of education between the distribution of average school years
and health quality on Java Island, the issue with and economic growth (Hanushek & Woessmann,
human capital becomes apparent, particularly 2010). But in this instance, it scatters erratically
when discussing it at the municipal (city) level. instead. Average years of schooling clustered
At this level, the quality of education and health mostly on the range of 10.0 to 11.0 and 8.0 to
on Java Island appears to be out of step with 9.0. The city with the lowest per capita GRDP is
economic growth, as will be discussed shortly. Banjar, but the lowest average years of schooling
Figure 1 and Figure 2 are the Klassen Typology is held by the city of Tegal, despite having a
for economic growth (proxied by per capita higher per capita GRDP exceeding other 14 cities.
GRDP), education (by average years of schooling), Figure 2 shows the distribution of life expectancy
and health (by life expectancy at birth) in cities at birth against per capita GRDP which is also
on Java Island 2019. Cities with rapid economic spreading irregularly across the chart. Majority of
growth and strong health/education systems the city clustered on life expectancy at birth value
make up Quadrant 1. High economic growth but between 71.0 to 75.0. The biggest discrepancy can
poor educational/health in Quadrant 2, and the be found in Cilegon City, which has the lowest
opposite in Quadrant 3. Quadrant 4 denotes an life expectancy at birth despite having one of the
equal low reading for both indicators. A large highest per capita GDPs on Java Island.

Figure 1. Klassen Typology on GRDP per Capita and Average Years of Schooling
Source: Statistics Indonesia, processed

Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 25

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

Figure 2. Klassen Typology on GRDP per Capita and Life Expectancy at Birth
Source: Statistics Indonesia, processed

Studies on the multidirectional effects to produce consistent results. Besides economic

of economic growth, education, and health growth, the two explanatory variables (health
simultaneously in Java Island are limited. A few and education) must then be described, which
known studies about the contribution of human produces simultaneity issues, and SEM would
capital to economic growth has been conducted be the best method for evaluating the impact of
on Java Island, but they are only in a one- which variables.
way relationship on economic growth. Chotib
& Suharto (2019) investigate the impact of 2. Research Method
human capital on economic growth in East Java, This research aims to test whether aspects
whereas Prasetyo (2020) focuses on Yogyakarta of economic growth, education, and health
and Central Java, and the (2017) paper by here are factors that influence each other.
Anwar does take the entirety of Java Island into High income will lead to better education and
account. Although health and education would healthcare; educated individuals typically
ultimately have an impact on economic growth, lead healthier lives and take better care of
it is established that education promotes health themselves; healthier people will therefore do
and vice versa (Todaro & Smith, 2015). Economic better in acquiring education; while worker
growth, educational, and health attainment have with high human capital will contribute more
been shown to be correlated with one another towards economic growth (Todaro & Smith,
in a tripartite relationship by studies by Dhrifi 2015). One indicator that can represent the
et al. (2021). Henceforth, this study intends quality of education is the average years of
to analyze the causalities between economic schooling. Barro (2001) used this metric to
growth, education, and health in Java Island measure the degree of human capital, as back
and presents empirical evidence of its impact then the concept of human capital is primarily
using simultaneous equations models (SEM). refered only to education. Nowadays, health is
This is due to the fact that while examining the commonly recognized as a component of human
influence of health on economic growth, the role capital that complements education. Acemoglu
of education is also taken into account. The use & Johnson (2007) and Dhrifi et al. (2021) used
of a basic linear specification does not appear life expectancy at birth as the proxy for the level

26 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

of health. Anwar (2017) used both indicators to and third structural equations. Hongyi & Huang
proxy human capital in his paper. (2009) tested the effects of student-teacher
To create the econometric model, this ratios in primary and secondary schools on
paper will use a theoretical framework of the economic growth; this research seeks to be more
macroeconomic production function, which precise by examining their impact on education
emphasizes the three-way linkages between the quality rather than directly on economic growth.
three variables. In order to do this, we’ll be using This study also includes the poverty rate as
an augmented version of Solow’s growth model a predetermined variable that affects health,
(Mankiw et al., 1992), which takes into account because poverty prevents people from getting the
labor, physical capital, and human capital as the proper healthcares they need (Todaro & Smith,
independent variables. 2015). Finally, food expenditures will be included
as a predetermined variables for health’s
Y = K α H β AL1- α- β (1) structural equation. The complete structural
equations are as follow:
Economic growth in this study is proxied
by per capita GRDP which can be written as (y), lnyit = α0 + α1 lnKit + a2 lnLit + a3 lnEdit
and human capital can be dissolve into education + a4 lnHeit + μ1it (3)
and health (Dhrifi et al., 2021; Hongyi & Huang,
2009). Then the equation is converted into an lnEdit = ß0 + ß1lnyit + ß2lnHeit + ß3 lnPrSchit
econometric model. + ß4 lnSeSchit + μ2it (4)

yit = α0 + α1Kit + α2Lit + α3Edit + α4Heit + eit (2) lnHeit = γ0 + γ1 lnyit + γ2 lnEdit + γ3 FdExpit
+ γ4lnPovit + μ3it (5)
This model can be replicated into three
structural equations with economic growth Economic growth proxied by per capita
(eq 3), education (eq. 4), and health (eq. 5) as GRDP (y), (K) is physical capital proxied by
the explanatory variables. The initial research gross fixed capital formation, (L) is labor, (Ed)
hypotheses are as follows: education and health shows quality of education with average years
have a positive impact on economic growth; of schooling as the proxy, (He) is health quality
education and health also have an impact on proxied by life expectancy at birth. (PrSch) and
education quality; and finally, education and (SeSch) are student-teacher ratio for primary
economic growth have an impact on health and secondary school respectively, higher ratio
quality. means higher students per teacher, while lower
As previously said, this study focuses on the ratio means more teachers for students. (FdExp)
three-way relationship between economic growth, is food expenditure per capita, and (Pov) is the
education, and health. In order to analyze it poverty rate. All data are transformed into natural
properly, an estimation using SEM will be logarithm (ln), and all of the structural equations
utilized, rendering economic growth, education, are overidentified. This is a panel data regression
and health as endogenous. Three structural using secondary data sourced from Central
equations will be used, all of which will use the Bureau of Statistics on the span of 5 years (2015-
endogenous variables. As the first structural 2019) for 34 cities in Java Island. These cities can
equation can already be derived from eq. (2), be further decomposed and compared into three
the second and third structural equations must classifications based on income levels derived
be constructed separately, and thus the relevant from “World Bank Classification by Income Level
variables must be identified. Predetermined 2019” which are “high”, “upper-middle”, and
variables are utilized to supplement the second “lower-middle” incomes.

Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 27

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

Table 1. Classification of Income Levels

Classification US Dollar ($) Indonesian Rupiah (Rp)
Lower-middle income $ 1,026 – 3,995 Rp 14,503,536 – 56,473,320
Upper-middle income $ 3,996 – 12,375 Rp 56,487,456 – 174,933,000
High income > $ 12,375 [> Rp 174,933,000

F-test is also being used to determine The key findings of this study, as shown in
the significance of the simultaneous effect of Tables 3, 4, and 5 above, suggest that economic
the independent variables on the dependent growth, education, and health significantly
variable. The three-stage least squares (3SLS) correlate each other, albeit the impacts may
estimation method was chosen because the usage differ, supporting the three-way linkages.
of ordinary least squares (OLS) in the SEM Ed has a positive and significant impact on y
that have overidentified structural equations, across all income levels. On high income cities,
appears unable to offer consistent estimations Ed has a positive effect on y with a coefficient
of model parameters (Gujarati & Porter, 2009). of 3.236; coefficient of 1.7991 on upper-middle
Furthermore, 3SLS outperforms two-stage least income; and coefficient of 2.4487 on lower-middle
squares (TSLS) when considering the potential income. High-income cities appear to be having
of correlation between error terms in structural the highest coefficient, meaning that cities with
equations (Dhrifi et al., 2021). In terms of classical higher GRDP per capita also tend to have better
assumption, all structural equations are normally quality education, and thus policies and actions
distributed on every level of income, and don’t to improve education quality will remarkably
have any multicollinearity problem. Also in this aid economic growth in these places. This may
case, the 3SLS method approach assumes that the be consistent with the study by Dhrifi et al.
structural disturbance in each equation has a zero (2021), which found that education have a more
mean, making it homoscedastic and not serially favorable and significant impact on an economy
correlated (Zellner & Theil, 1962). This suggests with higher income. What is intriguing is that y
that heteroscedasticity and autocorrelation has no significant impact on Ed in high income
testing are unnecessary in this study. cities, as opposed to upper medium and lower-
middle income cities, the former which have the
3. Results and Discussion highest significant coefficient of 0.2967. This could
Three-stage least squares (3SLS) is used to eventually lead to the assumption that education
estimate the above eq. (3), eq. (4), and eq. (5) over in higher-income economies is not caused by the
in order to analyze the three-way relationship high income itself, whereas education in lower-
between economic growth, health, and education. income cities would benefit greatly from economic
The estimation is conducted into cities that are growth.
decomposed into three income classes, in total He only influences y positively in high-
there are nine separate estimation regressions. income cities with the coefficient of 7.634 and
The Hausman test of endogeneity is used adversely on the rest, with the greatest negative
to determine whether a structural equation has coefficient in low-middle-income cities with the
simultaneity problems and whether the dependent coefficient of -14.441. Although this result is
variable is in fact endogenous (Gujarati & Porter, against the earlier hypothesis, the situation is
2009). According to Table 2, all structural comparable to that of Iskandar’s (2017) study,
equations do have simultaneity problems, and which found that employing a human capital
the variables y, Ed, and He can be confirmed as indicator would actually have a negative impact
endogenous variables. Thus, the SEM regression on economic growth instead. According to him,
using the 3SLS method can be carried out. this is due to inefficient allocation and the

28 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

significant disparity between districts and quality compared to those with higher GRDP
cities. Research by Acemoglu & Johnson (2007) per capita, thus making the correlation being
also found that health represented using life negative, confirming the issues explained in the
expectancy tends to have a negative effect on introduction. This could also imply that good
GDP per capita. This can be explained by the quality of health does not necessarily have an
fact that, as shows on Figure 2, cities with lower impact on the high productivity of an economy
GRDP per capita tend to have higher health and vice versa.

Table 2. Hausman Test of Endogeneity

Endogenous
Structural Eq. Residual t-statistic Probability
Variables
Eq. (3) y = Ed, He yResid 7.1509 0.0000***
Eq. (4) Ed = y, He EdResid 17.3975 0.0000***
Eq. (5) He = y, Ed HeResid 30.2361 0.0000***
Note: ***) significant at α 1%; **) α 5%; *) α 10%.

Table 3. Estimation Coefficient on High Income Cities

Variables Economic Growth Eq. (3) Education Eq. (4) Health Eq. (5)
y - 0.0128 -0.0242 ***
K 0.4814 *** - -
L -0.9770 *** - -
Ed 3.2357 ** - 0.9132 ***
He 7.6336 *** 0.6800 *** -
PrSch - 0.1521 -
SeSch - 0.1405 ** -
FdExp - - -0.0404 *
Pov - - -0.0723 ***
Obs 35 35 35
R2 0.89 0.40 0.63

Table 4. Estimation Coefficient on Upper-middle Income Cities

Variables Economic Growth Eq. (3) Education Eq. (4) Health Eq. (5)
y - 0.2132 *** 0.1660 ***
K 0.7351 *** - -
L -0.7889 *** - -
Ed 1.7991 *** - 0.2515 ***
He -9.5452 *** 1.1663 ** -
PrSch - 0.2928 * -
SeSch - -0.3036 *** -
FdExp - - -0.0950 ***
Pov - - -0.0351 ***
Obs 65 65 65
R2 0.78 0.30 0.50

Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 29

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

Table 5. Estimation Coefficient on Lower-middle Income Cities

Variables Economic Growth Eq. (3) Education Eq. (4) Health Eq. (5)
y - 0.2967 *** -0.0507 ***
K 0.1141 * - -
L -0.2018 ** - -
Ed 2.4487 *** - 0.0557 *
He -14.441 *** 6.6489 *** -
PrSch - 0.2242 ** -
SeSch - -0.0912 -
FdExp - - 0.0083
Pov - - -0.0112 **
Obs 70 70 70
R2 0.62 0.54 0.12
Note: ***) significant at α 1%; **) α 5%; *) α 10%.

However, both He and Ed do correlate every income level, with upper-middle income
positively to each other on all income levels, this cities having the largest coefficient of 0.7351,
is in line with the findings of Dhrifi et al. (2021). implying that investment should have the biggest
He affects Ed positively and significantly with the impact on cities with upper-middle income, as
coefficient of 0.680 on high income cities; 1.166 on cities in this classification also have growing
upper-middle income; and 6.649 on lower-middle industries. L has negative and significant impact
income, the former of which have the greatest toward y on all income levels. This implies that
impact. On the other hand, Ed has significant cities with lower per capita GRDP would require
and positive impact on He with high income more labor to create the same amount of output
cities having coefficient of 0.9132 on; upper- than cities with greater per capita GRDP, which is
middle income having 0.2515; and lower-middle inefficient. Based on Solow model, the use of more
income having 0.0557. Based on the respective efficient labor means the rate of technological
coefficients, policies and initiatives to improve development is also advanced (Mankiw &
education quality would benefit healthcare the Scarth, 2010), therefore this unfavorable result
most in places with higher growth, on the other of L appears to confirm that cities are more
hand, cities with lower growth would see the technologically advanced the higher their income.
greatest improvement in education quality as a PrSch has positive and significant correlation to
result of increased healthcare. It is this positive Ed only on upper-and-lower-middle income with
and varied correlation that shows the need to similar coefficient of 0.293 and 0.224 respectively.
balance human development priorities between As the nature of this ratio, positive correlation
education and health between cities with high and means that increase in average years of schooling
low incomes. This effort can be fulfilled by raising would likely be affected by increase in student
the budget of education and healthcare, enhancing quantity, in this case is the primary school.
the quality of buildings and infrastructure, and Meanwhile SeSch correlates Ed positively only
ensuring equal access to education and health. on high income cities with coefficient of 0.141 and
Another effort is to eradicate poverty and improve negatively on upper-middle income cities with
people’s living standards so that they are able to -0.304 coefficient, this means increase in average
get opportunities for education and better health. years of schooling would likely be affected by
Secondary findings in this study are focused increase in student quantity on high income
towards the predetermined variables, they show cities, and by increase of teacher in upper-middle
that K positively and significantly affects y on income cities. FdExp correlates He negatively

30 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

on high income and upper-middle income cities, in Java, the aforementioned findings that showed
meaning increased food consumption is inversely the unusual negative correlation between labor
connected to health. Pov negatively affects He and economic growth should also be resolved.
on every income level with highest coefficient is
on high income cities with coefficient of -0.072. 4.1 Limitation
This indicates that poverty will have a greater The limitation of this study is that the use
impact on health when the economy has a higher of indicator in this study, specifically on labor
standard of living. variables represented by the number of Work
Force, and health variables represented by the
4. Conclusions Life Expectancy at Birth, in this case (34 Cities
This study investigates causal linkages and in Java Island 2015 - 2019). They naturally tend
impacts among economic growth, education, and to show a negative correlation in cross-section,
health in Java Island by using simultaneous even though the value shows a steady increase
equations models. Of all the above, it can be in time series, this is most probably caused by
concluded that economic growth, education, and the high disparity on these cities. As a result,
health have substantial correlations, albeit the the estimation results of those two variables had
effects may differ. In line with what is disclosed a negative effect on Economic Growth, resulting
in the introduction, the primary findings indicate in contradictory estimation results with earlier
that there are indeed issues on human capital studies and the possible formation of new research
distribution, specifically health, in Java Island gaps.
indicated by the life expectancy at birth indicator
that negatively correlates with per capita GRDP. 4.2 Future Scope
Between education and health, on the other hand, Due to the limitations encountered in
are favorably associated. This study also finds completing this investigation, further study is
unusual results regarding the economic growth still required to pursue more conclusive results.
that is negatively impacted by labor, but the There are various recommendations that future
plausible explanation is that this was caused by studies should consider if they want to continue
inefficient worker on cities with lower per capita their research in this area. One suggestion is
GRDP. to use more various indicator of education and
Empirical findings in this study have health, might be an improvised index of many
important implications that could be made into factors combined that could reflect the actual
consideration. Based on what have been discussed, quality of both aspects closest to reality. Besides
special emphasis should be placed on improving this, the sample of cities needs to be increased to
health outcomes and life expectancy in these yield more accurate results considering the topic
cities. This can be accomplished by broadening itself is very cross-section intensive.
access to healthcare, establishing health
insurance programs, and creating a physically and 5. References
mentally healthy work environment. Regarding Acemoglu, D., & Johnson, S. (2007). Disease and
the education quality, there must be government development: the effect of life expectancy
policies and programs in place to ensure that on economic growth. Journal of Political
students may complete their education until Economy, 115(6), 925–985.
they all graduate from high school, and more Anwar, A. (2017). The Role of Human Capital To-
chances for tertiary education must be made ward Regional Economic Growth in Java.
available. Another aspect that is considered to be Jurnal Economia, 13(1), 79–94.
able to support the improvement of the quality
of human capital is efforts to alleviate poverty Barro, R. J. (2001). Human capital and growth.
and improve the welfare of the population. As American Economic Review, 91(2), 12–17.
with the improvement of human capital quality Bjork, G. C. (1999). The way it worked and why it

Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 31

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

won’t: structural change and the slowdown growth. The Quarterly Journal of Econom-
of US economic growth. Greenwood Pub- ics, 107(2), 407–437.
lishing Group. Mankiw, N. G., & Scarth, W. M. (2010). Macro-
BPS. (2020). Pertumbuhan Ekonomi IndonesiA economics (Canadian Edition). Macmillan.
Triwulan IV-2020. Berita Resmi Statistik Maryaningsih, N., Hermansyah, O., & Savitri, M.
No 06/01/th. XXIII. (2007). Pengaruh infrastruktur terhadap
BPS. (2021). Hasil sensus penduduk 2020. Berita pertumbuhan ekonomi Indonesia. Bulletin
Resmi Statistik No. 7/01/Th. XXIV. of Monetary Economics and Banking, 17(1),
Chotib, M., & Suharto, B. (2019). Optimization 62–98.
of human capital development on economic Ogundari, K., & Awokuse, T. (2018). Human
growth and poverty in east java. Interna- capital contribution to economic growth
tional Journal of Scientific and Technology in Sub-Saharan Africa: does health status
Research, 8(9), 652–657. matter more than education? Economic
Dhrifi, A., Alnahdi, S., & Jaziri, R. (2021). The Analysis and Policy, 58, 131–140.
Causal Links Among Economic Growth, Parkin, M. (2011). Macroeconomics, (Pearson Se-
Education and Health: Evidence from De- ries in Economics). Prentice Hall.
veloped and Developing Countries. Journal Prasetyo, P. E. (2020). Human capital as the main
of the Knowledge Economy, 12(3), 1477– determinant of regional economic growth.
1493. International Journal of Advanced Science
Eggoh, J., Houeninvo, H., & Sossou, G.-A. (2015). and Technology, 29(03), 6261–6267.
Education, health and economic growth in Silalahi, S. A. F. (2019). Dampak Ekonomi dan
African countries. Journal of Economic De- Resiko Pemindahan Ibu Kota Negara. Info
velopment, 40(1), 93. Singkat: Kajian Singkat Terhadap Isu Ak-
Gujarati, D. N., & Porter, D. C. (2009). Basic tual dan Strategis, XI (No.16/II/Puslit/
econometrics. McGraw-Hill/Irwin. New Agustus/2019), 19-24
York. Todaro, M. P., & Smith, S. C. (2015). Economic
Hanushek, E. A. (2013). Economic growth in de- development (ed.). USA: Person Education
veloping countries: The role of human cap- Limited.
ital. Economics of Education Review, 37, Zellner, A., & Theil, H. (1962). Three-Stage
204–212. Least Squares: Simultaneous Estimation
Hanushek, E. A., & Woessmann, L. (2010). Edu- of Simultaneous Equations. Econometrica,
cation and economic growth. Economics of 30(1), 54.
Education, 60, 67.
Hongyi, L. I., & Huang, L. (2009). Health, educa-
tion, and economic growth in China: Empir-
ical findings and implications. China Eco-
nomic Review, 20(3), 374–387.
Iskandar, I. (2017). Effect of human development
index fund on economic growth through a
special autonomy. Jurnal Ekonomi Pem-
bangunan: Kajian Masalah Ekonomi Dan
Pembangunan, 18(1), 40–49.
Mankiw, N. G., Romer, D., & Weil, D. N. (1992).
A contribution to the empirics of economic

32 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

6. Appendices

Appendix 1
Sample decomposition based on “World Bank Classification by Income Level 2019” with the
Rupiah exchange rate against the US Dollar based on the 2019 average exchange rate of US $ 1 = Rp
14,136:
a. Lower-middle income : $ 1,026 – 3,995 [Rp 14,503,536 – 56,473,320]
b. Upper-middle income : $ 3,996 – 12,375 [Rp 56,487,456 – 174,933,000]
c. High income : > $ 12,375 [> Rp 174,933,000]

Per capita Per capita Per capita

Cities Cities Cities
GRDP ($) GRDP ($) GRDP ($)
Kota
High Income 6,568 Kota Depok 2,129
Surakarta
Kota Jakarta Selatan 20,093 Kota Salatiga 4,873 Kota Cimahi 3,839
Kota
Kota Jakarta Pusat 54,485 7,494 Kota Tasikmalaya 2,279
Semarang
Kota Jakarta Barat 12,849 Kota Tegal 4,349 Kota Banjar 1,662
Kota
Kota Jakarta Utara 20,416 5,961 Kota Pekalongan 2,518
Yogyakarta
Kota Kediri 34,238 Kota Malang 5,909 Kota Blitar 3,384
Kota Surabaya 14,174 Kota Madiun 5,635 Kota Probolinggo 3,382
Kota Cilegon 16,860 Kota Batu 5,763 Kota Pasuruan 2,923
Kota
Upper Middle Income 5,532 Kota Mojokerto 3,716
Tangerang
Kota Jakarta Timur 11,895 Lower Middle Income Kota Serang 3,253
Kota Bandung 8,006 Kota Bogor 3,001 Kota Tangerang Sel 3,492
Kota
Kota Cirebon 5,266 2,682
Sukabumi
Kota Magelang 5,127 Kota Bekasi 2,358

Appendix 2
Normality and multicollinearity test results
High Income Cities
Eq. (3) Eq. (4) Eq. (5)
8 6 5

7
5
4
6
4
5
3

4 3

3 2
2
2
1
1
1

0 0 0
-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 -0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 -0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04

Jarque-Bera: 0.9168 (0.6322) Jarque-Bera: 1.1761 (0.5554) Jarque-Bera: 0.6136 (0.7357)

Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 33

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

Upper-middle Income Cities

Eq. (3) Eq. (4) Eq. (5)
10 20 12

10
8 16

8
6 12

6
4 8
4

2 4
2

0 0 0
-0.2 -0.1 0.0 0.1 0.2 0.3 -0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10 0.15 -0.04 -0.02 0.00 0.02 0.04 0.06

Jarque-Bera: 1.1310 (0.5680) Jarque-Bera: 2.1664 (0.3384) Jarque-Bera: 3.6802 (0.1587)

Lower-middle Income Cities
Eq. (3) Eq. (4) Eq. (5)
14 8 9

7 8
12
7
6
10
6
5
8 5
4
6 4
3
3
4
2 2
2 1 1

0 0 0
-0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 -0.10 -0.05 0.00 0.05 0.10 0.15 -0.06 -0.04 -0.02 0.00 0.02 0.04

Jarque-Bera: 5.4907 (0.0642) Jarque-Bera: 1.7871 (0.4091) Jarque-Bera: 2.4827 (0.2889)

High Income Cities

Eq. (3) K L Ed He
K 1
L 0.5333 1
Ed 0.7627 0.4602 1
He 0.2016 0.3705 0.6435 1
Eq. (4) y He PrSch Sesch
y 1
He 0.2335 1
PrSch -0.1751 -0.0619 1
Sesch 0.1039 0.4696 0.2916 1
Eq. (5) y Ed FdExp Pov
y 1
Ed 0.2815 1
FdExp -0.0947 0.5345 1
Pov 0.2021 -0.3457 -0.5664 1
Upper-middle Income Cities
Eq. (3) K L Ed He
K 1
L 0.7472 1

34 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

Upper-middle Income Cities

Ed 0.4137 0.4006 1
He 0.2180 -0.0676 0.2129 1
Eq. (4) y He PrSch Sesch
y 1
He 0.2023 1
PrSch 0.5398 -0.2805 1
Sesch 0.3428 -0.3086 0.7777 1
Eq. (5) y Ed FdExp Pov
y 1
Ed 0.6666 1
FdExp 0.6691 0.5501 1
Pov -0.5973 -0.3063 -0.5895 1
Lower-middle Income Cities
Eq. (3) K L Ed He
K 1
L 0.7174 1
Ed 0.6484 0.5246 1
He 0.4480 0.3937 0.5887 1
Eq. (4) y He PrSch Sesch
Y 1
He -0.0771 1
PrSch 0.1606 0.2005 1
Sesch -0.0445 0.3597 0.7720 1
Eq. (5) y Ed FdExp Pov
y 1
Ed 0.2861 1
FdExp 0.2404 0.7461 1
Pov -0.2394 -0.6729 -0.7447 1

Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 35

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

Appendix 3
Estimation Results on High Income Cities
System: UNTITLED
Estimation Method: Three-Stage Least Squares
Date: 08/11/22 Time: 19:10
Sample: 2015 2019
Included observations: 35
Total system (balanced) observations 105
Linear estimation after one-step weighting matrix

Coefficient Std. Error t-Statistic Prob.  

C(1) -20.42047 3.910712 -5.221677 0.0000
C(2) 0.481442 0.115464 4.169612 0.0001
C(3) -0.977022 0.080463 -12.14257 0.0000
C(4) 3.235717 1.401849 2.308179 0.0233
C(5) 7.633551 1.072127 7.120005 0.0000
C(6) -1.565094 0.955183 -1.638527 0.1048
C(7) 0.012840 0.017030 0.754006 0.4528
C(8) 0.680009 0.247245 2.750348 0.0072
C(9) 0.152137 0.115082 1.321996 0.1895
C(10) 0.140489 0.064308 2.184634 0.0315
C(11) 2.879637 0.339908 8.471820 0.0000
C(12) -0.024184 0.008744 -2.765783 0.0069
C(13) 0.913180 0.099526 9.175319 0.0000
C(14) -0.040409 0.025316 -1.796225 0.0943
C(15) -0.072260 0.013259 -5.449784 0.0000
Determinant residual covariance 6.10E-09

Equation: Y=C(1)+C(2)*K+C(3)*L+C(4)*ED+C(5)*HE
Instruments: K L PRSCH SESCH FDEXP POV C
Observations: 35
R-squared 0.897569     Mean dependent var 12.47913
Adjusted R-squared 0.883911     S.D. dependent var 0.483459
S.E. of regression 0.164723     Sum squared resid 0.814011
Durbin-Watson stat 0.400169

Equation: ED=C(6)+C(7)*Y+C(8)*HE+C(9)*PRSCH+C(10)*SESCH
Instruments: K L PRSCH SESCH FDEXP POV C
Observations: 35
R-squared 0.401498     Mean dependent var 2.347668
Adjusted R-squared 0.321698     S.D. dependent var 0.056418
S.E. of regression 0.046465     Sum squared resid 0.064771
Durbin-Watson stat 0.257954

36 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

Equation: HE=C(11)+C(12)*Y+C(13)*ED+C(14)*FDEXP+C(15)*POV
Instruments: K L PRSCH SESCH FDEXP POV C
Observations: 35
R-squared 0.630548     Mean dependent var 4.284294
Adjusted R-squared 0.581288     S.D. dependent var 0.037293
S.E. of regression 0.024131     Sum squared resid 0.017470
Durbin-Watson stat 0.444455

Estimation Results on Upper-middle Income Cities

System: UNTITLED
Estimation Method: Three-Stage Least Squares
Date: 08/11/22 Time: 20:47
Sample: 2015 2019
Included observations: 65
Total system (balanced) observations 195
Linear estimation after one-step weighting matrix

Coefficient Std. Error t-Statistic Prob.  

C(1) 50.96779 11.97200 4.257251 0.0000
C(2) 0.735133 0.150152 4.895924 0.0000
C(3) -0.788922 0.190653 -4.138007 0.0001
C(4) 1.799101 0.280474 6.414504 0.0000
C(5) -9.545177 2.611790 -3.654649 0.0003
C(6) 4.931631 2.190878 2.250984 0.0256
C(7) 0.213210 0.062964 3.386202 0.0009
C(8) 1.166319 0.545433 2.138335 0.0338
C(9) 0.292782 0.172666 1.695657 0.0917
C(10) -0.303557 0.107550 -2.822461 0.0053
C(11) 4.228329 0.301846 14.00825 0.0000
C(12) 0.166015 0.029411 5.644723 0.0000
C(13) 0.251532 0.073217 3.435419 0.0007
C(14) -0.095034 0.024824 -3.828377 0.0002
C(15) -0.035063 0.013119 -2.672702 0.0082
Determinant residual covariance 3.06E-08

Equation: Y=C(1)+C(2)*K+C(3)*L+C(4)*ED+C(5)*HE
Instruments: K L PRSCH SESCH FDEXP POV C
Observations: 65
R-squared 0.781912     Mean dependent var 11.21805
Adjusted R-squared 0.767372     S.D. dependent var 0.270131
S.E. of regression 0.130288     Sum squared resid 1.018498
Durbin-Watson stat 0.712692

Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 37

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

Equation: ED=C(6)+C(7)*Y+C(8)*HE+C(9)*PRSCH+C(10)*SESCH
Instruments: K L PRSCH SESCH FDEXP POV C
Observations: 65
R-squared 0.298037     Mean dependent var 2.323536
Adjusted R-squared 0.251239     S.D. dependent var 0.093852
S.E. of regression 0.081211     Sum squared resid 0.395716
Durbin-Watson stat 0.678443

Equation: HE=C(11)+C(12)*Y+C(13)*ED+C(14)*FDEXP+C(15)*POV
Instruments: K L PRSCH SESCH FDEXP POV C
Observations: 65
R-squared -0.499499     Mean dependent var 4.307540
Adjusted R-squared -0.599466     S.D. dependent var 0.027234
S.E. of regression 0.034443     Sum squared resid 0.071180
Durbin-Watson stat 0.466791

Estimation Results on Lower-middle Income Cities

System: UNTITLED
Estimation Method: Three-Stage Least Squares
Date: 08/11/22 Time: 21:34
Sample: 2015 2019
Included observations: 70
Total system (balanced) observations 210
Linear estimation after one-step weighting matrix

Coefficient Std. Error t-Statistic Prob.  

C(1) 68.21245 8.005386 8.520821 0.0000
C(2) 0.114140 0.059880 1.906140 0.0581
C(3) -0.201793 0.079265 -2.545816 0.0117
C(4) 2.448716 0.356810 6.862802 0.0000
C(5) -14.44132 1.885114 -7.660717 0.0000
C(6) -29.70091 3.623361 -8.197061 0.0000
C(7) 0.296686 0.051012 5.815957 0.0000
C(8) 6.648857 0.802388 8.286336 0.0000
C(9) 0.224150 0.086820 2.581771 0.0106
C(10) -0.091160 0.060010 -1.519086 0.1304
C(11) 4.591610 0.123090 37.30299 0.0000
C(12) -0.050718 0.010844 -4.677213 0.0000
C(13) 0.055749 0.069903 1.797514 0.0961
C(14) 0.008326 0.013877 0.599973 0.5492
C(15) -0.011220 0.005882 -1.957449 0.0479
Determinant residual covariance 1.26E-09

38 Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331

 Avalaible online at http://journals.ums.ac.id, Permalink/DOI: 10.23917/jep.v24i1.20194

Jurnal Ekonomi Pembangunan: Kajian Masalah Ekonomi dan Pembangunan, 24 (1), 2023, 24-39

Equation: Y=C(1)+C(2)*K+C(3)*L+C(4)*ED+C(5)*HE
Instruments: K L PRSCH SESCH FDEXP POV C
Observations: 70
R-squared -0.622601     Mean dependent var 10.44630
Adjusted R-squared -0.722453     S.D. dependent var 0.247527
S.E. of regression 0.324861     Sum squared resid 6.859740
Durbin-Watson stat 0.624337

Equation: ED=C(6)+C(7)*Y+C(8)*HE+C(9)*PRSCH+C(10)*SESCH
Instruments: K L PRSCH SESCH FDEXP POV C
Observations: 70
R-squared -0.548316     Mean dependent var 2.267425
Adjusted R-squared -0.643597     S.D. dependent var 0.109366
S.E. of regression 0.140210     Sum squared resid 1.277832
Durbin-Watson stat 0.558247

Equation: HE=C(11)+C(12)*Y+C(13)*ED+C(14)*FDEXP+C(15)*POV
Instruments: K L PRSCH SESCH FDEXP POV C
Observations: 70
R-squared 0.119436     Mean dependent var 4.279132
Adjusted R-squared 0.065247     S.D. dependent var 0.026910
S.E. of regression 0.026017     Sum squared resid 0.043998
Durbin-Watson stat 0.546533

Appendix 4
Formulas on System Equation using 3SLS, [EViews]

Jurnal Ekonomi Pembangunan, ISSN 1411-6081, E-ISSN 2460-9331 39