The Causality between Education and Health Funds Allocation to

Poverty in Indonesia

Muhammad Nasir

, Muhammad Farid

and Chenny Seftarita

Faculty of Economics and Business, Syiah Kuala University, Darussalam, Banda Aceh – Indonesia

Keywords: Education Fund, Health Fund, Poverty, Granger Causality

Abstract: This study aims to examine the causality between education and health fund to poverty in Indonesia. The

data used is quarterly from 1998Q1 to 2017Q4. This research uses granger causality model. The results

show on lag 1, there is no causality between education, health, and poverty. Similar results are also found in

lag estimates 4. Estimates of lag 2 indicate education and poverty have bidirectional relationships.

Meanwhile health and poverty have unidirectional relationships. The peak lag 3 illustrates only one-way

education on poverty vice versa. But health found no causality. The recommendation that the allocation

should have an impact and effective in the short term and increase the allocation of health budget.

1 INTRODUCTION

Poverty has been a problem facing by developing

countries. It has been a serious problem for the

developing countries. It is also a problem for

Indonesia as the country is among developing

countries. Poverty is also viewed as a complex

problem with many dimensions includes social,

economic, culture, and other aspects. In Indonesian

case, poverty causes the difficulties for the people in

fulfilling their needs. Poor people face the lack of

access for a better life.

Indonesian Statistic Board (BPS) via

www.bps.go.id stated that in 1996, the amount of

poor people in Indonesia were 22.5 million people.

The number had increased for the year 1998 with the

amount of 49.5 million people. The sharp increase

for that period of time one was caused by the

financial crisis in South East Asian Countries

including Indonesia. The economic performance in

1998 showed that Indonesian economy had grown

by minus 4 percent which caused many economic

problems such as high unemployment and increasing

in the poverty rate. But, many years after the

economic crisis, Indonesian economy had recovered

since 2005 with the decrease in poverty.

In order to reduce the poverty, Indonesian

government has the policy on increasing human

capital via education, increasing health care via

health insurance skim (BPJS), income support, and

mandatory education requirement (12 years of

schooling), and many other programs. Education has

been the focus of government with budget allocation

as much as 20 percent of total national government

budget (APBN) and regional government budget

(APBD).

The allocation of funds for education and health

is mainly plotted from the tax. Since 2000,

Indonesian government has plotted 20 percent of

APBN for education and 5 percent of APBN for

health care. The efforts in reducing poverty

continuously have been the key for the government.

The policy is also adjusted with the economic

conditions.

Theoretically, poor people face vicious circle of

poverty. According to Chambers in Syarifuddin

(2017), poor households and neighborhood has the

link in one circle that cause poor households in

poverty trap. There are five weaknesses that owned

by poor households, those are the limitation in

assets, weak physical condition, isolation,

vulnerability, and not empowered. On the other

sides, poverty can also be caused by the limitation in

capital as the economic factors. This can be

explained by income, saving, investment, and

productivity.

In terms of government expenditure, according

to Saifuddin (2017), the government expenditure

from one period to another period is not based on

492

Nasir, M., Farid, M. and Seftarita, C.

The Causality between Education and Health Funds Allocation to Poverty in Indonesia.

DOI: 10.5220/0009503204920496

In Proceedings of the 1st Unimed International Conference on Economics Education and Social Science (UNICEES 2018), pages 492-496

ISBN: 978-989-758-432-9

national income. In the economic downfall, for

instance the income from tax decreases. But, in

order to reduce unemployment, the government

needs to release development programs. Thus, the

expenditures have to be increased.

One of the important expenditure of government

is for education. According to Atmanti (2005), there

are many factors as the reasons for the importance of

education development. First, the higher education

level expands the knowledge of the people with high

rationality in thinking. Second, education makes the

people possible in learning technical knowledge that

is needed in ruling modern firms and other activities.

Third, a better knowledge got from education

becomes the stimulus in creating new innovations in

technical, economical manner, and other aspect of

life.

Furthermore, Ehrenberg et al. (2012) viewed

education as the investment. According to him,

fulfilling education means it needs some amount of

funding. Investment in human capital has cost used

in some period of times with the expectation of high

return in the future. In case of investment in

education, one people expect a better return in form

of higher income, increase in work satisfaction and

higher social status.

Muhi (2000) explained that human capital

investment is one of important priority for majority

of people. Majority of people have the expectation to

continue and finish education in the higher level.

Human capital has direct contribution in creating

national assets.

Another Indonesian government expenditure

focus is on health care. This expenditure can

strengthen the health status of the people. Health

program also can give the welfare for the people.

With government involvement, the people can

reduce their expenditure for health.

According to Azwar (2004), health a person is

not only can be seen from physic but also mental.

The body health includes physical, mental, and

social aspects. All those aspects affect the

performance of each individual in doing their

activities such as working and leisure times.

Based on the research background and some

theoretical review, this paper is aimed in analyzing

the causality between education and health fund

allocation on the poverty in Indonesia. The rests of

this paper are designed as follow. Second part is the

methodology, third is research findings, and the last

is conclusion.

2 METHODOLOGY

The method applied in this study was the causality

of poverty in Indonesia using education and health

funds. The period of observation was between years

1998 to 2017 with the observation as much as 19

years. Due to the limitation of data, time series

analysis needs long period of data, thus the

researcher had done the interpolation from yearly

data into quarterly data. Thus, the observation period

had become from 1998Q1 to 2017Q4 with the

sample as much as 76.

The kind of data used in this study is time series

data as secondary data source from Indonesian

Statistic Board (BPS), Ministry of Finance, World

Bank, and others.

The method of analysis used in this research is

quantitative analysis with time series data. The

model of analysis used is the causality analysis

between education and health funds and education

with Granger Causality Test. The use of this is in

order to understand the causality relationship

reciprocally between the variables, where in one side

the dependent variable is affected by independent

variables, and on the other side, the independent

variables can replace the dependent variable

(Saifuddin, 2017).

So far, the formula used in the study are as follow:

1. KMS

= 

+

ΔKMS

t-1

+ 

ΔGOVE

t........

(1)

2. ΔKMS

= β

+ β

ΔKMS

t-1

+ β

ΔGOVH

t......

(2)

3. ΔGOVE

= π

+ π

ΔGOVE

t-1

+ π

ΔKMS

t...

(3)

4. ΔGOVH

= μ

+ μ

ΔGOVH

t-1

+ μ

ΔKMS

t..

(4)

Where:

KMS = the amount of poor people

GOVE = share of the allocation of government fund

to education.

GOVH = share of the allocation of government fund

to health.

The estimation of regression model during the

period of the research uses Granger Causality.

Following Holzt-Eakin, Newey and Rosen, the

Granger Causality Test is formulated in the form of

vector autoregresive (Arfa, 2016) as follow:

Yit = a0 + Σk=1→m ak Yit-k + Σ1-1→n b1Xit-1

+ u1it (5)

Xit = α0 + Σk=1→m αk Xit-k + Σ1-1→n β1 Yit-

1 + u2it (6)

The time series procedures in the test were

applied such as Unit Root Test (Rosadi, 2012; Arfa,

2016), and the length of lag test as explained by

Gujarati (2003). According to him, in testing the lag

in Granger Causality Method, it needs the

determination of lag into some variables in order to

The Causality between Education and Health Funds Allocation to Poverty in Indonesia

493

give better estimation. The more length of lag, the

less will be the degree of freedom (df) of the model,

while the shorter lag will result in error in (Gujarati,

2003).

3 RESULT AND DISCUSSION

The Stationarity Test

Stationarity test has become necessary condition and

first step in estimation of model for some specific

time period mainly in the model of Granger

causality. This test is undertaken in order to test that

the variables in model has the stable pattern or

stationary or not. If the time series data directly

analyzed thus will give false regression analysis.

This will effect in bias conclusion and miss policy

implication. The result of test is as in Table 1.

Table 1: The Result of Stationary Test of ADP and PP

Variable

ADF PP

Level

I(0)

First-

Difference

I (1)

Level

I(0)

First-

Difference

I (1)

KMS -1.86 -3.66*** -1.97 -3.35**

GOVE -1.64 -3.88*** -1.87 -4.11***

GOVH -1.69 -2.67* -1.99 -3.59***

Source: Output Eviews, 2018. ***, **, and * show the level of significance of

1%, 5 %, and 10%.

The stationarity test in this study was using

Augmented Dickey-Fuller (ADF) and Phillips-

Perron (PP). Table 1 show that the poverty variable

(KMS), education variable (GOVE), and health

variable (GOVH) have unit root at level shown by

the insignificant value of ADF and PP or accepted

H0.

Then, the stationarity again test at first difference

and found that the variables of poverty, education,

and health do not have unit root or stationary with

ADF 1 percent and 10 percent, while PP are 5 and 1

percent.

Optimal Lag Test

The optimal lag test is important in Granger

Causality Test. The lag has the function in

explaining how long the effect of one variable on

other variable. Thus, it is needed to undertake the

optimal lag from each path. This can be done by

using Akaike Information Criterion (AIC), Schwarz

Bayesian Criterion (SC), and Hannan Quinn

Criterion (HQ). The expected value is the smaller

one.

Table 2 explained the optimal lag using lag limit

5. The result shows the same lag at 5. The result

shows the similarity for AIC at 3 lag, SC at 3 lag,

and HQ at 3 lag. Based on those three criteria, thus

in this research was used lag 3 as the optimal lag.

Table 2: Lag Information Criteria

Lag AIC SC HQ

0 11.266 11.358 11.303

1 3.112 3.482 3.620

2 -1.167 -0.518 -0.908

-2.320* -1.393* -1.949*

4 -2.253 -1.048 -1.772

5 -2.078 -0.594 -1.485

Sourcer: Estimation result

Granger Causality Test and the Implications

Table 3 show the causality relationship between the

variables using some lags. The estimation result

using suitable optimal lag is shown at lag 3, while

the use of lag 4 in order to verified the relationship

between variables. First, the reduction in poverty is

caused by other factors.

Table 3: The Results of Grabger Causality for Education, Health, and Poverty

Dependent

variable

Independent variable

KMS GOVE GOVH

KMS - 1.605 0.011

GOVE 1.509 - -

UNICEES 2018 - Unimed International Conference on Economics Education and Social Science

494

GOVH 0.437 - -

KMS 16.138*** 7.310***

GOVE 13.685*** - -

GOVH 1.648 - -

KMS - 2.892** 0.770

GOVE 2.137 -

GOVH 0.730 -

KMS - 1.810 0.539

GOVE 1.613 - -

GOVH 0.621 - -

Source: Estimation Result ***, **, and * show the level of significance at

1%, 5 %, and 10%.

Second, the estimation result at lag 2 give the

different result with lag 1. The value of F statistic for

education is 16.138 and for health is 7.310 which are

significant at 1 percent. This explain that the two

variables have the relationships with poverty up two

the second last period or 6 months has the significant

impact compared to first quarter. The conclusion is

that education and poverty have bidirectional

relationship or Granger causality. While health and

poverty have one direction relationship or

unidirectional.

Third, lag 3 shows that the optimal lag explain

that the variable of education has the effect on

poverty with F value is 2.892 and significant at 5

percent. In contrast to health and poverty, the

causality relationship was not happen. This means

that the allocation of education fund has the impact

on poverty at quarter 3 previously

Finally, lag 4 is the effect of education allocation

fund period -1 year on poverty. The result shows the

same estimation of causality of lag 1 where there are

no causality between education, health, and poverty.

4 CONCLUSION

Education and health are the basic need that are the

right of the people, but the problem is not all these

things are fulfilled especially for the poor people.

The estimation results dynamically show the

allocation of education fund effect poverty but it is

not immediately but needs time from 6 to 9 months.

But, the allocation of health fund has the effect on

the poverty but in 6 months. The allocation of

education and health funds by the government show

there is ineffectiveness in reducing poverty. Thus,

the government needs to increase the effectiveness

of the fund. Then, the share of health fund from

national budget (APBN) is needed to be increased

for the poor people.

ACKNOWLEDGEMENTS

The researchers would like to thank the Dean,

Lecturers, and also the student of Faculty of

Economics and Business of Syiah Kuala University

for the support of this study.

REFERENCES

Arfa, I. J. (2016). The Effect of Trade and Foreign Direct

Investment on Economic Growth in Indonesia: An

Application of Autoregressive Distributed Lag

Model. Thesis. Syiah Kuala University.

Atmanti, H. D. (2005). Investment in Human Resources

via Education. Jurnal Dinamika Pembangunan, 30-

39.

Azwar, A. (2004). Kecenderungan Masalah Gizi dan

Tantangan di Masa Datang. Pertemuan Advokasi

Program Perbaikan Gizi Menuju Keluarga. Jakarta.

Ehrenberg, R. G. (2012). Modern Labor Economics,

Theory and Public Policy, 11th Edition. New York:

Prentice-Hall.

Gujarati, D. (2013). Basic Econometrics, Fourth Edition.

Jakarta: Erlangga.

Muhi, A.H. (2000). Analisis Investasi Modal Manusia

Dalam Perspektif Pendidikan dan Pelatihan.

Rosadi, D. (2012). Econometrics and Applied Time Series

Analysis Using E-Views. Jakarta: Andi Publishing.

Saifuddin, M.S. (2017). Analysis of Causality of Revenue

and Expenditure of the Government in Aceh

Province Using Granger Causality Method. Thesis.

The Causality between Education and Health Funds Allocation to Poverty in Indonesia

495

Syiah Kuala University.

Syarifuddin (2017). The Impact of Fiscal Policy on

Poverty in Aceh Province. Dissertation. Syiah Kuala

University.

UNICEES 2018 - Unimed International Conference on Economics Education and Social Science

496