Analysis of the Cointegration and Causality

Relationship to Soybean Imports in Indonesia

Heffi Christya Rahayu1, Etty Puji Lestari2,
Tri Kurniawati Retnaningsih3, Anastasia Siti Nurhayati4
Faculty Economic, Universitas Pasir Pengaraian, Riau, Indonesia1,
Faculty Ekonomy, Universitas Terbuka, Indonesia2,3,4

{[email protected], [email protected], [email protected],

[email protected]}

Abstract. Soybeans are a strategic commodity in Indonesia, as soybeans are one

of Indonesia's main food crops. Demand for soybeans is very high, therefore
Indonesia has to import soybeans. The data used is time series data obtained
from the Central Statistics Agency, the Ministry of Agriculture, FAO and the
Ministry of Trade. The aim of this analysis is to evaluate the impact of soybean
productivity factors, domestic soybean prices, soybean production and soybean
harvested area on soybean imports in Indonesia. The data analysis method used
in this research is the econometric analysis method with VECM (Vector Error
Correction Model). The independent variables in the study used were soybean
productivity, soybean production, domestic soybean prices, soybean harvested
area, and using the dependent variable volume of soybean imports in Indonesia.
The data used in this study are secondary data from time series with a period of
41 years in the period 1980-2020.

Keywords: Soybean Productivity; Soybean Production; Domestic Soybean

Prices; Soybean Harvested Area; Soybean Import Volume

1 Introduction

Soybeans have been growing in demand from year to year. For decades, the difference
between soybean production and soybean demand in Indonesia has led to reliance on imports
of soybeans (Hasan et al, 2015). Based on the Ministry of Agriculture's 2019 soybean outlook
data, the national soybean production for the 2014-2018 period tended to fluctuate and grew
an average of 10.97% per year, but experienced a significant decline in 2016 and 2017 by
10.75% and 37.33%. A sharp increase is estimated to occur in 2018 of 79.66%, from 2017
production of 538.73 thousand tons to 967.87 thousand tons in 2018. The increase in
production is a positive excess of the increase in soybean harvest area in 2018 of 91.23% or
covering an area of 324.58 thousand hectares, from 2017 amounting to 355.80 thousand
hectares to 680.38 thousand hectares in 2018. Even though soybean productivity in 2018
decreased by 6.01% from 2017 amounting to 15.14 ku / ha to 14.23 ku / ha, but did not cause
a decrease in production because it was supported by a high increase in harvested area. In the
last five years, soybean productivity has tended to be stagnant, slightly increasing by 0.25%

ICLSSEE 2021, March 06, Salatiga, Indonesia

Copyright © 2021 EAI
DOI 10.4108/eai.6-3-2021.2305933
per year.
The increase in national soybean production for the 2014-2018 period is a positive effect
of a significant increase in harvested area outside Java, namely 27.71% per year and only
3.08% per year in Java. This also shows that several areas outside Java such as West Nusa
Tenggara, South Sulawesi, Aceh and Lampung have the potential to become soybean centers.
As we know so far the main centers of soybeans are still concentrated in the regions of Central
Java, East Java and West Java with a contribution to national soybean production of 58.19%.

2 Literature Review

2.1 Soybean Imports

Soybean imports in Indonesia are influenced by per capita income, the real exchange rate
of the rupiah against the US dollar and the price of imported soybeans. According to the
theory in Case and Fair (2004: 382), suggests that the level of imports is a function of income.
When a country's income increases, people will buy more of everything. This means that if
income increases, imports tend to increase. According to the theory in Mankiw (2008: 135),
states that the main difference between international transactions and domestic transactions
concerns the currency exchange rate. When people in different countries buy and sell from
each other, currency exchanges are inevitable. The exchange rate is the price of the currency
of a nation denominated in the currency of another country. Exchange rates are differentiated
by economists into two: nominal exchange rates and actual cusps.

2.2 Soybean Productivity

Soybean productivity is the amount of soybean production per certain area during a
production period, measured in tonnes per hectare which is used to manufacture or
manufacture materials consumed, for example in making, tofu, tempeh, etc. According to
botanists, soybeans are a plant originating from Manchuria and parts of China, and there are
wild types of soybeans belonging to the Glycine ussuriensis species. Then it spreads to the
tropics and subtropics and breeding is carried out so that various types of superior soybeans
are cultivated (Koswara, 2001).

2.3 Domestic Soybean Prices

According to Alex S Nitisemito (2007), price is the value of a product or service as

determined by a quantity of money where an individual or business is able to release products
or services owned by other parties based on this value. And according to Anindita (2004: 68),
the price is a sum of money required to satisfy the trade (and probably some goods). For goods
accompanied by services, trade may also be performed. Sukirno (2003: 75) said that ombining
supply and demand reveals how to assess prices through the interaction between buyers and
sellers. The high and low prices dictated by market supply and demand factors can be
interpreted this way.
Basically, this theory states that the value of a good is determined by the amount of labor
used to produce the goods. The more labor used to produce an item, the more expensive the
item is. A country experiences absolute loss or is less efficient in producing a good than other
countries. However, there are still advantages from each country, if both of them specialize in
the production and export of a good that experiences the smallest absolute loss (comparative
advantage) and exchanges a portion of its output for other goods Salvatore (1997: 4) with
various price levels with the assumption of ceteris paribus. Demand theory analyzes the
relationship between price and the quantity of goods requested. If the price of the goods
requested increases, the demand for the quantity of an item will decrease, assuming ceteris
paribus.

2.4 Soybean Production

Soybean production is the sum of the results of the activity of producing soybeans in units
of million tons. According to Rahim et al (2007), before the production process is carried out
on agricultural land, the process of procurement of agricultural inputs (production facilities) is
carried out in the form of the agro-chemical industry (fertilizers and pesticides), the agro-
automotive industry (agricultural machinery and equipment), and hatchery and nursery
industry. The performance of the Indonesian soybean harvest area from 1980 to 2013
fluctuated but tended to increase with an average increase rate of 0.62% per year. Indonesia's
soybean production centers are located in seven provinces, contributing 87.40% to national
soybean production over the last five years, and 27 other provinces have contributed 12.60%
(Ministry of Agriculture, 2015).

2.5 Soybean Harvested Area

Soybean harvested area is the land used for soybean farming as a whole in units of million
hectares. Land is defined as the physical environment consisting of climate, relief, soil, water,
and vegetation, as well as objects on it, as long as the use of land, including the effects of
human activities past and present, is affected. The effect of soybean land area on soybean
production in Indonesia is due to the fact that land area is an important input factor in
agricultural business. With the expansion of planting land, there will be an increase in
agricultural production. The increase in land area coupled with the application of cultivation to
the land can increase farmer productivity. So that the more land area farmers use, the soybean
production produced by farmers will also increase.

3 Methods

This research uses case studies in Indonesia and the data used for this research is data
sourced from related sources. This study aims to examine the effect of micro-economic
variables, namely soybean productivity, domestic soybean prices, soybean production and
soybean harvested area on the volume of soybean imports in Indonesia. The data to be used in
this research is secondary annual time series data (time series) from 1980 to 2020 from several
related sources. Data analysis techniques or econometric models define statistical relationships
between variables in a particular phenomenon.
The analytical methods used in this research are the Co-integration Test and the Vector
Error Correction Model (VECM) to see the short-term and long-term relationships in one time
series data. In this study, we took the cointegration test and the Vector Error Correction Model
(VECM) to examine the short and long term relationships between the microeconomic
variables of soybean imports, soybean production, soybean productivity, domestic soybean
prices, and soybean harvested area.
4 Results and Discussion

4.1 Descriptive Statistics

The results of descriptive statistics are as listed in table 4.1 as follows:

Table 1. The Results of Descriptive Statistics

IK HKD LPK PDK PVK
Mean 1293723. 4588.463 857.8180 1029.926 12.40756
Median 1220120. 3766.000 678.8500 935.1900 12.18000
Maximum 2671914. 12000.00 1665.710 1869.710 17.52000
Minimum 72465.00 350.0000 459.1200 521.3900 8.380000
Std. Dev. 727383.3 4124.990 343.2685 351.2167 2.343334
Skewness 0.284142 0.569289 0.743613 0.685996 0.282733
Kurtosis 2.258006 1.713873 2.163819 2.448804 2.556934
Jarque-Bera 1.492232 5.040411 4.973032 3.734725 0.881602
Probability 0.474205 0.080443 0.083199 0.154531 0.643521
Sum 53042629 188127.0 35170.54 42226.98 508.7100
Sum Sq. Dev. 2.12E+13 6.81E+08 4713331. 4934128. 219.6486
Observations 41 41 41 41 41

Based on table 4.1, it can be seen that the mean value of soybean import volume (IK) is
1293723, the maximum value is 2671914, the minimum value is 72465, and the standard
deviation value is 727383.3. The mean value of domestic soybean price (HKD) is 4588,463, a
maximum of 12000, a minimum value of 350, a value and a standard deviation of 4124.99.
The mean value of soybean harvested area (LPK) is 857,818, the maximum is 1665.71, the
minimum value is 459.12, the value and standard deviation value is 343,2685. The mean value
of soybean production (PDK) is 1029,926, the maximum value is 1869,710, the minimum
value is 521.39, and the standard deviation value is 351,2167. The mean value of soybean
productivity (PVK) was 12.40756, the maximum value was 17.52, the minimum value was
8.38, and the standard deviation value was 2.343334.

4.2 Unit Root Tests

From the results of the Augmented Dickey-Fuller (ADF) Unit Root test, it can be seen that
the data for ΔIK, ΔHKD, ΔLP), ΔPDK, and ΔPVK are data that do not contain unit roots at
the level or are stationary at the first difference level. It can be seen that ADF
statistics for each variable is smaller than =0,05, this means rejecting the hypothesis 0,
which means that the unit root does not exist and the data is stationary.

4.3 The Optimal Lag Length Test

The results of the optimal lag length test are as listed in table 4.3 as follows:

Table 2. The Results of the Optimal Lag Length Test

VAR Lag Order Selection Criteria
Lag LogL LR FPE AIC SC HQ
0 -1176.688 NA 1.46e+23 67.52504 67.74723* 67.60174
1 -1149.899 44.39277* 1.34e+23 67.42282 68.75598 67.88303
2 -1131.087 25.80039 2.09e+23 67.77638 70.22050 68.62009
 Lag LogL LR FPE AIC SC HQ
3 -1114.231 18.30010 4.30e+23 68.24179 71.79687 69.46900
4 -1073.238 32.79443 3.04e+23 67.32790 71.99394 68.93862
5 -1005.576 34.79748 8.88e+22* 64.89008* 70.66709 66.88431*
* indicates lag order selected by the criterion
LR : sequential modified LR test statistic (each test at 5% level)
FPE : Final prediction error
AIC : Akaike information criterion
SC : Schwarz information criterion
HQ : Hannan-Quinn information criterion

Based on Table 4.3, it is known that there are three criterions that determine the optimal
lag at lag five, among others, the criterions are the Final Prediction Error (FPE), Akaike
Information Criterion (AIC), and Hannan-Quinn Information Criterion (HQ). For sequential
modified LR test statistic (LR) determines the optimal lag at lag one, and Schwarz Information
Criterion (SC) determines the optimal lag at zero lag. From these results, it can be concluded
in this study using lag five, because most of the criterion chose lag five to be used in the study.

4.4 The Granger Causality Test

The results of the Granger Causality test is known that those that have a causal relationship
are those that have a probability value that is smaller than =0,05, so that H0 is rejected,
which means that a variable will affect other variables.

4.5 Cointegration Rank Test

The results of the Cointegration Rank Test are as listed in table 4.5 as follows:

Table 3. The Results of the Cointegration Rank test

Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized Max-Eigen 0.05
No. of CE(s) Eigenvalue Statistic Critical Value Prob.**
None * 0.956800 113.1086 38.33101 0.0000
At most 1 * 0.852094 68.80243 32.11832 0.0000
At most 2 * 0.657078 38.52904 25.82321 0.0006
At most 3 * 0.549285 28.68910 19.38704 0.0017
At most 4 * 0.355391 15.80800 12.51798 0.0136
Max-eigenvalue test indicates 5 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values

Based on the econometric analysis above, it can be seen that among the five variables in
this study, there are five cointegration at the 0.05 significance level. Thus, the results of the
cointegration test indicate that among the movements of HKD, IK, LPK, PDK, and PVK have
a stability / balance relationship and the similarity of movement in the long term. In other
words, in any short-run period, all the variables tend to adjust to one another, to reach their
long-run equilibrium.
4.6 The Vector Error Correction Model (VECM)

The VECM estimation results will obtain a short-term and long-term relationship between
soybean import volume (IK), domestic soybean price (HKD), soybean harvested area (LPK),
soybean production (PDK), and soybean productivity (PVK). In this estimate, soybean import
volume (IK) is the dependent variable, while the independent variables are domestic soybean
price (HKD), soybean harvested area (LPK), soybean production (PDK), and soybean
productivity (PVK). The data analysis technique or the econometric model defines the
statistical relationship between the variables using the selected lag which is 5, then the lag in
the cointegration test is 4 (minus 1 because the variable is differentiated).
The VECM estimation results are used to analyze the short-term and long-term effects of
the dependent variable on the independent variable. The results of the VECM Soybean Import
Volume (IK) in the short term are presented in table 4.6 as follows:

Table 4. The results of the VECM Soybean Import Volume (IK) in the short term
Variabel Koefisien t. statistik
CointEq1 0.31862 0.9177
D(IK(-1)) -0.673217 -1.92459
D(IK(-2)) -0.330461 -0.8941
D(IK(-3)) -0.099398 -0.36105
D(IK(-4)) -0.238129 -0.90674
D(HKD(-1)) -13.66676 -0.09043
D(HKD(-2)) 68.9778 0.39231
D(HKD(-3)) 59.2904 0.44017
D(HKD(-4)) -438.4538 -2.3715
D(LPK(-1)) -1654.561 -0.36779
D(LPK(-2)) 68.9778 0.39231
D(LPK(-3)) 59.2904 0.44017
D(LPK(-4)) -438.4538 -2.3715
D(PDK(-1)) 2368.096 0.53598
D(PDK(-2)) 3987.323 1.00804
D(PDK(-3)) 3579.959 0.85333
D(PDK(-4)) 9208.14 1.91301
D(PVK(-1)) -488817.8 -1.19084
D(PVK(-2)) -165940.6 -0.40989
D(PVK(-3)) 73245.91 0.19096
D(PVK(-4)) -434671.8 -1.01222
C 185271.9 0.87639

Based on the results presented in Table 4.6, in the short term there are no significant
variables. This can be seen from the t statistical value which is smaller than t table, so it can be
concluded that in the short term domestic soybean prices (HKD), soybean harvested area
(LPK), soybean production (PDK), and soybean productivity (PVK) have no significant effect.
to the volume of imports of soybeans (IK).

4.7 The Variant Decomposition Test

The results of the Variant Decomposition test is a summary of the results of the FEDV
analysis for the volume of soybean imports (IK) from the shocks given by each variable,
including itself. The FEDV analysis in Table 4.10 states that in the short term, i.e. year 3 the
shock to itself results in 78.06827% fluctuations in the volume of soybean imports, shocks to
domestic soybean prices result in 15.48477% fluctuations in soybean import volume, shocks
to soybean harvested area (LPK) resulted in 4.179580% fluctuation in soybean import volume
(IK), shocks to soybean production (PDK) resulted in 1.879516% fluctuation in soybean
import volume (IK), and shocks to soybean productivity (PVK) resulted in 0.387863%
fluctuation in soybean import volume (IK) . On the other hand, in the long term, namely in
year 10: shocks to itself result in 53.47788% fluctuations in soybean import volume, shocks to
domestic soybean prices result in 20.83482% fluctuations in soybean import volume, shocks
to soybean harvested area (LPK) result in 17.33145 % fluctuation in soybean import volume
(IK), shock to soybean production (PDK) resulted in 8.059311% fluctuation in soybean import
volume (IK), and shock to soybean productivity (PVK) resulted in 0.296526% fluctuation in
soybean import volume (IK).

4.8 Impulse Response

The results of the Impulse Response test are as listed in the table as follows:
Response to Cholesky One S.D. Innovations
Response of IK to IK Response of IK to HKD Response of IK to LPK Response of IK to PDK Response of IK to PVK
4 00,0 00 4 00 ,00 0 40 0,00 0 4 00 ,000 40 0,00 0

3 00,0 00 3 00 ,00 0 30 0,00 0 3 00 ,000 30 0,00 0

2 00,0 00 2 00 ,00 0 20 0,00 0 2 00 ,000 20 0,00 0

1 00,0 00 1 00 ,00 0 10 0,00 0 1 00 ,000 10 0,00 0

0 0 0 0 0

-1 00,000 -1 00 ,00 0 - 10 0,00 0 -1 00 ,000 - 10 0,00 0

-2 00,000 -2 00 ,00 0 - 20 0,00 0 -2 00 ,000 - 20 0,00 0

2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10

Response of HKD to IK Response of HKD to HKD Response of HKD to LPK Response of HKD to PDK Response of HKD to PVK
1,200 1 ,20 0 1,20 0 1 ,200 1,20 0

800 80 0 80 0 800 80 0

400 40 0 40 0 400 40 0

0 0 0 0 0

-4 00 - 40 0 - 40 0 - 400 - 40 0

-8 00 - 80 0 - 80 0 - 800 - 80 0
2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10

Response of LPK to IK Response of LPK to HKD Response of LPK to LPK Response of LPK to PDK Response of LPK to PVK
200 20 0 20 0 200 20 0

100 10 0 10 0 100 10 0

0 0 0 0 0

-1 00 - 10 0 - 10 0 - 100 - 10 0

-2 00 - 20 0 - 20 0 - 200 - 20 0
2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10

Response of PDK to IK Response of PDK to HKD Response of PDK to LPK Response of PDK to PDK Response of PDK to PVK
300 30 0 30 0 300 30 0

200 20 0 20 0 200 20 0

100 10 0 10 0 100 10 0

0 0 0 0 0

-1 00 - 10 0 - 10 0 - 100 - 10 0

-2 00 - 20 0 - 20 0 - 200 - 20 0
2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10

Response of PVK to IK Response of PVK to HKD Response of PVK to LPK Response of PVK to PDK Response of PVK to PVK
.4 .4 .4 .4 .4

.3 .3 .3 .3 .3

.2 .2 .2 .2 .2

.1 .1 .1 .1 .1

.0 .0 .0 .0 .0

-.1 - .1 - .1 -.1 - .1

-.2 - .2 - .2 -.2 - .2
2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10 2 4 6 8 10

Fig. 1. The Results of the Impulse Response Test

The plot results of the IRF can be seen in Figure 4.1 that there are 25 IRF plots for the next
10 years, which visually explains the response of a variable that arises because of a shock /
impulse of 1 standard deviation, both from itself. or other variables.

5 Conclusions and Recommendations

5.1 Conclusions

Based on the results and discussion, it can be concluded that:

a. Based on the optimal lag length test, the best model is obtained for data on soybean import
volume (IK), domestic soybean price (HKD), soybean harvested area (LPK), soybean
production (PDK), and soybean productivity (PVK).
b. Based on structural analysis, it can be concluded that the response of each variable to
shocks originating from itself is quite significant, because it is fluctuating. And the
responses from other variables are generally not significant to soybean productivity shock
 (PVK). On the other hand, the response of soybean productivity (PVK) to the shock of the
variable is very significant. Also In general, for future analysis, both in the long term,
domestic soybean price (HKD), soybean harvested area (LPK), soybean production
(PDK), and soybean productivity (PVK) have a significant effect on the volume of
imported soybeans (IK).

5.2 Recomendations

It should be noted that the volume of soybean imports (IK) is a policy issued by the
Government which is adjusted to the economic conditions that occur in Indonesia, especially
with regard to food security. Therefore, it is necessary to conduct an analysis in order to
strengthen the food economy, including examining the effects of the domestic soybean price
(HKD), soybean harvested area (LPK), soybean production (PDK), and soybean productivity
(PVK) on the volume of soybean imports (IK). By testing these four independent variables,
the government can predict the relationship between domestic soybean production factors and
soybean imports, which in the long run is expected to strengthen food security and food self-
sufficiency in Indonesia, especially the food crop sector, the soybean sub-sector. These are
economic events that can be used as references by various parties, either by the government or
the private sector, to be used in planning and decision making in the future.

References

[1] Anindita, R. 2004. Pemasaran Hasil Pertanian. Surabaya: Papyrus.

[2] Case, Karl E. dan Ray C. Fair. 2004. Prinsip-Prinsip Ekonomi Makro. Edisi. Kelima,
Cetakan Kesatu. Jakarta: PT. Indeks. Chuang YC
[3] Iswandari, D. A. 2018. Analisis Faktor-faktor yang Mempengaruhi Impor Kedelai di
Indonesia Tahun 1977 – 2015. Diakses dari https://docplayer.info/87090505-Analisis-
faktor-faktor-yang mempengaruhi-impor-kedelai-di-indonesia-tahun.html
[4] Kementerian Perdagangan. 2015. Analisis Outlook Pangan 2015 – 2019. Jakarta: Pusat
Kebijakan Perdagangan dalam Negeri, Badan Pengkajian dan Pengembangan
Kebijakan Perdagangan.
[5] Kotler, Philip. 1997. Manajemen Pemasaran. Jakarta: Prenhallindo.
[6] Mankiw, N. G. 2008. Pengantar Ekonomi Makro. Jakarta: Salemba Empat.
[7] Maryam. 2002. Evalusai Kesesuaian Lahan Untuk Pemukiman Melalui Pemanfaatan
Sistem Informasi Geografis Di Kota Semarang. Skripsi. h. 12
[8] Pusat Data dan Sistem Informasi Pertanian Kementerian Pertanian. 2015. Outlook
Komoditas Pertanian Tanaman Pangan. Diakses dari
http://epublikasi.setjen.pertanian.go.id/epublikasi/outlook/2015/Tanaman%20Pangan/O
utlook%20Kedelai%202015/files/assets/common/downloads/Outlook%20Kedelai%20
2015.pdf
[9] Rahim. Abd. Dan Hastuti. DRW. 2007. Ekonomi Pertanian. Jakarta : Penebar Swadaya
[10] Salvatore, Dominick. 2004. Theory and Problem of Micro Economic Theory. 3rd
Edition. Alih Bahasa oleh Rudi Sitompul. Jakarta: Penebit Erlangga.
[11] Sukirno, S. 2004. Makroekonomi, Teori Pengantar. Jakarta: PT Raja Grafindo Persada.