IOP Conference Series: Earth and

Environmental Science

PAPER • OPEN ACCESS You may also like

- Study on comparison of the advantages of
Factors influencing the soybean import volume in different soybean producing areas in
China from a temporal and spatial
North Sumatera Province perspective
Yu Wang, Shiwei Xu, Ganqiong Li et al.

- Adoption Determinants of Biofertilizer

To cite this article: T Supriana et al 2019 IOP Conf. Ser.: Earth Environ. Sci. 260 012003 Technology for Soybean in Rainfed Area
Dian Adi Anggraeni Elisabeth, Siti
Mutmaidah and Arief Harsono

- Study of Tempe Production from Dried

Peeled Soybeans
View the article online for updates and enhancements. Hasbullah and D Silvy

This content was downloaded from IP address 182.0.97.132 on 09/05/2024 at 05:31

AEFS 2018 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 260 (2019) 012003 doi:10.1088/1755-1315/260/1/012003

Factors influencing the soybean import volume in North

Sumatera Province

T Supriana*, M A Harahap and S K H Nasution

Agribusiness Department, Faculty of Agriculture, Universitas Sumatera Utara, Medan,
Sumatera Utara, Indonesia.

E-mail: *[email protected] or [email protected]

Abstract. Imports become an alternative policy to meet the domestic soybean demand. North
Sumatera Province has a high dependence on soybean imports because the amount of soybean
imported is greater than the amount of domestic production. The purpose of this study was to
analyse the influencing factors of the soybean import volume in North Sumatera Province. The
Engle-Granger Cointegration Test and the Error Correction Model (ECM) methods used to
identify the factors influencing the soybean import volume in the long-term and the short-term.
The results concluded that the soybean import was significantly influenced by the domestic
soybean production, the exchange rate, and the soybean price ratio. In the long-term, the soybean
price ratio significantly influences the soybean import volume. While in the short-term, exchange
rate and soybean price ratio significantly influence the soybean import volume.

1. Introduction
Soybean is one of the food crops that have high economic value. The development of soybeans has been
carried out intensively as it relates to various sectors. Soybean is processed into various foods such as
tofu, tempe, tauco, and soy sauce. It is the main food for most people in Indonesia, including people in
North Sumatera Province. However, the increase in soybean consumption is not followed by the increase
in domestic soybean production. It makes North Sumatera Province has to import soybean to fulfil
domestic soybean demand.
The amount of soybean production in North Sumatera Province tends to fluctuate. The production
had increased in 2008 with the acceleration program of soybean production that is Program Bangkit
Kedelai Nasional (PROKEMA). The program can increase domestic soybean production to touch 11
thousand tons. However, in 2012 production fell significantly to touch 5 thousand tons only. It is due to
the decline in the amount of harvested area and productivity because of the farmers' lack of interest in
soybean cultivation. In contrast, the consumption of soybeans in North Sumatera Province actually
increased every year to touch 65 thousand tons in 2016.
Imports become one of the inevitable ways to meet domestic soybean needs. The number of soybean
imports continued to increase from 2006 to 2016. It reached more than 169 thousand tons in 2016 [1].
This condition further widened the gap between soybean production and imports in North Sumatera
Province. The increasing of the soybean import volume in North Sumatera Province will have an impact
on food security and the farmers' interest to cultivate soybean. Therefore, this research is needed to
analyse the factors influencing the soybean import volume in North Sumatera Province.

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
AEFS 2018 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 260 (2019) 012003 doi:10.1088/1755-1315/260/1/012003

2. Materials and methods

2.1. Method for determining the research areas

This research was conducted in North Sumatera Province. The area of study was determined
purposively, with the consideration that North Sumatera Province experienced an increase in the volume
of imported soybeans, due to increased consumption of soybeans, and it was not able to be balanced
with the amount of domestic soybean production.

2.2. Data analysis method

2.2.1. Unit Root test. Unit root test or stationary test is aimed to know stationary data. False or spurious
regression will be generated if the data is not stationary. Unit root test was performed using the
augmented Dickey-Fuller (ADF) method, with the following hypothesis: Ho; there is a unit root (the
data is not stationary) and Ha; there is no root unit (the data is stationary). If the test at the level is not
stationary, then re-testing is required by conducting unit root test on the first difference of the series.

2.2.2. Ordinary Least Square test. The regression model must show a valid relationship equation or a
BLUE (Best Linear Unbiased Estimator). The model must meet the basic assumptions of the classical
Ordinary Least Square (OLS) [2].
Normality Test Criteria using the Jarque-Bera method is used to see whether the data is normally
distributed or not, with the following hypothesis: Ho; residuals are normally distributed and Ha;
residuals are not normally distributed. If the probability value of Jarque-Bera > 5% then do not reject
H0 which means the data is normally distributed and vice versa.
Criteria for autocorrelation test using LM method (Bruesch-Godfrey method) using hypothesis as
follows: Ho; there are no autocorrelation symptoms and Ha; there are autocorrelation symptoms. If the
critical area p-value Obs * R-squared > 5%, then accept H0, so there are no suspected autocorrelation
symptoms.
Multicollinearity means the existence of a perfect or definitive linear relationship, among some or all
of the explanatory variables of the regression model [2]. The method used to detect multicollinearity is
the partial correlation between independent variables (rule of thumb) method. The method is if the value
of correlation coefficient is high enough, that is above 0.8, then we can guess that the regression model
has multicollinearity disorder [3].

2.2.3. Cointegration test. Cointegration is a long-term relationship between variables that although
individually are not stationary, linear combinations between them can be stationary by transformation.
Cointegration test in this research using the Engle-Granger Cointegration Test method. The Engle-
Granger Cointegration Method using the following hypothesis: Ho; ADF-statistics > MacKinnon critical
value (no cointegration occurs) and Ha; ADF-statistics < critical value of MacKinnon (cointegration
occurs). The long-term equation model used in this research are:

ln 𝑉𝐼𝑡 = 𝛼0 + 𝛼1 ln 𝑄𝐷𝑡 + 𝛼2 ln 𝑁𝑇𝑡 + 𝛼3 ln 𝑅𝑃𝑡 + 𝑈𝑡 (1)

Where ln VIt represent the total soybean import volume of North Sumatera Province in period t; ln QDt
represent the total domestic soybean production in period t; ln NTt represent the Rupiah (IDR) exchange
rate against the US Dollar (USD) in period t; ln RPt represent the ratio of domestic soybean price to the
world in period t; α0 is the intercept; and Ut is the error distribution in period t.

2
AEFS 2018 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 260 (2019) 012003 doi:10.1088/1755-1315/260/1/012003

2.2.4. Error Correction Model (ECM). The estimation results on unit root and cointegration testing can
be used to estimate the model using ECM, as in the following equation:

D ln 𝑉𝐼𝑡 = 𝛼0 + 𝛼1 D ln 𝑄𝐷𝑡 + 𝛼2 D ln 𝑁𝑇𝑡 + 𝛼3 D ln 𝑅𝑃𝑡 + 𝜆𝑈𝑡−1 + 𝑒𝑡 (2)

Where D is the first difference; λ represent Error Correction Term (ECT); and et = error distribution in
period t. To find out if the model specification with ECM is a valid model then tested the coefficient of
Error Correction Term (ECT). If the test results on the ECT coefficient are significant, then the model
specification observed is valid.

2.2.5. Goodness of fit. The coefficient of determination test (R²) is done with the aim to see whether the
new variables of the independent variables in the model able to explain its influence to the dependent
variable. The coefficient of determination shows the ability of the regression line to explain the variation
of the dependent variable (percent proportion) which can be explained by the independent variable. R²
values range from 0 to 1. The closer to 1, the better it is.
The F-statistic test is a test used to determine whether the independent variables simultaneously have
a significant or insignificant influence on the dependent variable, with the following hypothesis; Ho:
independent variables simultaneously have no significant influence on the dependent variable; and Ha:
independent variables simultaneously have a significant influence on the dependent variable. If β1, β2,
β3 = 0 (prob F statistically significant at α = 5% or F-statistic < F-table), so Ho is accepted; if β1, β2, β3
≠ 0 (prob F statistically insignificant at α = 5% or F-statistic > F-table) so Ha is accepted.
The t-statistic test is done to know the significance of each independent variable in influencing the
dependent variable. The test is done by determining the value of t table then comparing t table value and
t-statistics value. The criteria in t-test are: Ho accepted if t-count < t-table table means that the
independent variables partially have no significant influence on the dependent variable and Ha accepted
if t-count > t-table means that the independent variables partially have a significant influence on the
dependent variable.

3. Results and discussion

Unit root test results at the level indicate that all variables are not stationary (Table 1). It can be seen
from the t-statistical value of ADF greater than the critical value of MacKinnon at the level of 1%, 5%,
and 10%.

Table 1. The unit root test results with the augmented Dickey-Fuller at level
Critical Value of MacKinnon
Variables Value of ADFt Inference
1% 5% 10%
Ln VI -1.942222 -3.788030 -3.012363 -2.646119 not stationary
Ln QD -1.866922 -3.788030 -3.012363 -2.646119 not stationary
Ln NT -2.809568 -3.788030 -3.012363 -2.646119 not stationary
Ln RP -2.069512 -3.788030 -3.012363 -2.646119 not stationary

Since all data has not been stationary at the level, then testing should be continued by conducting
unit root test at first difference. As shown in Table 2, all variables have been stationary at the first
difference. It can be seen that the t-statistical value of the ADF is smaller than the critical value of
MacKinnon with varying degrees of probability.
Normality: Based on the results of data processing on the model, obtained the probability value of
0.907 > 0.05, then Ho accepted. Thus, it can be concluded that the residual is normally distributed.
Autocorrelation: Based on the result of data processing with LM (Bruesch Godfrey) test, known the p-
value Obs * R-squared of 0.0985 > 0.05, then Ho is accepted. Therefore, it is free from the problem of
autocorrelation. Multicollinearity: In the regression equation model of the factors influencing the

3
AEFS 2018 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 260 (2019) 012003 doi:10.1088/1755-1315/260/1/012003

soybean import volume in North Sumatera Province, all independent variables have a relationship value
smaller than 0.8, meaning the independent variables are free from multicollinearity.

Table 2. The unit root test results with the augmented Dickey-Fuller at the first difference
Critical Value of MacKinnon
Variables Value of ADFt Inference
1% 5% 10%
Ln VI -4.941761 -3.808546 -3.020686 -2.650413 stationary
Ln QD -5.206172 -3.831511 -3.029970 -2.655194 stationary
Ln NT -4.977889 -4.498307 -3.658446 -3.268973 stationary
Ln RP -4.545365 -4.498307 -3.658446 -3.268973 stationary

Cointegration testing with the Engle-Granger Cointegration method on the regression equation shows
that the ADF-statistic value for residual cointegration equation is less than the critical value of
MacKinnon both at the level of 1%, 5%, and 10% (Table 3). The Engle-Granger cointegration test results
show that the soybean import volume (VI), the domestic soybean production (QD), the exchange rate of
IDR to USD (NT), and the ratio of domestic soybean price to world soybean price have long-term
balance.

Table 3. The augmented Dickey-Fuller test results on residual Error Correction Term (ECT) equation
Critical Value of MacKinnon
Variable Value of ADFt Inference
1% 5% 10%
ECT -4.437371 -2.66 -1.95 -1.60 stationary

The Long-term estimation of soybean import volume factors in North Sumatera Province shows R-
square (R²) value of 0.832 (83.2%) (Table 4). It explains that the variables used in the model can explain
the diversity of factors as much as 83.2%, while the remaining 16.8% is explained by other factors not
included in the model.

Table 4. The results of estimation of the influencing factors of the soybean import volume in
North Sumatera Province in the long-term
Variables Coefficient Probability
C -24.50287 0.2651
Ln QDt -0.824215 0.4658
Ln NTt -5.063756 0.1186
Ln RPt 9.971803 0.0074
R-Squared 0.832851
Adjusted R-Squared 0.804993
F-statistic 0.000000

The estimation results of the model of factors influencing the soybean import volume in North
Sumatera Province in the long-term, are presented in Table 4, while in the long-term equation model
obtained is as follows:

ln 𝑉𝐼𝑡 = −24.5028 − 0.8242 ln 𝑄𝐷𝑡 − 5.0637 ln 𝑁𝑇𝑡 + 9.9718 ln 𝑅𝑃𝑡 (3)

Error Correction Model (ECM) is a technique for correcting short-term imbalances leading to long-
term equilibrium. ECM results show that the ECT coefficient value of 0.9886 indicates that the short-
term imbalance will be adjusted within 9 years 8 months (Table 5). The value of ECT coefficient of
0.0004 is significant at the level of 5% (0.0004 < 0.05). It indicates that the Error Correction Model

4
AEFS 2018 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 260 (2019) 012003 doi:10.1088/1755-1315/260/1/012003

(ECM) is valid and can illustrate the factors influencing the soybean imports volume in North Sumatera
Province, which are the domestic soybean production, the exchange rate of IDR to USD, and the ratio
of domestic soybean price to the world price.

D ln 𝑉𝐼𝑡 = 0.4536 − 0.1704D ln 𝑄𝐷𝑡−1 − 7.6216D ln 𝑁𝑇𝑡−1 + 8.4821D ln 𝑅𝑃𝑡−1 − 0.9886𝐸𝐶𝑇𝑡−1

(4)

The estimated ECM estimation value of R-square (R²) is 0.64489 (64.48%) (Table 5). It indicates
that 64.48% of the domestic soybean production, the exchange rate of IDR to USD, and the ratio of
domestic soybean price to the world price influencing the import of soybean, while the rest of 35.52%
is explained by other factors not included in the model.

Table 5. The results of estimation of the influencing factors of the soybean import volume in
North Sumatera Province in the short-term
Variables Coefficient Probability
C 0.453687 0.3039
D Ln QDt-1 -0.170467 0.8518
D Ln NTt-1 -7.621609 0.0030
D Ln RPt-1 8.482117 0.0041
ECTt-1 -0.988682 0.0004
R-Squared 0.644890
Adjusted R-Squared 0.566113
F-statistic 0.001557

The F-statistic test is used to know the influence of the independent variable simultaneously to the
dependent variable. Based on the results of Cointegration and ECM estimation, the F value of 29.896
7.2641 > F statistic at 5% significant level, then H0 is rejected. It means that the domestic soybean
production, the exchange rate of IDR to USD, and the ratio of domestic soybean price to the world price
simultaneously have a significant influence to the soybean import volume in North Sumatera Province.
The t-statistic test is used to determine the influence of each independent variable to the dependent
variable. The t-statistical test results with a significant level of 5% indicate that in the long-term variables
influencing soybean imports in North Sumatera Province is the ratio of domestic soybean price to the
world price. While in the short-term the exchange rate and the ratio of domestic soybean price to the
world price influence the soybean import volume in North Sumatera Province.

Table 6. The results of the t-statistics test

Variables Long-term Short-term
Total production not significant not significant
Rupiah exchange rate not significant significant
Price ratio significant significant

3.1. The influence of production amount on soybean import volume

In the long-term and short-term economy, domestic soybean production does not influence the soybean
import volume in North Sumatera Province. It is allegedly due to government policies in the
development of soybean agribusiness that did not show significant progress [4]. In addition, the
increasing population factor implicates to the high demand for vegetable protein through higher soybean
commodities [5].

5
AEFS 2018 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 260 (2019) 012003 doi:10.1088/1755-1315/260/1/012003

3.2. The influence of exchange rate on soybean import volume

The Long-term estimates show that the IDR exchange rate to the USD has no significant influence on
soybean import volume in North Sumatera Province. It is expected in the long-term, the population will
increase and it is supported by an increase in income and awareness of the fulfilment of proper nutrition.
Thus, this depreciation of the Rupiah will not influence the soybean import volume because the cost of
depreciation will be charged to the consumers through the price paid [5]. The IDR exchange rate against
USD has a negative coefficient. It is in accordance with the theory that if the domestic currency
depreciates, then the importer who will buy soy will require more dollars. So that importer will spend
bigger cost to import soybean. It will give consequences importers will reduce the import of soybeans.
In the short-term, the IDR exchange rate against the USD in the previous period (first lag) has a
consistent direction to hypotheses and theories and has a significant impact on the soybean imports
volume. In contrast to the long-term conditions, where the exchange rate does not significantly influence
the imports volume. It is suspected because in the short-term there is very high volatility that causes
importers (economic actors) must predict the changes in exchange rate. While in the long-term, with a
good understanding of economic development and rational action, importers suspect the long-term
exchange rate movement is relatively stable [5].

3.3. The influence of price ratios on soybean import volume

The ratio of domestic soybean price to the world price in the long-term and the short-term has a
significant influence to soybean import volume in North Sumatera Province. It means that in the long-
term, the ratio of soybean prices has a significant influence on the demand for soybean imports. So that
if the price of local soybeans rises, the demand for soybean imports will increase, so that soybean prices
are still stable. In addition, it is allegedly due to the existence of long-term contracts by exporting
countries, agreement in international trade organizations, and government policies regarding the
determination of import tariffs and unstable quota [5].

3.4. Policy implications

Firstly, the results of short-term and long-term estimation show that the influence of domestic soybean
production is not significant on the decrease of soybean import volume. Therefore, there is a need for a
real strategy, program, and support to increase soybean production and productivity so that it can hamper
the rate of soybean import in North Sumatera Province.
Secondly, there is the need for strong institutional strengthening and capital support for soybean
agroindustry so that if the rupiah exchange rate depreciates, the soybean industry business does not need
to close its business directly. The government can cooperate with financial institutions and banks to be
able to provide credit or capital assistance to soybean industry business to be able to carry out its business
activities. In addition, institutional strengthening through cooperation is very important as a channel to
credit or technology assistance. It is expected that the group/institution will together help the industry
effort so it will be easier to get input, capital, and marketing.
Thirdly, one of the incentives that farmers need to run their farms is the price. The price becomes an
important signal for farmers to cultivate soybeans. It is necessary to set a "reasonable price" mechanism,
to ensure domestic soybean prices can compete with imported soybean prices. The relatively lower
imported soybean price becomes a disincentive for farmers, especially if it is associated with a trade-off
of growing more profitable crops. If the incentive system through the price is difficult to handle by the
government, then efforts to keep farmers planting soybeans will be very difficult.
Fourth, the government can maintain the protection policy (tariff setting) which has been taken, that
is the implementation of import tariffs. With specific tariff rates and quotas, a price level created will
not compete with local soybean prices and should be combined with non-tariff policies such as
monitoring, regulation, and import restrictions.
Fifth, according to the estimation result in the short-term, the IDR exchange rate also significantly
influence the soybean import volume. Therefore, the government through the monetary authority of
Bank Indonesia should have the right mechanism to maintain the stability of the rupiah exchange rate in

6
AEFS 2018 IOP Publishing
IOP Conf. Series: Earth and Environmental Science 260 (2019) 012003 doi:10.1088/1755-1315/260/1/012003

the short and long-term. Because it will trigger domestic inflation fluctuations, and finally the domestic
inflation fluctuations will trigger domestic commodity price fluctuations.

4. Conclusions
The soybean imports volume in North Sumatera Province was significantly influenced by the domestic
soybean production, exchange rate, and the soybean price ratio. In the long-term, the soybean price ratio
has a significant influence on the soybean import volume in North Sumatera Province. While in the
short-term, the IDR exchange rate to the USD and the soybean price ratio have a significant influence
on soybean import volume. The amount of domestic soybean production has not been able to reduce the
soybean import volume in North Sumatera Province. In order to reduce the soybean imports volume, the
government need to compile strategies and programs to increase soybean production and productivity;
strengthen the institutional and capital support for soybean agroindustry; arrange the incentive systems
for viable soy farmers; maintain the policy of imposition of import tariffs; and maintain the stability of
the rupiah exchange rate.

References
[1] Badan Pusat Statistik Provinsi Sumatera Utara [BPS-Statistics of North Sumatera Province] 2017
Provinsi Sumatera Utara Dalam Angka Tahun 2017 [Sumatera Utara Province in Figures 2017]
(Medan: Badan Pusat Statistik Provinsi Sumatera Utara [BPS-Statistics of North Sumatera
Province])
[2] Gujarati D N 2004 Basic Econometrics Fourth Edition (New York: The McGraw Hill Companies)
[3] Widarjono A 2009 Ekonometrika. Pengantar dan Aplikasinya Edisi Ketiga [Econometrics. Basic
and Applications Third Edition] (Yogyakarta: EKONISIA)
[4] Asmara A and Anggi 2015 Kebijakan Swasembada Kedelai. Tantangan dan Potensi Capaian
dalam Pembangunan Pertanian Berorientasi kepada Peningkatan Kesejahteraan Masyarakat
[Soybean Self-Sufficiency Policy. Challenges and Potential Achievements in Agricultural
Development Oriented to Enhancing People's Welfare] (Bogor: IPB Press)
[5] Feryanto 2015 Faktor-faktor yang Mempengaruhi Impor Kedelai Indonesia Periode 1984-2013
[Factors Affecting Indonesian Soybean Imports for the Period 1984-2013] Seminar Nasional
Agribisnis Kedelai: Antara Swasembada dan Kesejahteraan Petani [National Seminar on
Soybean Agribusiness: Between Self-Sufficiency and Farmer Welfare] (Yogyakarta: Universitas
Gadjah Mada) pp 282-95