Faktor Penentu Ekspor Manufaktur Di Indonesia: Analisis Data Panel

DETERMINANTS OF INDONESIA’S MANUFACTURING
EXPORT: A PANEL DATA ANALYSIS

AHMAD SOHIBIL KAHFI

POSTGRADUATE SCHOOL
BOGOR AGRICULTURAL UNIVERSITY
BOGOR
2016

ii

iii

STATUTORY DECLARATION
I, Ahmad Sohibil Kahfi, hereby declare that the master thesis entitled
‘Determinants of Indonesia’s Manufacturing Export: A Panel Data Analysis’ is my
original work under the supervision of Advisory Committee and has not been submitted
in any form and to another higher education institution. This thesis is submitted
independently without having used any other source or means stated therein. Any source
of information already stated in the part of reference of this thesis. Herewith I passed the

thesis copyright to Bogor Agricultural University
Bogor, Januari 2016
Ahmad Sohibil Kahfi
H151137094

iv

RINGKASAN
Ahmad Sohibil Kahfi. Faktor Penentu Ekspor Manufaktur di Indonesia: Analisis Data
Panel. Dibimbing oleh NOER AZAM ACHSANI dan FLORIAN PLOECKL.

Nilai ekspor Indonesia terus mengalami penurunan sejak 2012. Masalah ini
menarik perhatian pemerintah Indonesia untuk segera diselesaikan dengan cara
meningkatkan performa ekspor total. Salah satu sektor yang dapat ditingkatkan adalah
sektor manufaktur. Namun pada 2014, kontribusi dari sektor manufaktur terhadap nilai
ekspor total Indonesia hanya sebesar kurang lebih 40%, dimana kontribusi tersebut harus
ditingkatkan untuk memenuhi permintaan dunia terhadap produk manufaktur. Tujuan
dari penelitian ini adalah untuk menganalisis efek dari faktor penentu untuk ekspor
manufaktur di Indonesia.
Beberapa faktor utama yang dapat meningkatkan ekspor manufaktur antara lain

adalah nilai tukar rupiah, foreign direct investment (FDI), gross domestic product
(GDP), dan kebijakan perdagangan. Faktor tersebut dianalisis menggunakan regresi data
panel dengan pendekatan random effect model. Hasil dari analisis menunjukkan bahwa
perubahan relative dari nilai tukar, foreign GDP, jarak antara dua negara, dan tariff
secara signifikan mempengaruhi ekspor manufaktur di Indonesia.
Berdasarkan hasil penelitian, terdapat beberapa rekomendasi untuk pemerintah
Indonesia untuk meningkatkan performa ekspor manufaktur. Rekomendasi pertama
adalah dengan menjaga ekspor Indonesia ke negara-negara besar dan dengan
memperluas pasar ekspor Indonesia ke negara lain. Rekomendasi kedua adalah dengan
menjaga stabilitas nilai tukar rupiah dan mendukung industry local untuk menggunakan
teknologi yang lebih maju. Kemudian, dengan mendukung penyederhanaan proses
import melalui system satu atap dan mengefektifkan penerapan tariff melalui negosiasi
FTA.
Kata kunci: produk manufaktur, ekspor, nilai tukar, FDI, foreign GDP, kebijakan
perdagangan, regresi data panel, random effect model.

v

SUMMARY
Ahmad Sohibil Kahfi. The Determinants of Indonesia’s Manufacturing Export: A Panel

Data Analysis. Under supervision of NOER AZAM ACHSANI and FLORIAN
PLOECKL.
Indonesia’s export value has been decreasing since 2012. This raised government
concern to address this issue by increasing total export performance. One sector that can
be improved in this strategy is the manufacturing sector. In 2014, the contribution of
manufacturing sector to Indonesia’s total export value was only about 40% which should
be increased to fulfill global manufacturing products’ demand. The aim of this study is
to analyse the impact of determinant factors on Indonesia’s manufacturing export.
Factors that can support Indonesia’s manufacturing export include the exchange
rate, foreign direct investment (FDI), gross domestic product (GDP) and trade policies.
The factors are examined using panel data regression with a random effect model. The
results reveal that relative change of exchange rate, foreign GDP, distance between two
countries and average tariffs significantly affect Indonesia’s manufacturing export.
Based on the results, there are recommendations for Indonesian government to
increase its manufacturing export performance. First recommendation is to maintain
Indonesia’s export to rich countries and to diversify its export market. Secondly, to
stabilise the rupiah exchange rate and to encourage industries to use more advanced
technologies. Next is to suggest simplification of import processes by promoting a
national single window system and tariffs improvement through FTA negotiation.
Keywords: Manufactures, export, exchange rate, FDI, foreign GDP, trade policies, panel

data regression, random effects model.

vi

©Copyright belongs to IPB, 2016
All rights Reserved Law
Prohibited quoting part or all of this paper without including or mentioning the
source. The quotation is only for educational purposes, research, scientific
writing, preparation of reports, writing criticism, or review an issue; and citations
are not detrimental to the interest of IPB.
Prohibited announcing and reproducing part or all papers in any form without
permission of IPB.

vii

DETERMINANTS OF INDONESIA’S MANUFACTURING
EXPORT: A PANEL DATA ANALYSIS

AHMAD SOHIBIL KAHFI

Master Thesis
As a requirement to obtain a degree
Master of Science in
Economics Program

POSTGRADUATE SCHOOL
BOGOR AGRICULTURAL UNIVERSITY
BOGOR
2016

viii

External Advisory Committee Examiner: Dr. Ir. Yeti Lis Purnamadewi, M.Sc.Agr.

ix

ACKNOWLEDGEMENT
My greatest gratefulness goes to almighty god, ALLAH SWT, who provides me
opportunities to study until I complete this dissertation. His guidance supports me to
solve all challenges during my study.

Also, I present my deepest gratitude, for my supervisors, Prof. Noer Azam
Achsani and Dr. Florian Ploeckl. Obviously, they have provided precious knowledge,
ideas, time and all other supports to help me accomplish this dissertation.
Following this, nothing that I can deliver but great thanks for my families,
especially for my parents who continuously pray to ALLAH for my success and my wife
giving her best efforts to support me finish this study.
I also want to thanks to all my colleagues from the University of Adelaide, IPB,
the Ministry of Trade Republic of Indonesia and the south road family. Thanks for
having very pleasant experiences together. I hope our friendship will be long-lasting.
At the end, I would declare that all mistakes in this study are my own.
Bogor, January 2016
Ahmad Sohibil Kahfi

x

xi

Table of Contents
LIST OF TABLES
LIST OF FIGURES

1. INTRODUCTION
2. BACKGROUND
The importance of manufacturing for the economy
Manufacturing export performance in Indonesia
3. LITERATURE REVIEW
Exchange Rate
Foreign Direct Investment
Gross Domestic Product
Trade Policies
4. METHODOLOGIES AND DATA
Methodologies
Data
5. RESULT
Finding the best approach
Impact of each determinant
6. CONCLUSION
REFERENCES
APPENDIX
BIOGRAPHY

xii
xii
1
2
2
3
4
4
5
6
7
8
8
11
11
11
12
15
17
20

21

xii

LIST OF TABLES
Table 1. Hausman’s test between FEM and REM
11
Table 2. Impact of each determinant on the Indonesia’s manufacturing export (random
effects)
12

LIST OF FIGURES
Figure 1. Indonesia’s total export value and its change from 2004 to 2014
1
Figure 2. Manufacturing export value and total export value in Indonesia from 2004 to
2014
3

1. INTRODUCTION
In 2011, Indonesia made history by doubling its export value to approximately USD

203 billion over a period of five years from 2006 (Figure 1). Generally,countries take more
than five years to double their export value,yet only Indonesia, Belgium, Russia,
Switzerland, the United States, and Brazil have successfully achieved this in a five year
period, as stated by the former Trade Minister, Gita Wirjawan (Suryanto, 2012).
However, this improvement did not continue in the following year. The economic
crisis in 2011 caused export demand from Indonesia’s trading partner countries to decline;
consequently Indonesia’s export value fell to USD 190 billion (Sukarno, 2012). Another
reason for the decline in export value was the drop in export mining commodities’ prices,
as stated by the head of Statistics Indonesia, Suryamin (Syafputri, 2013). The export value,
then,dropped consistently in the following years to approximately USD 176 billion in 2014
(Figure 1) (Ministry of Trade, 2015).
40

250.00

30
20

150.00
10

%

USD Billion

200.00

100.00
0
50.00

-10
-20

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Non oil and gas

Oil and Gas

% change of total export value (RHS)

Source: Ministry of Trade, 2015
Figure 1. Indonesia’s total export value and its change from 2004 to 2014
To address the decline in export trade, Indonesia’s government, through the Ministry
of Trade, established a target which to triple Indonesia’s export value in the five year period
from 2014 to 2019. However, the Indonesian government has also introduced a policy
prohibiting export of raw materials in order to guarantee natural resource sustainability and
develop local industries (Gunawan, 2014). Therefore, in order to meet the target, the
government aims to focus on increasing export performance in the manufacturing
sector,which has more value added thanprimary products.
As the Ministry of Trade prepares strategies to boost Indonesia’s manufacturing
export value, two questions arise which this study aims to answer:

2

1.

What are factors significantly affect the export performance of the manufacturing
sector in Indonesia?
2. How big is the impact of each factor on the export performance of the manufacturing
sector in Indonesia?
This study consists of six chapters commencing with the introduction. The second
chapter briefly describes the importance of manufacturingand manufacturing export
performance in Indonesia. Following this, previous case studies and issues are examined
through a review of the literature. The fourth chapter presents the methodology and data for
this study, followed by a discussion of the results in chapter five. The final chapter draws
conclusionsand offer some recommendations for the future.
2. BACKGROUND
Manufacturing essentially contributes to an economic development. This section
concisely explains manufactures’ role on the economy and the manufacturing export
performance in Indonesia.
The importance of manufacturing for the economy
In 1960, the term, newly industrializing economies (NIEs) was introduced due to
the massive growth of industrialisation in Asia. Indonesia, whose economic growth was
about 6 percent a year, joined the NIEs together with Malaysia and Thailand. One specific
characteristic of these countries was a huge improvement in exports from the
manufacturing sector as industrialisation was integrated into international trade (Trade and
Development Report, 1996). Therefore, it is important for a country to develop
manufacturing export performance by diversifying export from natural-based to
manufacturing-based to support economic growth.
Transforming export activities from being heavily resource-based to manufacturing
based is very important for promoting economic growth. Coxhead and Li (2008) argue that
if the export growth of a country is mainly dominated by resource-based products, the
possibility of economic growth in the future is reduced. This is because exporting natural
resources hinders improvement in exporting other commodities due to labour and capital
being mainly utilised in the resource sector rather than in expandingthe manufacturing
sector (Corden and Neary, 1982). Similarly, Soderbom and Teal (2003) assert that relying
on the exportof primary products makes an economy vulnerable, particularly when there
are trade shocks. They suggest one method to reduce such vulnerability is by varying
export activities from primary commodities to manufactured goods. Such diversification in
exports not only reduces vulnerability, but can also increase export performance and
contribute to economic growth. Similarly, Hoekman and Nicita (2011) highlight that the
market access of developed countries is broader than that of poor countries owing to the
fact that developed countries, in most cases, export manufactured commodities, whereas
poor countries typically export primary products, such as agriculture.

3

Manufacturing export performance in Indonesia

250.00

60.00

200.00

50.00
40.00

150.00
30.00

%

USD Billion

Accordingto data from the Ministry of Trade (2015), Indonesia’s manufacturing
exports gradually increased from 2004 to 2011 except for a small drop in 2009 (Figure 2).
This trend was almost similar to trends in Indonesia’s total export. In 2004, the
manufacturing export value was about USD 36 billion, then it rose consistently to slightly
above USD 50 billion in 2008. Although it dropped to about USD 47 billion in 2009, it
recovered in the following year and remained stable at around USD 70 billion from 2011 to
2014. In contrast,total export value suffered decline after the 2011 economic crisis, but the
manufacturing sector seemed to be robust enough to stand with this economic shock
(Soderbom and Teal, 2003).

100.00
20.00
50.00

10.00

-

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Manufacturing export
Other exports

Source: Ministry of Trade, 2015
Figure 2. Manufacturing export value and total export value in Indonesia from 2004 to
2014
Moreover, from 2004 to 2014, the manufacturing sector contributed between 35 and
50 percent of Indonesia’s total export.The highest percentage was in 2004 which was just
over 50%. In 2014, the manufacturing sector contributed to the aggregate export by about
40%. However, the former Minister of Trade, Rachmat Gobel wanted to increase this
contribution to 65% to fulfill the international demand of manufacturing products.
Therefore, the following section discusses factors that are able to increase manufacturing
export performance in Indonesia.

4

3. LITERATURE REVIEW
Export is products and services produced domestically but sold overseas (Mankiw,
2010). Exporting is one part of international trade activity. A country exports products to
the global market when it has excess supply. This means that the export of a country is an
excess of domestic supply of goods and services which are produced although its citizen do
not consume them (Deliarrnov, 1995). Goldstein and Khan (1985) stated that the export
performance of a country is also determined by export demand from other countries.
Theoretically, the factors affecting exports performance of a country come from supply and
demand.
Exchange Rate
One major factor of export discussed in many research studies is the exchange rate,
which is the rate or ratio between two currencies used by people to trade across countries.
Depreciation or a decrease in the value of the money of a particular country causes its local
products to be cheaper for other countries, which experience an appreciation under these
conditions (Mankiw, 2010). Consequently, it increases overseas consumption and raises
domestic export demand. However, Krugman, Obstfeld and Melitz (2012) state that this
situation requires the Marshall-Lerner condition, in which depreciation can increase export
value if export volume is elastic to changes of the exchange rate or the elasticity’s absolute
value of export volume to exchange rate changes is more than one.
In the case of Indonesia, Hossain (2009) and Pratikto (2012) both found that the
exchange rate significantly affects Indonesia’s export performance, although they adopted
different estimations to reach this conclusion. Hossain (2009) used the autoregressive
distributed lag (ARDL) bound testing approach to analyse how global income and export
prices influenced the Indonesia’s exports from 1975 to 2005. This author found that relative
export prices negatively affect Indonesia’s export volume. Following this, Pratikto (2012)
examined impact of the real effective exchange rate (REER) on export and import value in
Indonesia by using a vector autoregression (VAR) approach. This research revealed that the
REER is necessary to explain the volatility of export and import value in Indonesia.
These research contributions made specific recommendations to the Indonesian
government. Firstly, Hossain (2009) suggested that appreciation of the rupiah, after the
huge depreciation period, explains the slowdown in exports during that period. Therefore, it
was recommended that the Indonesian government introduce a policy to maintain the
rupiah exchange rate at the optimal level in order to support the increase of Indonesia’s
manufacturing export value. On the other hand, Pratikto (2012) stated that the exchange
rate is not the sole factor determining export performance by showing that depreciation
cannot simply increase the export or reduce the import. Therefore, the exchange rate only
cannot be considered as a major factor determining Indonesia’s trade performance, and the
Indonesian government should also consider other aspects to improve Indonesia’s export
performance.
Looking at China as another example, Shao, Xu and Qiu (2012) studied the
determinants of manufacturing exports. Some determinants they considered were GDP,
GDP per capita, and the exchange rate. These authors took sample data from 1995 to 2007

5

and analysed the data using the Cox PH model. Their results suggest that the exchange rate
has a significant effect on China’s manufacturing exports because Chinese industries
majorly rely on less advanced technologies. Therefore, China depends much more on cheap
price to compete with products from other countries. Appreciation of the Chinese currency
(RMB) causes Chinese commodity prices to become more expensive, and consequently,
reduces its manufacturing exports. According to this finding, the authors recommended that
the Chinese government conducts deeper studies into export markets in developed countries
which have more stable and durable economies to ensure Chinese manufacturing export
continuity. The authors also suggested that while a stable RMB exchange rate is preferable
for Chinese manufacturers, the Chinese government should not interfere with the RMB
exchange rate and let the exchange rate depend on the market. Finally, the Chinese
government should also encourage innovation and use of advanced technologies to increase
competitiveness of Chinese products.
Foreign Direct Investment
The second main factor to be considered in export trade is foreign direct investment
(FDI). Tangavelu and Narjoko (2014) explain that such investment comes in the form of
capital and transferring technologies, skills, knowledge, and connection. Therefore, FDI
can support industries to expand productivity and develop competition among local
industries by improving resource allocation more effectively. A practical example of the
use of FDI was explained by Vernon (1966) in Denisia (2010). Vernon developed
production cycle theory to explain FDI made by the US companies in Western Europe after
the second World War in the manufacturing industries. It was begun with production of
new innovative products by the US companies for local consumption. Then high demand
from Euro of those manufactured products caused the US firms, having the advantage of
technology on international competitors, began to export. Then, once the technologies
became known, European firms have started imitating American products. Therefore, US
companies were forced to perform production facilities on the local markets to maintain
their market shares in those areas. This theory managed to explain certain types of
investments in Europe Western made by U.S. companies between 1950 and 1970.
There are other benefits of FDI. Ebadi and Ebadi (2015) affirm that FDI can also
help domestic industries to improve the product quality due to the recognition of
international standards by domestic manufacturing which can increase export opportunities.
Furthermore, Kokko (1996) as cited in Tangavelu and Narjoko (2014) explains that FDI
causes structural change in the composition of commodities from total exports because of
gains in market access and competitiveness.
According to Zhang (2005), FDI can not only increase the export performance of a
country, but conversely, can also reduce export performance because investing countries
target partner countries as their export market. This finding is consistent with findings from
Horst (1972) and Pain and Wakelin (1998). Horst (1972) found that FDI caused a reduction
in exports from the US to Canada because high import tariffs set by Canada encouraged the
US to increase its FDI in Canada to support American industries to localise production for
the domestic Canadian market. Similarly, Pain and Wakelin (1998) found that FDI in

6

Japan, Italy and Denmark reduced their export performance because the FDI was targeted
to only reach the domestic market rather than export.
More recently, Amighini and Sanfilippo (2014) found that investment hinders
export improvements. These authors analysed the impact of FDI and other factors on
Africa’s export development by modeling export development using variables such as
GDP, inflation rate, the exchange rate and FDI. They then estimated their model utilizing a
generalized method of moments (GMM) estimator. The results showed that investment to
GDP negatively affected manufacturing exports indicating that investment does not allocate
effectively to support domestic industries.
In Indonesia, according to Rahmaddi and Ichihasi (2013), over the last 25 years,
foreign investment has been dominated by Japanese investment. Japan invested in various
industries in Indonesia, such as the metal, machinery, electronic and vehicles industries.
This investment caused an increase in Indonesia’s manufacturing exports by more than
20% each year and almost six-fold in total from 1987 to 1996. For this reason, Rahmaddi
and Ichihasi (2013) investigated the effects of this investment on Indonesia’s
manufacturing exports. The authors defined manufacturing products as those in SITC 5 to
SITC 8 and applied panel data regression. The study found that FDI significantly supports
manufacturing export growth in Indonesia. More specifically, a one percent addition in FDI
in Indonesia supports an almost 0.1% increase in Indonesian manufacturing exports.
Furthermore, the authors specified that the industries gaining the most significant benefits
from FDI are the metal, machinery and electronic industries as a one percent addition in
FDI leads to a 0.25% increase in exports for those industries.
Gross Domestic Product
Gross domestic product (GDP) is the total production and expenditure of goods and
services in a country (Mankiw, 2010). Based on Goldstein and Khan (1985), local GDP
determines the supply of exports and foreign GDP has an impact on the demand for
exports. Furthermore, GDP is also effective in supporting export diversification.
Shao et al (2012) and Rahmaddi and Ichihasi (2013) found that foreign GDP
significantly affects a country’s export performance. In particular, Shao et al (2012) found
that GDP and GDP per capita of partner countries significantly affect manufacturing
exports in China. Manufacturing export duration from China is longer in high GDP
countries, indicating that China primarily depends on trading partners with large economies
such as the USA and the EU. The authors explained that this condition has not changed
even though China’s government has tried to diversify its export markets. According to
these findings, the authors recommended that the Chinese government conducts studies
about export market in developed countries having more stable and durable economy to
ensure Chinese manufacturing export sustainability.
Similarly, Rahmaddi and Ichihasi (2013) showed in the last 25 years, GDP growth
significantly influenced Indonesia’s manufacturing export value. They found that one
percent increase in GDP growth in Indonesia caused 2.7% growth in manufacturing export
because GDP growth is generally followed by advancement of production processes,
expansion of production capacities and development of export-supporting facilities and
infrastructures.

7

Trade Policies
Trade policy is an important instrument for encouraging manufacturer development,
particularly under the current trend of trade liberalisation. Generally, trade policies are
dividedinto two categories: tariff and non-tariff measures. Tariffs are the most popular
policies used to restrict trade. However, global acknowledgment of trade liberalisation has
caused countries to reduce the use of tariff instruments. Therefore, another way to restrict
trade is through the use of non-tariff measures (NTMs); consequently, the number of
countries applying NTMs has increased and its coverage of products has become wider
(World Bank and IMF, 2008). Hoekman and Nicita (2011) note that the use of NTMs
instruments represents the level of development of a country. A more developed country
will apply more NTMs, as evidenced by the high number of developed countries exploiting
NTMs.
Akinkugbe (2009) stated that the practices of tariff barriers may hinder export
improvement even though upgrading of infrastructure and effective governance can expand
export activities. These findings come from examining the effect of trade facilitation in 20
African countries on manufacturing export performance between 1995 and 2004 using a
panel data regression method. Similarly, according to Hoekman and Nicita (2011), tariff
policies are still widely exerted by poor countries with limited access to the global market.
Therefore, they suggest that lowering tariffs can improve a country’s market access.
Free Trade Agreements (FTAs) are a major instrument for countries seeking to
expand their market access and raise export value. For this reason, Yang and Zarzoso
(2014) and Nguyen (2014) analysed the impact of FTAs on export performance,
particularly manufacturing exports. Yang and Zarzoso (2014) investigated the impact of the
ASEAN-China Free Trade Agreement (ACFTA) on agricultural and manufacturing trade
creation and trade diversion dating from 1995 to 2010. They adopted a panel data gravity
equation as their statistical analysis tool. Focusing on the manufacturing sector, they found
that the coefficients of FTA in their model were positive and significant. This means that
FTA encourages manufacturing export performance of FTA member countries.
Moreover, Nguyen (2014) explored how Japan’s economic partnership agreements
(EPAs) affect the export performance of Japan’s partner countries. The EPAs caused Japan
to reduce tariff barriers and encouraged Japan to invest in its partner countries. These
agreements benefit Japan, particularly in the industrial sector in two ways. Firstly, Japan
can benefit from the low labour costs from the developing countries in which it invests. The
second advantage is that reducing import tariffs will boost Japanese companies overseas to
export to Japan. Therefore, the author suggested that the Japanese government conduct
further research into Japanese trade agreements to provide better knowledge of the impacts
of its agreements.
In the case of Indonesia, Chen (2014) investigated the impacts of the Trans Pacific
Partnership (TPP) on trade between Canada and five developing countries including
Indonesia from 1995 to 2010. Firstly, this agreement resulted in the reduction of Canadian
import tariffs, which is generally followed by an increase in Canadian imports. By using
panel regression, the author showed that Indonesia’s exports to Canada rose by almost 70%
because of this agreement.

8

Moreover, Marks (2015) examined the impact of the ASEAN-China Free Trade
Agreement (ACFTA) on Indonesia’s economy by using a simple partial equilibrium
simulation. Firstly, This author highlighted that the initial impact of the ACFTA was for
each member country to reduce its tariff barriers. Targets for reduction varied between 50%
on ‘highly sensitive products’ and zero on ‘normal-track items’. Then, focusing on the
bilateral trade between Indonesia and China, based on the author’s estimation, the ACFTA
caused China to reduce its import tariffs from Indonesia from 5.1% to 1.1%. This reduction
led to an increase in Indonesia’s export value to China by almost $2 billion, mainly from
mining and manufacturing such as machinery, vehicles and food.
According to these results, researchers suggest governments of FTA member
countries review their trade policies to improve the FTA. They recommend that the FTA
should not only focus on reducing tariff barriers but also consider other aspects, such as
reaching agreement on product standards and customs process simplification.
To sum up, there are for four main factors discussed in this study, which are the
exchange rate, FDI, GDP and trade policies. Additionally, there are potentially other
factors, such as currency similarity, language similarity and landlock.In the case of
Indonesia, those factors cannot be included because there are no countries which fulfill
those categories.
4. METHODOLOGIES AND DATA
This section briefly describes the model, approaches and data used in this study. To
examine the impact of each determinants of manufacturing export, this study applies a
gravity model and analyzes them by using panel data regression with fixed effects and
random effects approaches.
Methodologies
According to previous research, one of some methods that is appropriate to determine
factors that significantly affect the Indonesia’s manufacturing export performance is a gravity
model. This is because the gravity model captures bilateral factors that affect trade such as
geographical distance, and other economical factors (Yang and Zarzoso, 2014).

Gravity Model
The gravity model suggests that trade between two countries positively relates to GDP of
those countries, but negatively relates to trade barriers between those countries. According to
Chunxiang and Mengge (2012), one of the natural trade barriers is distance which represents trade
cost. Further distance leads to a higher trade cost which consequently causes less export. Bilateral
trade between two countries is formulated as:
=

0(

� ) 1(

� ) 2(

) 3 (� )

4

(4.1)

Where
is the export value from country i to country j.
� and
� are gross domestic
products in country i and country j respectively.
is the distance between country i and

9

country j. � is a vector of other determinants of export between county i and country j(Sheldon,
Mishra, and Thompson, 2013).
According to the equation above, the general gravity model consists of factors that may
significantly affect export value. In this study, those factors are the exchange rate, FDI, distance,
import time, population and FTA implementation.

Panel Data Regression
Panel data are data which combine cross-section and time-series data together.
Therefore, to make a regression of these data is different to a regression of a regular crosssection or time-series only. A panel data regression has a double subscript on its variable
representing individuals and time series. Baltagi (2008) formulated the model of panel data
regression as:
� =

+

′

= 1, … , ;

+

= 1, … , �

(4.2)

Where denotes individuals and represents time. Hence, error component,
, of this
model consists of individual effects that cannot be observed and other disturbance.
=� +

(4.3)

Where � shows time-invariant individual effect and
shows other interference.
Approaches of the panel data regression include the fixed effect model and the
random effect model. The fixed effect model (FEM) considers the individuality of each
cross-section unit and let the intercepts differ for each individual ( 1 ). Yet, this model still
assumes that the slope coefficients are constant across individuals. Gujarati (2003)
formulated the model as:
=

+

1

2 2

+

3 3

+

(4.4)

The intercept 1 has subscript i to show that the intercepts of individuals may be different
because of particular characteristics of each individual. However, it has no subscript t to
suggest that each individual is time invariant.
Although the random effect model (REM)has the same basic model as FEM,
= 1 + 2 2 + 3 3 + , this model does not treat 1 as fixed, but assumes that
1 is a random variable having mean value of 1 , hence the intercept value for each
individual is:
1

=

1

+�

= 1, 2, … ,

(4.5)

Where � is a random error term with a zero mean value and variance of �� 2 .
By combining equation (4.4) and (4.5), the equation becomes (Gujarati, 2003):
=
=

1

+
1+

+
2 2 +
=� +

2 2

3 3
3 3

+� +
+

(4.6)
(4.7)

10

The composite error term
consists of error from the cross-section (� ) and the
combination between time series and cross-section error component ( ).
Hausman’s test
To choose the more appropriate approach between FEM and REM, a test is
required. One of test that can be considered is the Hausman’s test (Gujarati, 2003).
Hausman (1978), as cited in Baltagi (2008), stated that coefficient of REM ( � ) is
consistent with coefficient of FEM (
) under the null hypothesis 0 :
= 0.
Hausman argued that
is consistent without considering whether 0 is true or not.
However, �
is only consistent and asymptotically efficient under 0 . Gujarati (2010)
concluded that if the null hypothesis is rejected, the REM approach is not appropriate;
consequently, it is better to use the FEM approach.
Specification
This paper applies the gravity model as proposed by Sheldon, Mishra, and Thompson
(2013) with some more variables are added to the equation.1 After applying a logarithm
transformation, the model becomes:
ln⁡
(

) = �0 +

+
+
+
+

1

ln

ln
7 ln �

5

11
16

�

+

�
06 +
11 +

2

ln � �
−1

12
17

+

+
6

+ 8 ln
07 +
12 +

ln
13
18

3

� �
+
� � −1
08 +
13 +

14
19

4

ln

+

−1

9 ��

09 +
14 +

+
15

��
10

10

(4.8)
Where ln⁡
( ) is the logarithm of Indonesia’s real manufacturing export value to country j2
in year t, which is ranging from 2005 to 2014.ln
� is the logarithm of real GDP of
country j in year t. ln � � is the logarithm of real exchange rate against country j in

year t.

� �

� � −1

is the ratio of real exchange rate of year t and the prior year representing a

relative change of exchange rate in year t. ln
is the logarithm of bilateral distance
between Indonesia and country j. ln
−1 and ln
−1 are
the logarithms of the amount of positive total FDI inflows (investment) and the amount of
negative total FDI inflows (disinvestment) to Indonesia from country j in year t-1.
ln �
�
is the logarithm of the population of country j in year t.
ln
is the logarithm of number of importing days in country j in year t.
��
is the average tariff applied by country j in year t. �� is a dummy variable of
FTA implementation between Indonesia and country j in year t.
06 to
14 are
The equation excludes a factor of Indonesia’s GDP, because this study only focuses on Indonesia. The use
of time dummies takes up the effect of Indonesia’s GDP changes.
2
Countries used in this research are listed in table 3.

1

11

time dummies to capture particular effects in each year affecting Indonesia’s manufacturing
export value.
Data
This study utilises yearly panel data, dating from 2005 to 2014, which consists of 28
countries resulting in a total of 280 observations. Countries, chosen in this study, are the
top 28 importing countries in 2014, which contribute approximately 90 percent of the total
of Indonesia’s manufacturing exports.
Indonesia’s nominal manufacturing export data are obtained from the Ministry of
Trade, Republic of Indonesia. The nominal export data are, then, adjusted to real value by
using the export price index with a year base of 2005 obtained from Bank Indonesia’s
Producer Prices Indices (2015). Real GDP data in constant USD at 2005 prices, the
nominal exchange rate data, population data and data of the import days of each country are
sourced from the World Bank’s World Development Indicator (2015). To obtain the real
exchange rate, nominal exchange rate data are multiplied by the relative price between two
countries represented by the ratio of consumer price index (CPI) of each trading partner and
Indonesia (Siregar and Rajan, 2004). The consumer price index is also obtained from the
World Bank’s World Development Indicator (2015). Data of nominal FDI inflows to
Indonesia are acquired from UNCTAD. According to UNCTAD (2013), these data are on a
net basis, therefore, their value each year might be positive representing investment or
negative representing disinvestment. There is a problem when these data are transformed to
logarithm value. Hence, to address this problem, these data are divided into two variables:
investment, consisting of positive value of FDI inflows, and disinvestment, consisting of
the absolute value of the negative value of FDI inflows. According to Cavallari and
d’Addona(2013), nominal FDI data are scaled by the GDP deflator of each partner country
from the World Bank’s World Development Indicator (2015) to obtain real values. Data of
distances between each country and Indonesia are from CEPII (2015) and the average tariff
data from the World Bank’s WITS (2015).
5. RESULT
Finding the best approach
According to Gujarati (2010), one formal test to find the more appropriate model
between FEM and REM is a Hausman’s test. The result of Hausman’s test is presented on
Table 1.

Chi-square

Table 1. Hausman’s test between FEM and REM
Hausman Test
Statistic
d.f
10.37
18

Prob.
0.919

12

The null hypothesis of the Hausman’s test is difference in coefficients of FEM and REM is
systematic. Based on Table 1, p-value for the Chi-square statistic is 0.919 which is greater
than = 0.05, therefore the null hypothesis cannot be rejected. This means that coefficients
between FEM and REM are not significantly different. Consequently, the more appropriate
approach for this study is REM.
Impact of each determinant
Table 2 shows the result of the panel data regression with random effects estimation
of Indonesia’s manufacturing export value. The estimation shows that real foreign GDP, the
ratio of real exchange rate, distance, import time and average tariff have a statistically
significant effect on Indonesia’s manufacturing export performance. However, results
indicate that the coefficients of real exchange rate level, FDI inflows, population and the
implementation of FTA are not statistically significant. Hence, those factors cannot be
assumed to affect Indonesia’s manufacturing export performance.
Table 2. Impact of each determinant on the Indonesia’s manufacturing export (random
effects)
Random Effect Model
Coef.
Std. Err.
Constanta
11.297**
1.556
Ln (Real GDP)
0.299**
0.064
Ln (Real Exchange Rate)
0.045
0.040
Ratio Real Exchange Rate
-0.665**
0.172
Ln (Distance)
-0.813**
0.142
Ln (Investment)t-1
-0.002
0.002
Ln (Disinvestment)t-1
-0.001
0.003
Ln (Population)
0.107
0.074
Ln (import days)
-0.275**
0066
Average Tariff
-0.019*
0.009
FTA Implementation
0.042
0.038
Year 06
0.028
0.052
Year 07
0.133**
0.045
Year 08
-0.028
0.047
Year 09
-0.182**
0.047
Year 10
-0.086
0.056
Year 11
0.032
0.053
Year 12
-0.019
0.054
Year 13
-0.090
0.054
Year 14
-0.082
0.055
* and ** are statistically significant at level of five and one per cent, respectively.
Variable

P-value
0.000
0.000
0.252
0.000
0.000
0.398
0.771
0.150
0.000
0.038
0.266
0.583
0.003
0.543
0.000
0.125
0.537
0.721
0.096
0.139

The estimated coefficient of real foreign GDP capturing market size in Indonesia’s
trading partner countries is positive (0.299) and significant at the level of 1%. This means
that a one percent addition of GDP in one of Indonesia’s trading partners contributes to an
almost 0.30% increase in Indonesia’s manufacturing exports to that country. This result is
consistent with that of Shao et al (2012).

13

The coefficient of the ratio of the exchange rate, representing a relative change in
the exchange rate, is (-0.655) which is a significant effect on manufacturing export value at
the level of 1%. This means that one percent increase in this ratio causes export value
decreases by about 0.65%. This indicates that a depreciation of the rupiah will reduce
Indonesia’s manufacturing exports. There are, at least, two possible reasons supporting this
argument. The first reason is the nominal effect of the rupiah depreciation. It causes
Indonesian products to be cheaper for international buyers using other currencies.
Therefore, buyers spend less money than before depreciation although they buy the same
amount of Indonesian products. As a consequence, the total value of Indonesia’s export
significantly decreases. The second reason is that Indonesia’s manufactures rely much on
the imported raw material. Rupiah depreciation causes imported raw materials to become
more expensive resulting in a reduction of import of raw material for manufacturing.
Consequently, manufacturing production falls and its export value decreases.
The coefficient of distance between Indonesia and its partner countries is negative (0.813) and significant at the level of 1%. This means that Indonesian exports less
manufacturing products to countries located at a greater distance. For example, distance
between Indonesia and Thailand is two times larger than the distance between Indonesia
and Malaysia, therefore, Indonesia’s manufacturing export value to Thailand is expected to
be less than its value to Malaysia by about 81% if other factors do not change.
The effect of the number of days for the importing process is negative (-0.275) and
significant at the level of 1%. This means that a one percent reduction on importing days
causes an increase in manufacturing exports by 0.27%. A practical example of this finding
is if Malaysia reduces its average importing days from 10 days to 9 days, Indonesia’s
manufacturing export to Malaysia is expected to increase by about 2.7 percent under a
condition of no changes in other factors.
The impact of the average import tariff applied by each partner can be seen in the
coefficient of the average tariff which is -0.019 and significant at the level of 5%. This
means that if the tariff is one percent higher; the export value reduces by about 0.02%. This
suggests that higher tariffs lead to lower export value. For instance, if the average import
tariffs applied in Malaysia increase from around 5% to about 6%, it is expected that
Indonesia’s manufacturing export to Malaysia will fall by 0.02% if all other factors do not
change.
On the other hand, FTA implementation does not have significant effect on
Indonesia’s manufacturing exports. Although the coefficient of the FTA dummy is positive,
this value is not significant at the level of 5%. This indicates that the implementation of the
FTA only has an effect through tariffs, which significantly affect manufacturing export.
Moving to FDI inflows, neither investment (the positive value of FDI inflows) nor
divestment (the negative value of FDI inflows) significantly affects manufacturing exports
because both coefficients are not significant at the level of 10%. There are possible reasons
explaining this finding. Firstly, FDI inflows in Indonesia from each country may be
concentrated on particular sectors. Therefore, its effect of manufacturers as a whole is
considerably weak. Another reason might be that FDI inflows cannot directly affect export
performance in a year after, meaning that time lag between investment and its impact may
be bigger than one year. Yet, although they are not significant, the negative value of the
investment coefficient should raise concern. According to Zhang (2005) and Amighini and

14

Sanfilippo (2014), the negative relationship between investment and export performance
means that either FDI is not allocated efficiently or FDI is targeted to reach only the
domestic market and not the export market.
Focusing on time dummies, in which 2005 becomes a year base, based on the
estimation, export value in 2007 and 2009 significantly differ from other years. The
coefficient of the time dummy in 2007 and 2009 are 0.133 and -0.183 respectively. To
know the real effect of the time dummy, it needs a mathematical operation to retransform it
from the logarithm form.
For the 2007’s time dummy:
ln
ln

= 0 + 12
07 where
= 11.297 + 0.133
07
=

11.297+0.133

0

= 11.297 and

12

= 0.133

07

(4.9)
(4.10)
(4.11)

07 = 0, then:

If

=
Whereas if

11.297+0.133(0)

=

11.297

= 80,579.54

11.297+0.133(1)

=

11.430

= 92,041.97

07 = 1, then:
=

For the 2009’s time dummy:
ln
ln
=
If

= 0 + 14
09 where
= 11.297 − 0.182
09

11.297−0.182

0

= 11.297 and

09

12

= −0.182

(4.12)
(4.13)
(4.14)

09 = 0, then:
=

Whereas if

11.297−0.182(0)

=

11.297

= 80,579.54

11.297−0.182(1)

=

11.115

= 67,171.21

09 = 1, then:
=

According to calculation above, it can be concluded that there was an addition to
export value in 2007 of USD 11.46 billion, whereas in 2009, the manufacturing export
value dropped by USD 13.48 billion.
According to Bulman report (2008), the significant increase in Indonesia’s
manufacturing exports in 2007 was supported by three conditions. Firstly, the high growth
in industries producing automotive, chemical, and agricultural products, with the highest
growth being in the automotive industries, growing by 37 percent in 2007. Secondly, the
increase in commodity prices, particularly palm oil rising by 55 percent in 2007, also

15

supported improvement in Indonesia’s manufacturing export performance. The third factor
was that in 2007 Indonesia successfully reduced its dependency on the US as Indonesia’s
biggest export market. In 2007, the US was still the second largest export market for
Indonesia even though its contribution to Indonesia’s non-oil and gas export performance
decreased, from about 17% in 2000 to 12% in 2007.At the same time, Indonesia’s export to
Malaysia, China and India increased by 48%, 22%, and 21% respectively, which in total,
contributed to 22% of Indonesia’s non oil and gas export. This export market
diversification seems to be beneficial for Indonesia’s manufacturing export performance by
increasing its value. Additionally, the improvement of export performance in 2007 also
contributed to economic growth in Indonesia, reaching 6.3 percent which was the highest
growth since the 1990s (Bulman report, 2008).
On the other hand, in 2009, Indonesia’s manufacturing exports suffered a significant
decrease. This reduction started from the middle of 2008. Based on the Purna report (2009),
this reduction was mainly caused by the financial crisis of 2008 which led to a slowdown in
the global economy. According to Indonesia economic quarterly report (2009), the highest
drops in manufacturing exports were to Japan, Singapore, the US and China. These declines
were due to a fall in the prices of Indonesia’s export commodities such as mining products,
oil and rubber.
6. CONCLUSION
Indonesia’s total export value has been decreasing since 2012 due to the global
economic crisis and the drop of commodity price in 2011. This study, focusing on export in
manufacturing sector, is conducted to support the Indonesian government to overcome this
problem. By applying panel data regression using random effect approach, this study
analyses the impact of export determinants on Indonesia’s manufacturing export
performance to the biggest 28 partner countries from 2005 to 2014.
Results of this study show that real foreign GDP and distance between Indonesia
and its partner countries have a statistically significant effect on Indonesia’s manufacturing
export. Real foreign GDP positively affects manufacturing export, whereas distance
negatively affects the export performance. However, Indonesia cannot interfere with these
factors, therefore there are no direct policy recommendations that can influence these
factors, except to conduct deeper research to understand and maintain Indonesia’s export to
rich countries which have a higher export share than other countries. In addition, the
government should also consider other countries with which to diversify its export market
because if GDP of those countries increase in the future, this will encourage an increase in
Indonesia’s manufacturing export. Moreover, this can also minimize the impact of trade
shock suffered by partner countries in the future.
Moving to other factors, ratio of the real exchange rate, number of import days and
average tariffs statistically negatively affect Indonesia’s manufacturing export.
Accordingly, there are suggestions for the Indonesian government to improve
manufacturing export values. Firstly, the Indonesian government should stabilise the rupiah
exchange rate to prevent the drawback of rupiah depreciation. Moreover, as suggested by
Shao et al (2012), the Indonesian government should encourage domestic industries to use

16

more advanced technologies to reduce dependency of Indonesia’s export on price
competitiveness influenced by the rupiah exchange rate. In regard to import days and
average tariffs, the Indonesian government can suggest simplification of import processes
by, for example, promoting a national single window system and tariffs improvement
through FTA negotiation in order to improve international trade between countries.
However, the results also reveal that the level of exchange rate, FDI inflows
(investment and divestment), foreign population and FTA implementation do not have a
statistically significant effect on Indonesia’s manufacturing export performance. Focusing
on FTA implementation, the Indonesian government is suggested to support the expansion
of FTA scope beyond tariff reduction, such as agreement of products standards and customs
process simplification. Next, regarding FDI inflows, although they do not statistically affect
manufacturing export as a whole, it cannot be assumed that FDI inflows do not support
Indonesia’s manufactures as study by Rahmaddi and I