1st National and International Graduate Conference

March 27, 2015, KU Home, Bangkok, THAILAND

Impact of Free Trade Agreement on Indonesias Agricultural Exports

Heri Akhmadi*

Abstract
This study assesses whether Indonesian participation on free trade agreement increases export of
agriculture commodity. Annual data from 140 partner countries for years 2004-2013 has been used. The
augmented gravity model is chosen to analyzed the factors affected Indonesian Agricultural Exports.
Result showed that Indonesian agricultural exports are positively correlated with the size of economic,
partner countries population and the enrollment on free trade agreement, while they are negatively
correlated with geographical distance. Thereby, the membership in free trade agreement leads to increase
Indonesian agricultural exports.
Keyword: Free Trade Agreement, Gravity Model, Agricultural Exports

Introduction
Over the past two decades since early 1990s, free trade agreement had become the focus of
various groups of countries. As of 15 June 2014, some 585 of Regional Trade Agreements (RTAs) had been
notified to the World Trade Organization (WTO). Economic integration, in particular can lead to increase
trade and other benefits in the form of a more competitive trade region by removal of trade and nontrade barriers and free flow of goods and services.
Indonesia, like another country in the world also involved on several regional economic
integration. Based on RTA database that notified to WTO, until 2014 Indonesia has become member of at
least seven free trade agreements (World Trade Organization [WTO], 2014).
The membership in regional trade agreement is expected to increase trade among parties due to
decreasing trading cost and removing trade barrier. This policy ultimately can enhance market size and
increase the competitiveness of countries product, which in the end could increase economic growth and
welfare.
Despite many potential benefit, trade liberalization gets much criticism regarding its effect on
*

Student, MABE Program, Faculty of Economics Chulalongkorn University; Email: heriakhmadi@gmail.com

1st National and International Graduate Conference

March 27, 2015, KU Home, Bangkok, THAILAND
economy particularly the liberalization in agriculture sector since agriculture still becomes major sector by
many countries especially developing countries. For Indonesia, agriculture sector is still considered as the
economic backbone due to the contribution of this sector to countrys Gross Domestic Product (GDP) and
supply around two fifth of countrys labor force. Based on Statistics Indonesia (Badan Pusat Statistik [BPS]
2013), the share of agriculture sector to Indonesian GDP are around 14.43%-15.29% in the last 5 year
since 2009 to 2013. In term of employment, according to National Labor Force Survey 2013 conducted by
Statistic of Indonesia (BPS), agriculture sector provide 39,220,261 employment or about 34.78% of
Indonesian workforce.
Trade liberalization undertaken by Indonesian government through membership in several free
trade agreements (FTAs) on the one side may encourage economic growth and increase trade because
Indonesias commodity would had larger market and get more efficient trade procedure within FTA. But
in the same time, participation in FTA also gives more access for another country to Indonesian market,
this means would threaten domestic agriculture commodity because they will compete with other
countries imported product which may cheaper and better in quality.
This condition is interesting to investigate whether Indonesias agricultural exports are influenced
by Indonesias membership in FTAs. It is also appeal to find out whether these facts effect on Indonesias
decision to participate in free trade agreement. Hence the objectives of this study is to assess the
economic impact of Free Trade Agreement (FTA) on Indonesias Agriculture exports.

Literature Review
A number of researches have been conducted to examine of the impact of Free Trade
Agreement (FTA) on agriculture trade. A previous study about Agriculture, Trade, and Regionalism in South
Asia conducted by DeRosa and Govindan (1996) was examined South Asias agriculture and trade relations
and the implementation of the South Asian Association for Regional Cooperation (SAARC) Preferential
Trading Arrangement (SAPTA). Using quantitative analysis method with a simple economic model, this
study found that free trade agreement would expand intra-trade substantially, especially in agriculture
commodities. Trade creation in agriculture sector is found to be limited only $86 million or 2 percent, but
trade diversion will be extensive under free trade agreement for about 75 percent ($628 million).
Grant and Lambert (2005) investigated the effects of Regional Trade Agreement (RTAs) on
agriculture trade. Their research aims to assess the effects of 8 Regional Trade Agreements (Including:
NAFTA, MERCOSUR, EU-15, CER, AFRICA, AFTA, APEC and Andean Pact) in agricultural trade by specifying
an extended gravity model. The result shows that the income elasticity (GDP + per capita GDP) are
positive and statistically significant, the magnitudes of these estimates is 0.90 and 0.70 for the exporting
and importing country respectively. Distance is negative and significant in 5 of the 9 individual agricultural
commodities and for all agriculture and non-agriculture regressions. The result also indicates that AFTA
[59]

1st National and International Graduate Conference

March 27, 2015, KU Home, Bangkok, THAILAND
members actually traded less with each other after the formation of the agreement because AFTA
members do not specialize in production or exports of agricultural commodities. Thus we expect to see a
positive trade diversion coefficient implying that AFTA members rely more on imports from larger
countries that specialize in production and exports. In summary, this research result supports the claim
that RTAs have generally increased agricultural trade between members but the magnitude of this
increase depends on the average flow it is being compared to.
Another study by Grant and Lambert (2008) about Trade Agreement and Agricultural Trade using
the gravity model has been shown that Regional Trade Agreement (RTA) increase members agricultural
trade. The increase in members trade was much larger in agriculture compared to non-agriculture in all
panel specifications. The average effect of RTAs was to increase members agricultural trade by 72%
compared to just 27% for members non-agriculture trade.
A research about Turkish agricultural exports to the European Union (EU) was conducted by
Erdem and Nazliogrlu (2008). A gravity model was developed to analyze the determinants of Turkish
agricultural export to European Union. The gravity model estimation results are reported that the
estimated coefficient of the CU shows that Turkish agricultural exports to the EU countries, which are the
member of the customs union, are 130.82 percent higher than others. As conclusion, the results show
that conventional variables (the size of the economy, the importer population, and distance) of the
gravity model and the memberships of Turkey in custom union (free trade agreement) have significant
impacts on Turkish agricultural exports to EU.
A recent study about Impact of Free Trade Agreement on Indonesias Agriculture Trade
conducted by Dianniar (2013). Using the gravity model, this research shows that Indonesias participation
in AFTA and ACFTA does not have a significant impact on Indonesias agricultural trade flows. However,
this does not necessarily mean that become a member of FTA would not favorable for Indonesia. The
participation on AFTA contributed a little additional welfare gain for Indonesia because ASEAN remains a
weak regional group with a small market size in the global economy and because most international
trade by Indonesia is with non-ASEAN countries. Hence, the expected gain from tariff reductions under the
CEPT scheme is very small because the tariff reduction is applied only to ASEAN members.

Methodology
The impact of free trade agreement on Indonesias agricultural exports can be analyzed with
gravity model. The gravity model is widely used tool to analyze factor affecting agricultural of trade flows
such as free trade agreement, exchange rate, common border, language commonality and arable land
(Erdem and Nazlioglu, 2008).
The traditional basic gravity model established by Tinbergen (1962) underlying the value of
exports from country i to country j, Xij is a positive function of countries gross domestic product (GDP), but
[60]

1st National and International Graduate Conference

March 27, 2015, KU Home, Bangkok, THAILAND
negatively related to the distance between countries. While many literatures agree to the empirical
model that GDP and distance is the main explanatory variable, many studies uses other variable to be
included as another explanatory variable. Ghosh and Yamarik (2004) showed a list of 48 independent
variables that has been used in literature to estimate the gravity model in various combinations.
Sohn (2005) concludes that in gravity model, the most common dependent variables are exports
and bilateral trade flows. While the explanatory variables are factors indicating demand and supply of
trading countries, and impedance factors of trade flow between countries. The proxies for demand and
supply are measures of countries economic and market size such as income level, population, area size
and GDP per capita.
Greene (2013) on his paper about Export Potential for U.S. Advanced Technology Goods to India
Using a Gravity Model Approach said that the most often used as dependent variable in gravity model are
total trade (exports + imports), exports and imports. While on the right hand side as independent
variable, most researcher include country income level, geographical distance, land area, population, real
exchange rate, market openness, FTA membership and other geographic characteristic (Island, landlocked,
etc.).
To this end, this research follows Erdem and Nazlioglu (2008) and Greene (2013) gravity model
specification, the model is as follows:
LnEijt = 0 + 1lnGDPijt + 2lnDistij + 3lnPjt + 4Landjt + 5Exjt + 6FTAij + ijt
Where: Eijt is total Agricultural export from Indonesia (i) to partner countries (j) measured in
current US$ dollar. GDPij is the sum of real GDP of Indonesia and partner countries measured in US$
dollar. Distij is geographical distance between capital cities of Indonesia (i) and partner countries (j) in
kilometers. Pjt are Indonesian partner country j population. Landjt represents total irrigated land of
Indonesian partner country j in hectare. Exjt are Indonesian partner country j currency real exchange rate
per US$. FTAijt is dummy variable if Indonesia and partner country participate in free trade agreement
(FTA).
Panel data is used in this study. The reason for using this type of data is because panel data can
give empirical analysis in a way that is not feasible if just use a cross-section or time series data. There are
three types of panel data estimation model: Pooled OLS, Fixed Effect Model (FEM), and Random Effect
Model (REM).
To choose which is the best effective model for this research between FEM and REM, Hausman
test is applied. The null hypothesis that underlying the Hausman test is that REM is more preferable. The
test statistic developed by Hausman has an asymptotic 2 distribution. By testing the model, we are able
to choose the best estimation model based on the available data set and eventually we will make the
proper analysis. (Wooldridge, 2013).
[61]

1st National and International Graduate Conference

March 27, 2015, KU Home, Bangkok, THAILAND
Egger (2000) pointed out that the random effects model (REM) would be more appropriate when
estimating trade flows between randomly drawn samples of trading partners from a larger population.
While, the fixed effects model (FEM) would be a better choice than REM when one is interested in
estimating trade flows between a predetermined selections of nations. Since our sample only includes
trade exchanges between Indonesia and its trading partners, the FEM might be the most appropriate
estimation. However, Hausman test is also conducted to check whether the REM is more efficient that the
FEM estimation.
The yearly data of agricultural export commodities are obtained from Statistics Indonesia,
Indonesian Ministry of Agriculture and United Nation Commodity Trade Database (UN-COMTRADE). The
data on GDP (real value in the 2005 base year), population and exchange rate are obtained from World
Development Indicators (WDI) database of the World Bank. Data of Free Trade Agreement (FTA) are come
from World Trade Organization (WTO) regional trade agreement database, ASEAN secretariat and t he
official website of the FTA. While data on distance are collected from Centre dEtudes Prospectiveset
dInformations Internationales (CEPII) and data of irrigated land are collected from Food and Agriculture
Organization (FAO). This study will use panel data that are collected for periods 2004-2013. This period
was selected as it covers the time when all FTA which followed by Indonesia were inforced.

Result
The estimation results of gravity model are presented in Table 1. The Hausman test shows that
there is no correlation between individual random effects and explanatory variables, indicating that the
REM is consistent and efficient. Tests results verify our model selection and refer to the one-way REM
including only individual effects. Furthermore, gravity model using REM usually employed when
destination countries selected randomly from a larger population (Greene, 2013). Gujarati and Porter
(2009) on his book Basic Econometrics pointed out, if number of cross section data is large and time
series data is small, REM may be preferable. However, in this study, selected countries are chosen
randomly from all Indonesian trading partners, so REM would be best suited.
The R2 value reported in Table 1 show an estimation of 0.3755. Its means the model explains
37.6 percent variation in Indonesian agricultural exports during 2004-2013. The sign of the explanatory
variable are as expected and statistically significant mainly for original gravity model variable (size of
economic, distance and population). The sum of GDPs, the trading partners population and the dummy
variables Free Trade Agreement (FTA) are positively affected on Indonesian agricultural exports, while
distance has negative impact. Another independent variables, irrigated land and real exchange rate are
negative and statistically insignificant affect Indonesian agricultural exports.

[62]

1st National and International Graduate Conference

March 27, 2015, KU Home, Bangkok, THAILAND
Table 1 Gravity Model Estimation for Indonesias agricultural exports, 2004-2013 Using Random Effect
Model Dependent variable: ln (Indonesian agricultural exports)
Independent Variable
Coefficient
Standard Error
t-statistics
Constant
-78.63196
(10.08631)
-7.795906***
LnTGDP
3.066175
(0.403628)
7.596543***
LnDIST
-0.769011
(0.313463)
-2.453278**
LnPOPj
1.172582
(0.283694)
4.133270***
LAND
-0.107478
(0.173391)
-0.619860
EXCj
-0.072511
(0.099392)
-0.729541
FTA
1.115355
(0.563970)
1.977685**
2
R
0.375555
F-Statistic
30.82897
N
1400
Note: ***/**/* significant at 1%, 5%, and 10% level. N is number of observations.
The estimated coefficient of the LnTGDP as a proxy for the wealth effect of the size of economy
has the expected positive sign and statistically significant. The magnitude of this variable is 3.07, implying
that a 1 percent increase in the sum of GDPs will increase Indonesian agricultural exports by 3.07 percent.
The coefficient of TGDP provides most of the explanatory power for the random effect model.
The estimation result of distance as one important variable on gravity model show as expected
in negative value and significant impact on Indonesian agricultural export. Its means that distance as a
proxy of transportation cost is still become hindrance factor for Indonesia to develop trade with countries
partners. Distance is significant determined of Indonesian agriculture exports with the coefficient of -0.77
implies that a 1 percent increase in the distance leads to decrease Indonesian export by 0.77 percent.
While population of partner country as represent in the coefficient of LnPOP shows positive
effect and statistically significant. Suggesting that larger importer population, means greater market and
consumption is good to enhanced agricultural exports. The magnitude of this independent variable is
1.18, indicating that a 1 percent increase in the partner countries population will increase Indonesian
agricultural export by 1.18 percent.
The irrigated land variable has an expected negative sign but statistically insignificant effect on
Indonesian agricultural exports. This means that the extent of irrigated land owned by partner country
does not affect on Indonesia to export more. Similarly the variable real exchange rate also statistically
insignificant and has negative coefficients.
The dummy variable free trade agreement (FTA) as a proxy of countrys economic integration
shows positive sign and statistically significant. The magnitude of this variable is 1.12, bear a meaning that
Indonesian export to partner countries which are members of the FTA will be higher than for countries
[63]

1st National and International Graduate Conference

March 27, 2015, KU Home, Bangkok, THAILAND
that are not affiliated with the FTA. This result is in line with the theory and research that the
participation on free trade agreement will increase trade due to decreasing trade barriers among countries
involved. Because the basic principle of any economic integration is the elimination of trade and nontrade barriers among two or more countries with the ultimate goal is to increase trade (Kehoe, 2007)

Conclusion
The objective of this study is to employ an augmented gravity model of international trade to
empirically analyze the impact of free trade agreement (FTA) on Indonesias agricultural exports during
the years 2004-2013. The gravity equation included standard gravity variables plus dummy variable FTA.
The results are based on the study of 140 Indonesian trading partners over a 10 year period. Regression
analysis was performed on panel data in three ways: pooled OLS, the random-effect model, and the
fixed-effect model. The random-effect model was selected because it fits the data better and is more
efficient than either OLS or the fixed-effect models.
The result shows that conventional variable of the gravity model (i.e. GDP as size of economy,
distance as proxy of trade cost and importer population) and dummy variable free trade agreement (FTA)
have significant impact on Indonesian agricultural exports. Unexpected are shown by independent
variable irrigated land (LAND) and real exchange rates do not bring significant impact on Indonesian
agricultural exports.
The policy implication that can be suggested from this research is that Indonesia should explore
more benefit from their membership in FTAs, particularly related to agricultural products trading
agreements. The government of Indonesian should take correct measures to increase trade volume with
other countries not only with big economic countries that have high income per capita but also with
other countries that in the similar level with Indonesia.
Lattermost, this research is employed few in its explanatory variable. Hence, for further
development of this study it is obligatory to consider include more explanatory variable that already
proved by other previous research, such as dummy variable common language, colonial link and
investment.

References
Centre dEtudes Prospectiveset dInformations Internationales (CEPII). 2014. Geographical Distance.
Retrieved from www.cepii.fr/anglaisgraph/bdd/distances.htm.
Derosa, D. & Govindan, K. 1996. Agriculture, Trade, and Regionalism in South Asia. Journal of Asian
Economics 7 (2): 293-315.

[64]

1st National and International Graduate Conference

March 27, 2015, KU Home, Bangkok, THAILAND
Dianniar, U. 2013. The Impact of Free Trade Agreements on Indonesias Agricultural Trade Flows: An
Application of the Gravity Model Approach. Master Thesis, International Institute of Social
Studies. The Hague.
Egger, P. 2000. A Note on the Proper Econometric Specification of the Gravity Equation. WIFO.
Erdem, E. & Nazlioglu, S. 2008. Gravity Model of Turkish Agricultural Exports to the European Union.
International Trade and Finance Association Working Paper 2008, The Berkeley Electronic Press.
Food and Agriculture Organization (FAO). 2014. Area Equipped by Irrigation Data. Retrieved from
www.fao.org/nr/water/aquastat/irrigationmap/index.stm
Ghosh, S., Yamarik, S., 2004. Are preferential trade agreements trade creating? An application of extreme
bounds analysis. Journal of International Economics 63: 369-395.
Grant, J. & Lambert, D. 2005. Regionalism in World Agricultural Trade: Lesson from Gravity Model
Estimation. Agricultural and Applied Economics Association.
__________. 2008. Do Regional Trade Agreements Increase Members' Agricultural Trade?. American
Journal of Agricultural Economics 90: 765-782.
Greene, W. 2013. Export Potential for U.S. Advanced Technology Goods to India Using a Gravity
Model Approach. Office of Economic Working Paper. US International Trade Commission. No.
2013-03B.
Gujarati D. & Porter, D. 2009. Basic econometrics. 5th ed. Mc Graw-Hill International.
Kehoe, W. 2007. Regional and Global Economic Integration: Implication for Global Business. University
of Virginia. Virginia.
Sohn C. 2005. Does the gravity model explain South Koreas trade flows?. Japanese Econ Rev 56 (4):
417-30.
Statistics Indonesia (Badan Pusat Statistik [BPS]). 2014. National Labor Force Survey 2013.
Tinbergen, J. 1962. Shaping the World Economy, Suggestions for an International Economic Policy.
New York: Twentieth Century Fund.
UN-COMTRADE. 2014. United Nation Commodity Trade Database. Retrieved from comtrade.un.org/data.
Wooldridge, J. 2013. Introductory Econometric, A Modern Approach. South Western CENCAGE Learning.
WORLD-BANK. 2014. World Development Indicators. Retrieved from data.worldbank.org/data-catalog/
world-development-indicators.
World Trade Organization 2014. Regional Trade Agreement Database. Retrieved from rtais.wto.org/UI/
PublicAllRTAList.aspx. Accessed: July 2014.

[65]