Modern Economy, 2011, 2, 132-141
doi:10.4236/me.2011.22018 Published Online May 2011 (http://www.SciRP.org/journa l/ me )
Copyright © 2011 SciRes. ME
Long-Run Effects of Customs Union between European
Union and Turkey: Is It Zero-Sum Game?
Nedret Demirci, Levent Aydın
Department of Economics, Kırıkkale University, Kırıkkale, Turkey
E-mail: leventaydin60@gmail.c om
Received January 11, 2011; revised February 17, 2011; accepted February 26, 2011
Abstract
Without being a full member of the European Union, Turkeys participation to the Unity raised some sub-
stantial and continued questions in Turkey about the economic results of a regional integration. Some argue
that, particularly adoption of the Union’s common external tariff regime would have a devastating effect on
Turkeys trade. The aim of this paper is to search the pros and cons of customs union associated with com-
mon external tariff between Turkey and European Union. Our study especially focuses on the effects of
common external tariff to Turkeys trade. The simulation results show that Turkey has been benefiting from
the noticeable tariff liberalization as well as improved market conditions in line with EU since it ratified
custom union with EU.
Keywords: Trade Liberalization, Customs Union, Common External Tariff, GTAP Model, Turkey
1. Introduction
Turkey is a country located at the crossroads of different
regions and cultures bridging Europe from the West and
Asia from the East. Due to this fact, relationships be-
tween Turkey and Europe have an historical basis. In fact,
beginning from the 19th century, during the westernize-
tion process of the Turks, Europe and Turkey have on-
going relationships not only in economic but also politi-
cal, social and military fields. Along with the establish-
ment of the new Turkish Republic in 1923, especially in
new era starting with the end of the Second World War,
Turkey more enthusiastically adopted Western (specifi-
cally European) values and has a closer and continued
connection with the European Countries. Accordingly,
from their beginning, he is a member of many western
institutions and integrations like United Nations, N ATO,
the OECD, the Council of Europe, and lately European
Union.
Regarding the adventure of the Turkey’s participation
to European Union, that process should be considered
not only as an economic attempt but also the very sub-
stantial part of the westernization and social development
project. After the World War II, one year later than the
Treaty of Rome entered into force and European Eco-
nomic Community was legally established on 1 January
European Economic Community. That time strikingly
was approximately fifteen days later than the Greece
applied for association with the EEC. In spite of a delay
due to the military intervention occurred in 1960, after
four years lasting difficult negotiations with EC, Turkey
signed “Ankara Agreement” on 12 September 1963. That
was the accession agreement into the EEC and creating
an association between The Republic of Turkey and Eu-
ropean Economic Community. This agreement went into
force the following year on 12 December 1964. The fin al
goal of the Ankara Agreement was to set up a customs
union between the parties and after a very long standing
process, unlike the other participant countries, Turkey
signed a customs union coming into effect on 1st of Jan-
uary, 1996.
Although Turkey has not been yet adopted as a full
member of the European Union, Her entrance into the
customs union raised some heavy debates on its future
results. These debates consist of especially economic
questions rather than social and logical aspects of the
customs union. For example some argue that the accep-
tance of the customs union without being full member of
the Union is zero sum game that EU will gain and Tur-
key will lose at the end. On the other hand, being tied
with the EU’s common customs tariffs, is another nega-
tive aspect of the customs union that lays a heavy burden
N. DEMIRCI ET AL.
Copyright © 2011 SciRes. ME
133
on the bac k of the Tur ke y1.
2. Literature Review
According to the estimates of Harrison, Rutherford and
Tarr [2] the gains to Turkey of the customs union with
the EU are between 1 - 1.5 percent of its GDP per year
depending on the complementary policies adopted by
Turkey. These gains are recurring, in the sense that they
can be expected each year, and they take into account the
costs of imposing higher taxes to compensate for the
revenue loss of the tariff. In their point of view, since
Turkey is complementing its tariff elimination on EU
imports with tariff reductions on imports from third
countries, Turkey will become a rather open economy in
the non-agricultural sectors. This complementary tariff
reduction to third countries should be regarded optimis-
tically since it reduces the trade diversion costs of the
customs union, and results in additional gains from trade.
Bekmez [3] to measure gains and losses from cu stoms
Union used a single country, multi-sector computable
general equilibrium (CGE) model with implicit inclusion
of the EU and ROW (Rest of the World), and He as-
sumed various policy scenarios. He divided the Turkish
economy into twenty-two sectors: two agricultural, eight-
een manufacturing, and two service sectors. An impor-
tant specification of the study was to consider the differ-
entiation of imports and exports as EU and ROW. His
simulation results showed that the imports and e xports of
Turkey have changed in favor of the EU under tariff re-
duction policies. The rest of the world’s trade with Tur-
key also tends to increase due to preferential trade agree-
ments with the non-EU countries. Under the customs un-
ion scenario, a 2 percent decrease in GDP and an 8 per-
cent decrease in government revenue will be experi-
enced.
Diao et al. [4] developed an intertemporal, multi-sec-
toral general equilibrium model to analyze the effects of
rising fiscal debts and trade liberalization on foreign trade,
capital accumulation and growth rate in the context of
Turkeys post-1990 experience. Their results suggest that
imbalances in the government fiscal accounts cause a
contraction of sectoral outputs and real GDP beyond the
levels expected from trade liberalization. They found that
the longer the delay in fiscal policy adjustment, the more
harmful the tariff liberalization will be. Tax adjustment
neutralizes the effects of tariff liberalization, but invest-
ment and imports are stimulated due to decreases in tariff
rates. As a result, the level of consumption increases.
This expands the trade deficit, thus increasing foreign
capital inflows. Because Turkey has a comparative ad-
vantage in the manufacturing and service sectors, the net
ex-ports of Turkey in these two sectors tend to increase.
This growth in exports w ill be faster than that of imports
after the eighteenth period. As a result, under the first
experiment, the economy as a whole will enjoy welfare
gains from liberalization. The steady-state capital stock
increases by 14.5 percent and consumption by 2.2 per-
cent with respect to the pre-reform equilibrium. The wel-
fare gain in the first ten periods is 0.16 percent, and
reaches 0.71 percent by the end of the thirtieth period.
However, in the second and third experiments, the results
differ. Turkey suffers from fiscal problems due to the
absence of compensating revenue sources. In the second
experiment, welfare losses will be seen in the first ten
periods, but these losses disappear over time. However,
in the third scenario, the welfare losses resulting from
tariff harmonization become worse over time.
Akkoyunlu-Wigley et al. [5] searched the effects of
the CU on the pricing behavior and market structure of
the Turkish manufacturing industry sectors and they es-
timated price cost mark up and Herfindahl concentration
ratio equations (HERF). The y used for estimation a panel
data set of 12 manufacturing industry sub-sectors for the
period 1994 to 2000. According to their findings the vo-
lume of manufacturing industry trade with the EU has
actually increased on average. They found that increased
trade volume with the EU during the CU period, created
a positive effects upon the Turkish economy through the
channel of imports, more specifically, by way of in-
creasing the competitive pressure to reduce markups and
the market power. Moreover, the rising export perform-
ance of the manufacturing industry vis-à-vis the EU is
likely to be the effect of falling mark-ups. Hence, sig-
nificant welfare gains seem to have followed the changes
in the pricing behavior and the market structure.
The economic theory of customs union is complicated
since this theory combines two reverse results. On the
one hand, while customs unions creating a new trade
with the members of the union; on the other hand, di-
verts t he trade with low er producer co untries whi ch will
remain outside the union after formation of the customs
union2. Freer trade and more protectionism are together
in customs union. Therefore, welfare effect (net out-
come) will be the combination of these two opposite
resul ts [7 ].
1In fact, whether
individual countries necessarily gain by entering a
customs union (CU) or, world welfare is higher under a CU is un-
proven and there is no explicit c
onfirmation. See for more detailed
discussion about this subject to Abrego at al. [1].
2A growing number of governments have begun to participate in re-
gional trade agreements. Th ese agreements enable member coun tries to
make preferential trade in the region. A great quantity o
f country is or
in the search of being a member of such integration and some 55 to 60
of world trade occurs within such trading blocks. For more information
see Schiff and Winters [6].
N. DEMIRCI ET AL.
Copyright © 2011 SciRes. ME
134
3. Overview of Economic Relations between
Turkey and EU
The Association Agreement between Turkey and EC put
forward two main goals to be fulfilled by Turkey. F ir s tly,
the Agreement envisaged three stages that Turkey had to
achieve [8]. These stages can be counted as preparatory,
transitional and final stages.
The preparatory stage covers 1963-1970 intervals and
aims to prepare Turkey to the EEC. This process has
been planned as a period in which the Community would
provide unilateral concessions and financial aid to Tur-
key while Turkey would take appropriate actions to im-
prove its economy and to prepare itself for the transi-
tional stage.
The frame of subsequent transitional stage was design-
nated by the additional protocol signed on 23 November
1970 and entered into force on 1 January 1973. That pe-
riod was going to be completed in two different catego-
ries between 12 and 22 years aiming finally to create a
customs union between EC and Turkey.
Although there is no a definite calendar for it, the
agreement also included the possibility of a third final
stage which would bring Turkey to full membership.
Accession was not an automatic process and dependent
on the realization of some concrete objectives.
Second main goal depends on the Additional Protocol.
According to the related Articles of the Protocol that
went into effect in 1970: 1) in connection with the Title 1
and article 2, free movement of goods will be provided, 2)
according to the Title 2, freedom of movement for work-
ers between Member States of the Community and Tur-
key shall be secured by progressive stages in accor-
dance with the principles set out in Article 12 of the
Agreement of Association between the end of the twelfth
and the twenty-second year after the entry into force of
that Agreement, 3) in relation to Title 2, EEC and Tur-
key shall determine the timetable and rules for the pro-
gressive abolition of restrictions on freedom of estab-
lishment and on freedom to provide services, 4) Closer
alignment of Turkish economic policies to EEC and fi-
nancial aids to promote Turkish economic development
shall be consider e d.
When the Additional Protocol is examined in a de-
tailed manner it is obvious that besides freedom to trade
of goods and services, it aims free movement of labor
and capital, freedom of establishment and to harmonize
Turkish economic policies, especially taxes, to EEC. An-
other prominent detail of the Association Agreement was
Turkeys acceptance of the EU’s Common External Tar-
iff (CET) in its trade with third and non-member coun-
tries. In order to reduce the negative effect of all these
policies financial aids was provided to Turkey.
The beginning of the 1980s is a turning point for the
Turkish economic system. With the 24 January decisions
in 1980, Turkey entered into a new economic period by
abolishing import substitution policies, which have been
implemented for a very long time, and initiating various
types of tools of outward-oriented economic policy. Tur-
key realized a comprehensive policy changes in all fields
of the state government with the abovementioned deci-
sions. These new liberal perspective in every stage of the
governance and a new set of economic policies resulted
in a very soon positive feedback for these policy altera-
tions and a considerable confidence to Turkeys economic
and democratic performance. Liberalizing economic po-
licies and wave of democratization encouraged Turkey to
apply for full membership to EU on April 14, 1987.
The application was in line with the Article 237 of the
Rome Treaty. According to that article “Any European
State may apply to become a member of the Community.
It shall address its application to the Council, which
shall act unanimously after obtaining the opinion of the
Commission”. After more than two and half years, the
Commission responded to Turkeys application on the
December 19, 1989. In this response the Commission
neither accepted nor directly rejected Turkey’s applica-
tion and underlined Turkey’s eligibility for membership
but since the implementation of the Single European Act,
and the objective of completion of the internal market by
the end of 1992, EC couldn’t initiate any new accession
negotiations before 1993 [8]. On the other hand, EU em-
phasized on some economic and political problems that
Turkey has to overcome before acquiring the full mem-
bership. Among some major economic problems there
were very major structural disparities in agriculture and
industry, macroeconomic imbalances, a high level of
industrial protectionism, a low level of social protection
[9] and some political issues were primarily Cyprus issue,
democratic weaknesses and human rights. Nevertheless,
in order not to leave Turkey out of the sphere of influ-
ence The Commissi on proposed to Turkey Cust oms Union.
4. The Pros and Cons of Customs Union
With a comprehensive economic reform performed at the
beginning of 1980’s, Turkey abandoned import substitu-
tion strategy that had been implemented more than two
decades. That was a genuine revolution for the Turkish
economy and thereafter an outward-oriented growth stra-
tegy was put into effect. As parallel with the 24 January
Decisions, before the financial liberalization, trade be-
came more liberal and as a first step of that liberalization
quantitative restrictions on import were elimin ated .
In compliance with the 24 January Decisions, Turkey
has made remarkable reductions on import tariffs of the
N. DEMIRCI ET AL.
Copyright © 2011 SciRes. ME
135
EU produced industrial goods by 1989. One of the most
effective reforms was made in January 1984. Subsequent
reforms were held in 1986 and in 1989. Econo my-wide
reductions on the customs duty were so large that the
nominal protection rate went down from 70 percent in
1984 to 40 percent in 1989 and to 28.25 percent in 1991.
That means a decline in effective protection rate from 79
percent in 1984 to 54 percent in 1989 [10]. Import ratio
to GNP has also increased from 1980 till 1990 and ac-
cording to calculation of Togan, the openness ratio wa s 9
percent in 1980 and that went up to 27.9 percent due to
rapid trade expansion by 2000 [11].
Before entering to the Customs Union Turkey had al-
ready been reducing tariffs on industrial goods produced
by European Union Countries and at the beginning of
January 1995, just before the year Customs Union was
signed with EU, over 80 percent of all Turkish imports or
taxes were free of duties. The implementation had al-
ready reached 90 percent of the target on the first twelve
year list and 80 percent of that on the second twenty-two
year list of the scheduled reductions [9].
After a difficult process, Customs Union agreement
was signed on 6 March 1995 between Turkey and EU
and that agreement entered into force as of 1 January
1996. According to the Decision No 1/95 Of the EC-
Turkey Association Council of 22 December, import or
export customs duties and charges having equivalent
effect shall be wholly abolished and quantitative restrict-
tions on imports and all measures having equivalent effect
shall be prohibit ed bet wee n the Comm unit y and Turkey.
Section 4 Article 13 explains another important issue
that Turkey had to align itself. According to this upon the
date of entry into force of this Decision, Turkey shall, in
relation to countries which are not members of the
Community, align itself on the Common Customs Tariff.
On the other hand, without being a full member of
European Union participating in Customs Union and
especially accepting the common external tariffs of EU
revealed some questions whether signing Customs Union
with EU is beneficiary or harmful. Another aspect of the
Customs Union was that Turkey had to implement the
European Unions Common Customs Tariff on imports
of industrial goods from third countries. Besides He has
also adopted most of the preferential trade agreements of
the EU. It has frequently spoken at the circles of scholars
in Turkey that due to not being ready for entrance to the
Union, removing all barriers even against to third coun-
tries Turkey would encounter a considerable trade loss.
To examine this question we develop a computable gen-
eral equilibrium model in the next section. In our study
we will especially concentrate on common external tariff
with rest of the world to measure potential loss or gain of
Turkey.
5. Modeling Framework and Simulation
Design
To analyze the potential impact of a common external
tariff in the context of customs union between EU and
Turkey, we have used counterfactual simulation with a
computable ge neral equilibr ium (CGE) mo del. CGE mod-
els atte mpt to turn the abstract models of general equ ilib-
rium theory into a practical tool for policy anal ysis. This
section provides a brief overview of the CGE approach,
the particulars of the model used in this paper.
Since CGE models attempt to capture the features of
the real-world economies, they incorporate data on the
structure of production and trade in the economy under
consideration. In general, starting point will be a national
input-output table and a set of trade matrices. These data
represent the state of the economy in question at one
point in time in the base year.
Having introduced the basic terminology and provided
overview of the way in which CGE models are con-
structed and used, we turn to brief overview of the prop-
erties of the model used in this paper.
5.1. The GTAP Model
The specific model used here is the Global Trade Analy-
sis project (GTAP) model. This is a publicly available,
mul tiregional CGE model that has been extensively used
in the literature.
The GTAP model3 is a standard multi-regional, static
3
GTAP stands for the Global Trade An alysis Project, a global network
of researchers and policy makers con
ducting quantitative analysis of
international policy issues. The standard GTAP Model is a multi-re-
-gion, multi-sector, computable general equilibrium model, with per-
fect competition and con stan t returns to scale. Bilat
eral trade is handled
via the Armington assumption and is implemented using
GEMPACK
(General Equilibrium Modeling Package). GEMPACK provides soft-
ware for calculating accurate solutions of an econo mic model, starting
from an algebraic representation of the equa
tions of the model. These
equations can be written as levels equation
s, linearized equations or a
mixture of these two Harrison, Pearson, [12
]. The GTAP dat abase, in its
latest version, describes the entire w orld economy as 113 countries (or
regions) in terms of 57 sectors, as well as all bilateral trade flows be-
tween these regions.
This database for ms the basis for a r ange of CGE
models that start from the same theoretical framework but are adapted
to addressing different economic contexts or research questions.
The
prominent role of GTAP in the trade p olicy debate has insp
ired further
developments of the da
tabase and models to deal with changes over
time (the regular model is static and thus does not pro
vide trajectories
of changes over time), international migration (capturing the flow of
persons and remittances between na
tions), energy use (capturing the
impact of bio-fuels in relation to developments in markets for non-re-
newable fuel) and climate change. For the latter, additional databases
are developed with more detail on land use (production by agro-eco-
logical zones in each region) and carbon sequestration. The latter de-
velopments have led to an increasing role of GTAP-
based analyses in
the Intergovernmental Panel o n Climate Change (IPCC) to assess po li-
cies for limiting greenho use g as emissions. The f ramewo rk of Stan
dard
GTAP model is well documented in chapter 2 of Hertel (1997) and
available on the Internet (http:// www.gtap.agecon.purdue.edu).
N. DEMIRCI ET AL.
Copyright © 2011 SciRes. ME
136
CGE where all markets are assumed to be perfectly com-
petitive and technologies exhibit constan t returns to scale,
model fully documented by Hertel [13].
The GTAP model defines consumers as having iden-
tical preferences that allocate income among private con-
sumption, government consumption and savings (Cobb-
Douglas). The single representative household then
maximizes a constant difference of elasticity (CDE) ob-
jective function. In each case consumption is of a CES
composite of domestically produced and imported goods.
The production in each identified sector and in each
identified region is represented by a nested constant elas-
ticity of substitution (CES) function. The model incur-
porates The Armington [14] assumption and as such,
each firm uses A CES composite domestically produced
and imported intermediate goods in fixed proportions
with a value added CES composite based on five en-
dowed factors of production (land, natural resources,
unskilled labor, skilled labor, and a capital).
In addition, this paper introduces a comparative static
long-run extension to the GTAP model, which treats
capital stock as an endogenous variable. This extension
consists of certain modifications to the structural form of
the model and the development of a new closure. It is
assumed that capital is mobile across regions and thus
that rates of return will approach the converging growth
rates in the steady state [15].
We adopt microeconomic closure that reflects the
choice of time frame. It is important to emphasize that
the frame element cannot be interpreted in terms of cal-
endar years, but rather in terms of the adjustments that
are allowed to take place in the (unobservable) transition
to a new equilibrium. In the model closure we adopt a
neoclassical approach.
5.2. Data and Aggregation
We used version 4 of the GTAP database which is de-
tailed in McDougall, Elbehri and Troung [16]. The full
database contains information on 66 regions and 50
commodities and has a base year of 1997 the simulations
presented here are based on an aggregated version of data
consisting 3 regions 7 com moditie s .
Within GTAP database, tariffs and other taxes are
represented as ad valorem equivalents of the actual ap-
plied rates. The GTAP database incorporates extensive
information on distortions in good sectors, including
trade taxes/subsides, output and input taxes/subsides, and
consumption taxes/subsides. The database does not at
present account for distortions in trade in services. Tar-
iffs and export subsidies are defined on a region-to-re-
gion basis, which allows the model to be easily used for
analysis regional trading arrangements. This also implies
that differing applied rates across regions reflect the dif-
ferent import compositions of trade within a given ag-
gregation.
We focus on the effects on the directly involved coun-
tries extending our analysis to the third countries in the
context of customs union with EU and Turkey. Sectoral
aggregation has been set up such as to allow us to pro-
vide a consistent picture of the effects of the CET for
both non agricultural commodities. Thus, the 57 GTAP
sectors have been aggregated into 7 representative ones
of which 3 are manufacture, whi l e the w orld consis t s of 3
regions (Table 1). Manufacture sectors are aggregated as
follows; light manufactures are unskilled labor intensive,
heavy manufactures are skilled labor intensive, and tech-
nical manufactures are capital intensive sectors. The
baseline has been updated to the year 1997 since the im-
port tariffs for CET have to be captured in the context of
custom union with Turkey. Sectoral aggregation strategy
in detail was given in Table A1 in Appendix.
5.3. Simulation Procedure and Design
Simulations have been designed as to represent key pol-
icy elements of the EU enlargement process and they
correspond to the cumulative effects of the trade liber-
alization. Thus simulation 3 depicting the most complete
set of policy shocks applied in our simulations:
1) CET: adoption of the Common External Tariff by
Turkey and Third countries.
2) FT: free trade between the EU and Turkey.
3) CET + FT: full trad e scenario —summarizes in one
comprehensive outcome results for the two pre-
vious cases.
We first simulate the formation of a common external
tariffs between EU and third countries (outside in the
Union) by adopting the these bilateral tariffs between
Turkey and third countries from the base year of 1997,
holding all other distor tion levels in t he system constant.
Because of the likely sensitivity regarding agriculture in
Table 1. Aggregation.
Sectoral Aggregation Regional Aggregation Production
Factors
Agriculture Third Countri es Land
Natural resource EU Unskilled
Food manufacture Turkey Skilled labor
Light manufacture Capital
Heavy manufacture Natural
Resources
Technical manufacture
Services
N. DEMIRCI ET AL.
Copyright © 2011 SciRes. ME
137
the agreement, we also simulate the removal of bilateral
tariffs excluding those in agricultural sector. The scena-
rios are all run as comparative static simulations. We again
emphasize that this gives us information on possible end
outcomes, but not on transition path, and that the time
element is represented by the alternative microeconomic
closures.
The Common External Tariff (CET) settlement
process generated a tariff structure greater, in some cases,
and smaller, in others, than in the previous situation. So,
the adopting CET by the year of 1997 will cause a diffe-
rentiated impact among commodities, depending on the
tariff structure. Table 2 shows the “shocks” (variations)
needed in the 1997 tariff structure to reach the CET val-
ues for EU in the last column, the values of the CET it-
self. Table 2 below shows the tariff structure in 1997
respectively Turkey and EU imports from third countries
(last second and third column). The values in these tables
are the power of the tariffs, or the relation between im-
port values at domestic prices and the same values at
world prices, CIF.
In the second simulation (FT), bilateral imports tariffs
have been eliminated in non agricultural commodities.
In third simulation, it will be simulated the effects of
trade liberalization and settlement of a Common External
Tariff (CET) between Turkey and the EU. As noted be-
fore, this simulation seeks to approach the results for EU
as a whole. In this sense, what is going to be simulated
here is the elimination of import tariffs between Turkey
and the EU and the settlement of the CET for the third
countries, at value levels showed in Table 2. The expe-
riment comprises both the elimination of import and
common external tariffs against third countries. The ta-
riff structure will be modified adequately, and a new
equi- librium will be computed.
As one can see, the tariff changes needed to adjust the
EU’s tariff structure to the CET values that will be mar-
kedly different between products, as it stated before. As
an example, Turkey would need a 0.3% decrease in its
food manufacture import tariff from third countries, on
the contrary, the 5.4% increase in its light manufacture
import tariff from the third countries.
6. Results
The regional percentage change in export and import in
Table 3 confirms the presence of trade creation in trade
between Turkey and EU. Turkey increases by 6.1% ex-
ports to third countries and increases by 3.9% imports
from the third countries. Both exports and imports in-
crease more than 10% due to the customs union. There-
fore there is a large percentage increase in exports and
imports between Turkey and third countries by adopting
of CET.
The most important trade indicator showed in Table 3
is changes in balance of trade defined as the value of
exports at world prices less the v alue of imports at world
prices, expressed in millions of dollars. Clearly, this in-
dicator is relevant only when the closure allows flexibil-
ity of the current account (otherwise it is zero by defini-
tion). The results indicate that Turkey’s current account
position further worsens when combined with CET in the
long run. However table indicates improvement in the
trade balance for EU in customs union between EU and
Turkey (90.7 million dollars) and deterioration adopting
common external tariffs for Turkey (-34.8 million dol-
lars). Nevertheless, the value of current account position
of FT with CET is still remains positive (63 million dol-
lars).
Table 4 shows the percentage changes in the sectors in
Turkish economy as a result of CET and FT simulation
outcomes. These changes are driven almost exclusively
Table 2. Bilateral imports for regions and import tariff rates.
Bilateral imports at Market prices
(Million $) Bilateral imports at World prices
(Million $) Import Tarif fs % change
for CET
Turkey-Third
Cont. EU-Third Cont. Turkey-Third
Cont. EU-Third Cont. Turkey EU
Agriculture 2 641 41 684 2 202 36 684 19.9 13.6 -31.6
Natural Resource 5 524 115 483 5 505 114 974 0.3 0.4 28.3
Food Manufacture 1 068 37 796 829 29 357 28.8 28.7 -0.3
Light Manufacture 1 501 98 001 1 387 90 190 8.2 8.7 5.4
Heavy Manufacture 5 421 135 654 5 154 131 578 5.2 3.1 -40.2
Technical Manufacture 7 251 336 747 6 951 324 955 4.3 3.6 -15.9
Services 4 009 266 709 4 009 266 709 0.0 0.0 0.0
Source: GTAP version 5 database and own calcul ations.
N. DEMIRCI ET AL.
Copyright © 2011 SciRes. ME
138
Table 3. Long run macro and trade results.
CET FT CET + FT
Turkey Third Country EU Turkey EU Turkey Third Country EU
Real GDP % 1.16 0.010.03 3.71 0.005 4.870.010.02
Terms of Trade % –1.38 0.040.04 0.77 0.010.61 0.020.03
GDP deflator % –1.98 0.050.09 0.83 0.011.15 0.020.07
Import Volumes % 3.93 0.09 –0.1 10.58 0.07 14.60.02 0.03
Export Volumes % 6.11 0.060.06 10.67 0.08 16.7 0.03 0.01
Trade Balance Million $ –47.74 82.5734.83798 90.7845 782 63
Equivalent Variation Million$ 954 2 9232 673 6241 691 7 1961 0331 982
Sources: Simulation results.
Table 4. Implications of industry outputs of cet: long run.
CET FT CET + FT
Turkey Third Count. EU Turkey EU Tu rkey Third Count. EU
Agriculture 0.920.02 0.05 0.67 0.02 1.590.01 0.07
Natural Resource 6.710.05 0.08 3.410.02 10.120.05 0.07
Food Manufacture 1.110.02 0.03 1.9 0 3.01 –0.04 0.03
Light Manufacture 1.950.09 0.24 22.190.48 24.140.260.24
Heavy Manufacture –4.08 0.140.26 1.71 0.052.37 0.120.21
Technical Manufactu re 2.00 0.010.03 1.42 0.06 3.420.01 0.03
Services 1.47 0.000.02 3.38 0 4.850.020.02
Capital Goods 2.05 0.010.06 5.8 -0.01 7.840.040.06
Sources: Simulation results.
by the light manufacture textile, apparel, and natural re-
sources sector of forestry, oil, coal and gas. In CET and
FT simulation, output of light manufacture increases by
24.1%, and natura l resources increases by 10.1%. Capital
goods are the only other productive sector to show any
significant changes, and here output increases by 7.8%.
Output in the heavy manufacture decreases by 2.37%
even though there is an increase by 1.7% in custom un-
ion process in this sector.
The welfare measure used in the results is the equiva-
lent variation (EV) for each regional household, ex-
pressed in millions of 1997 US dollars. This can be in-
terpreted as the change in regional household income at
constant prices that is equivalent to the proposed change.
The results indicate that Turkey benefit in net welfare
terms under all three simulations. EU’s gains are a very
solid $691 million under customs union, a similar out-
come to the Third countries gains of $2,923 million in
case of CET.
The GTAP model allows us to separate different
components of a given welfare change into changes in
allocative effici ency, endowment effi ciency effect, chan ge
in terms of trade and investment/savings effect as shown
Table 5. The welfare impact of the adoption of CET for
Turkey is shown to be positive (gain) but, for third coun-
tries is shown to be negative (loss) .
The model also calculates an index of welfare derived
directly from the utility function, the Hicksian Equiva-
lent Variation (EV), as well as shows its decomposition
into parts. This variable, graphed in millions of 1997
US$, is obtained through the product of the initial in-
come times the percent variation in the “per capita” util-
ity, and can be decomposed in three effects: an allocative
effect (AE), endowment efficiency (EE) and a terms of
trade effect (TTE). This variable expresses the size of the
Hicksian compensation of a price variation. As can be
seen in Table 5, the EV is positive for both regions, and
considerably greater in third countries than in Turkey,
due to the greater size of the third country economies.
Analyzing the decomposition, it can be seen that the
N. DEMIRCI ET AL.
Copyright © 2011 SciRes. ME
139
Table 5. Welfare implications of CET (US$ Million).
CET FT CET + FT
Turkey Third Count. EU Turkey EU Turkey Third Count. EU
Allocative Efficiency (AE) 557602 99 1082 562 1181 –5 –40
Endowment E fficiency (EE) 6771232 1704 4828125 653221321357
Terms of trade effect(TTE) 1524924600 336 264264 925661
IS 165 84249 –5 –9 –254 179 75
Total 2 9232674 954 6241 691 719610331982
Source: Sim ula tion re sult s.
size of the endowment efficiency effect is greater than
total EV: 954 millions for Turkey. The terms of trade
contribution to EV, however, is negative: -US$ 600 for
Turkey. As noted before, there is a worsening in the
terms of trade in both countries after the adoption of
CET, due to the fall in export p rices. The summing up of
the effects result in a positive total EV for the region,
indicating an increase in aggregated welfare due to the
trade liberalization process. The third countries EV also
increase by an amount of US$ 2,923 million. Brazil, then,
would appropriate more than half of total EV generated
by the integration.
As it can be seen, Tables 6-8 show a different situa-
tion for Turkey and EU. Moreover, Tables 6-8 below
bring a direct measure for both trade creation and diver-
sion effect in Turkey and EU in the simulation. The val-
ues are calculated by taking the difference between the
si mulation and the base year import values in Turkey and
EU, by commodity and region. The last column of each
table shows the total variation in imports.
Table 6. Turkey. Change in import values (CET). Million of
US$.
Third
Countries EU countries Turkey Total
Agriculture –26 –5 0 –31
Natural Resource 352 250 0 602
Food
Manufacture –2524 0 –49
Light
Manufacture –296 247 0 –49
Heavy
Manufacture 4 3874317 0 69
Technical
Manufacture 4 4114451 0 –40
Services –6240 0 –103
Total 8 7418 341 0 399
Source: Sim ula tion re sults.
In case of the adoption of CET, the import values from
EU countries fall in the simulation. There is actually an
increase in import values from regions outside EU, what
could be seen as trade diversion. While the total imports
in Turkey from the EU would fall by US$ 8741 million,
the total imports from third countries implementing CET
would increase by an amount of US$ 8741 million (Ta-
ble 6).
In the second experiment, for Turkey, there is an in-
crease in imports of US$5728 millions from the EU. This
is due mainly to the increase in imports of heavy
(US$1020), and technical (US$2728) manufactures from
the EU (Table 7). The interesting point here is that all
manufacture products are more seriously “diverted” in
the trade with EU. This suggests a modification in the
route of trade insid e EU after the FT plus CET, with im-
port flows of manufactures being reoriented through EU
countries, and then to third countries (Table 8).
7. Conclusions
In this paper we have simulated the economic effects of
Table 7. Turkey. Change in import values (FT). Million of
US$.
Third
Countries EU
countries Turkey Total
Agriculture 232 64 0 295
Natural resource 284 43 0 328
Food manufacture295 631 0 337
Light manufacture59 1000 0 941
Heavy manufacture460 1020 0 560
Technical
manufacture –1233 2728 0 1495
Services 294 242 0 536
Total –1236 5728 0 4492
Source: Sim ula tion re sult s
N. DEMIRCI ET AL.
Copyright © 2011 SciRes. ME
140
customs union associated with common external tariff
between Turkey and EU. The main emphasis of the pa-
per is the effect that wou ld have on the Turkish economy.
The simulations were carried out with a GTAP computa-
ble general equilibrium model using the GTAP database
5.0, which takes common external tariff in line with Eu-
rope Agreements between the Turkey and third countries
into account. We distinguished between three scenarios
for evaluating the effects of integration. The baseline
integration scenario (the EUTurkey free trade area)
covers only reductions in trade barriers in the context of
customs union. Then, as a second scenario, we assumed
the common external tariff implementing to third coun-
tries. In the third scenario, we assumed both customs
union and common external tariff to analyze combined
effects of them.
This paper then uses the GTAP model to assess what
the likely benefits from full trade liberalization between
Turkey and the EU are likely to be. The results suggest
that there will be gains to Turkey of some $7196 million.
These gains come about exclusively through its own re-
duction in import tariffs to zero on EU imports, and this
increased welfare stems mostly from an increased alloca-
tive and endowment efficiencies. The EU and other trad-
ing partners lose in welfare terms. Although the customs
union and common external tariffs are not a zero-sum
game, one can argue that Turkey has won but the EU has
lost in the context of customs union plus common exter-
nal tariff.
We note that the GTAP model has not increased any
other domestic taxes neutralizing the reduced income the
Turkish government faces due to tariff reductions. In ad-
dition, the pro-competitive effects of internal market
have not been covered in the model due to the lack of
data availability.
Turkey has been benefiting from the noticeable tariff
liberalization as well as improved market conditions in
line with EU since it ratified customs union with EU.
Therefore, we can say contrary to what is expected and
argued in some circles before Turkey has become the net
beneficiary of the customs union.
Although before entering to the Customs Union Tur-
key had already been redu cing tariffs on industr ial goods
produced by European Union Countries, it can be con-
cluded that more stable macroeconomic conditions and
closer cooperation with the EU respecting to fully opera-
tion of technical and financial assistance can improve
Turkeys gains from the customs union associated with
common external tariff.
8. References
[1] A., Lisandro, R. Riezman and J. Whalley, “How Often
Are Propositions on the Effects of Regional Trade
Agreements Theoretical Curiosa?” Journal of Interna-
tional Economics, Vol. 68, No. 1, January 2006, pp. 59-
78. doi:10.1016/j.jinteco.2005.03.003
[2] G. W. Harrison, T. F. Rutherford and D. G. Tarr, “Eco-
nomic Implications for Turkey of a Customs Union with
the European Union: A Quantitative Based Policy Analy-
sis,” The World Bank International Economics De- part-
ment International Trade Division, Policy Research
Working Paper, No. 1599, May 1996.
[3] S. Bekmez, “Sectoral Impacts of Turkish Accession to
the European Union: A General Equilibrium Analysis,”
Eastern European Economics, Vol. 40, No. 2, April 2002,
pp. 57-84.
[4] X. Diao, L. R. Terry and A. E. Yeldan, “How Fiscal
Mismanagement May Impede Trade Reform: Lessons
from An Intertemporal, Multi-Sector General Equilibrium
Modal For Turkey,” The Developing Economies, Vol. 37,
No.1, March 1999, pp. 59-88.
[5] A. Akkoyunlu-Wigley and M., Sevinç “Effects of the
Customs Union with the European Union on the Market
Structure and Pricing Behaviour of the Turkish Manu-
facturing Industry,” Applied Economics, Vol. 38, No. 20,
2006, pp. 2443-2452. doi:10.1080/00036840500427643
[6] M. Schiff and L. A. Winters, “Dynamics and Politics in
Regional Integration Arrangements: An Introduction,”
The World Bank Economic Review, Vol. 12, No. 2, May
1998, pp. 177-195.
[7] R. G. Lipsey, “The Theory of Customs Unions: Trade
Diversion and Welfare,” In: J. Bhagwati, P. Krishna and
A. Panagariya, Eds., Trading Blocs: Alternative Approaches
to Analyzing, Preferential Trade Agreements, MIT, 1999.
[8] E. Erdogdu, “Turkey and Europe: Undivided but not
United,” Middle East Review of International Affairs, Vol.
6, No. 2, June 2002, pp. 40-51.
[9] B. Yilmaz, “The Relations of Turkey with the European
Union: Candidate Forever?” Center for European Studies
Working Paper Series 167, 2008.
[10] H. Olgun and S. Togan, “Trade Liberalization of Interna-
tional Trade in Turkey during the 1980s,” The Pakistan
Development Review, Part II, Winter 1989, pp. 949 -960.
[11] S. Togan, “Turkey: Toward EU Accession,” The World
Economy, Vol. 27, No. 7, July 2004, pp. 1013-1045.
[12] W. J. Harrison and K. R. Pearson, “Computing Solutions
for Large Ge ne ral Equi lib riu m Mo del s U sing GE MPACK,”
Computational Economics, Vol. 9, No. 2, 1996, pp. 83-127.
[13] T. W. Hertel, Global Trade Analysis: Modeling a nd Ap-
plications,” Cambridge University Press, Cambridge, 1997.
[14] P. S. Armington, “A Theory of Demand for Products
Distinguished by Place of Production,” IMF Staff Papers
16, 1969, pp.159-176.
[15] T. Walmsley, “Long-Run Simulations with GTAP: Illu-
strative Results from APEC Trade Liberalization”, GTAP
Technical Paper No. 9, 2008.
[16] R. A McDougall, A. Elbehri and T. P. Truong, Global
Trade, Assistance and Protection: The GTAP 4 Data
Base,Center for Global Trade Analysis, Purdue University,
West Lafayette, 1998.
N. DEMIRCI ET AL.
Copyright © 2011 SciRes. ME
141
Appendix
Table A1. Commodities aggregation strategy.
Code Aggregated commodities
AGRIC Paddy rice, Wheat, Cereal grains nec, Vegetab les, fruit, nuts, Oil seeds, Sugar cane, sugar beet, Plant-based fibers, Crop s nec,
Bovine cattle, sheep and goats, horses, Animal products, Raw milk Wool silk-worm cocoons, Bovine cattle, sheep and goat,
horse meat prods,
FOOD Meat products nec, Vegetable oils and f ats, Dairy products, Processed rice, Sugar, Food products n ec, Beverages and tobacco
products
EXTRACT Forestry, Fishing, Coal, Oil, Gas, Minerals nec, Petroleum, coal products
LITMNFC: Textiles, Wearing apparel, Leather products, Wood products,
HVYMNFC: Paper products, publishing, Chemical, rubber, plastic products, Mineral products nec, Ferrous metals, Metals nec,
TECHMNFC: Metal products, Motor vehicles and parts, Transport equipment nec, Electronic equipment, Machinery and equipment nec,
Manufactures nec
SVCES: Electricity, Gas manufacture, distribution, Water, Construction Trade, transport, Financial, business, recreational services,
Public admin and defense, education, health, Dwellings & services