European Trade Study Group - ETSG - Effects of Trade Liberalization between … · 2013. 4. 9. · Effects of Trade Liberalization between the EU and Ukraine in a CGE Model are small

Effects of Trade Liberalization between the EUand Ukraine in a Computable General

Equilibrium (CGE) Model

Miriam FREY12, Zoryana OLEKSEYUK-VIBER3

1University of Regensburg2Osteuropa-Institut Regensburg3University of Duisburg-Essen

Contents

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2 Literature overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

3 Model description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

4 Data and policy experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

5 The simulation results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

5.1 Aggregate effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

5.2 Disaggregate results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

5.3 Robustness and sensitivity analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

5.4 Comparison with other studies and critique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

6 Summary and policy implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

A Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

A1 Countries’ aggregation into trading regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

A2 Initial input and output structure of production sectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

List of Tables

Table 4.1: Shares for household disaggregation (in %) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

Table 4.2: Model elasticities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Table 4.3: Calculated import tariffs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Table 5.1: Aggregate results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Table 5.2: Disaggregate results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

List of Figures

Figure 3.1: Model structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Figure 4.1: Structure of Ukrainian commodity trade. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Figure 5.1: Regional structure of Ukrainian foreign trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Figure 5.2: Sensitivity analysis of welfare changes to the elasticities of substitution andtransformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2

Abstract

The establishment of the currently negotiated Free Trade Agreement (FTA) between the EU andUkraine is the next step towards Ukraine’s deeper integration into the world economy. Someeconomists argue that the classical form of a FTA based on the removal of tariffs would notinduce a significant rise of welfare and would not have a serious positive impact on trade flowsbetween Ukraine and the EU. We check this argument by calculating the effects of an importtariff elimination in a Computable General Equilibrium (CGE) model for Ukraine incorporatingthe changed economic conditions after Ukraine’s WTO accession. The results confirm that tariffelimination has no significant impact on the country’s welfare as the households’ consumptionremains either unchanged or is reduced slightly (-0.08% for non-poor households). This confirmsthat the elimination of strongly reduced tariff rates after Ukraine’s WTO accession generates nowelfare gains because of the already low level of protection. For this reason the country shouldaim for a deeper integration into the EU incorporating the elimination of non-tariff barriers.

JEL-Classification: C68, F13, F15, O52Keywords: Ukraine, EU, Trade, Integration, CGE

Effects of Trade Liberalization between the EU and Ukraine in a CGE Model

1 Introduction

After Ukraine’s accession to the WTO in 2008 the creation of aFree Trade Agreement(FTA) between Ukraine and it’s most important trading partner the European Union (EU)is the next significant and realistic step towards Ukraine’sdeeper integration into theworld economy. The negotiations on the deep and comprehensive FTA (FTA+) withinthe framework of the New Enhanced Agreement started in March2007. So far there havebeen more than 10 rounds of negotiation and additional rounds are planned for 2011.

Some economists argue that the classical form of a FTA, whichonly involves the re-moval of export and import duties, compared to a deep one, which additionally incor-porates the harmonization of the Ukrainian norms and standards, would not lead to asignificant rise of welfare and would not have a serious positive impact on trade betweenUkraine and the EU.4 The purpose of this paper is to check this argument by examin-ing the implications of a unilateral tariff elimination as apart of a simple FTA betweenUkraine and the EU. In particular, we calculate the effects of an import tariff eliminationin a Computable General Equilibrium (CGE)5 model for Ukraine.

The paper is organized as follows. The next section providesan overview of the existingliterature. The structure of the model is described in section 3 followed by the specifica-tion of the data source and the policy experiment. A detailedanalysis of the results isgiven in section 5 including robustness checks and criticalcomments. The last sectionconcludes with some policy implications.

2 Literature overview

The different forms of Ukraine’s integration into the worldeconomy are widely evaluated.Most previous studies are devoted to the WTO accession. In the framework of a standardCGE model Pavel et al. (2004) simulate the full WTO accessionof Ukraine including tar-iff reduction, improved market access and adjustments of domestic taxation and identifya significant welfare gain and an increase in real GDP. These findings are supported byJensen et al. (2005) who predict an overall welfare gain of 5.2% of Ukrainian consump-tion and a rise of real GDP by 2.4% in a modified model (e.g. somesectors produce underincreasing returns to scale). Kosse (2002) confirms that thetariff reduction is indeedthe most important part of the full WTO accession. She separately analyzes the impactof an import tariff reduction on national welfare and finds the WTO membership to bebeneficial for Ukraine.

Subsequent studies focus on Ukraine’s trade relations withthe EU, especially after theten Central and Eastern European countries joined the EU in 2004. An analysis of theFTA between Ukraine and the EU shows that the FTA+ would have astronger positiveimpact on Ukraine’s welfare compared to the simple one wherethe overall welfare effects

4See Shumylo (2007).5See Jones (1965), Shoven and Whalley (1992).

1


are small or even slightly negative.6 In a recent study Maliszewska et al. (2009) modelthe impact of the different FTAs between the five European Neighborhood Policy (ENP)countries (Armenia, Azerbaijan, Georgia, Ukraine and Russia) and the EU. The conclu-sions are similar to the ones in the previous study. Among theENP countries, Ukrainegains most from the simple FTA with a net welfare increase of 1.73%. But it could benefiteven more from a FTA+ (increase of welfare by 5.83%). Francois and Manchin (2009)study the same question for the CIS region and Ukraine as a country study. They findnegative real income effects for CIS and Ukraine (-0.83% and-2.12%, respectively) incase of the classical FTA simulation and a decrease of Ukrainian real income by 0.4%under the FTA+ scenario.

The main drawback of this as well as of all other studies on theeffects of the FTAbetween the EU and Ukraine is the use of out-dated tariff rates. This paper contributes tothe ongoing discussion as it shows the impact of unilateral tariff elimination incorporatingthe changed economic conditionsafter Ukraine’s WTO accession.

3 Model description

Our model is a modified version of the static CGE model of Pavelet al. (2004). In additionto the updated database the modifications include the creation of new trading regions andproduction sectors, the disaggregation of the representative household into four types andthe implementation of sector-specific capital. It is implemented in GAMS/MPSGE7 andconsiders 38 sectors, four types of households, the government, investments and ninetrading regions. The structure of the model is shown in Figure 3.1.

The supply side of the Ukrainian economy is characterized bythe assumptions of per-fect competition and constant returns to scale. There are four factors of production: skilledand unskilled labor (ls,i), capital (ki) and sector-specific capital. Labor and capital (ex-cept sector-specific capital in the state-owned mining (a04) and pipeline transportation(a24P)) are perfectly mobile across sectors. The top nest ofthe production function ischaracterized by a Leontief-type structure:

yi = min{V Ai, IDi,j}, (1)

whereyi represents the total output of sector i (including domesticsales and exports),IDi,j is the intermediate demand for good j by industry i, andV Ai is the value added thatis given by the Cobb-Douglas function:

V Ai = c k(1−

∑

s αs,i)i

∏

s

lαs,i

s,i , 0 ≤ αs,i ≤ 1,∑

s

αs,i < 1, c > 0. (2)

The subscripts denotes the two types of labor: skilled and unskilled. Intermediate in-puts are either produced domestically or imported. Each firmuses a CES composite ofdomestic and imported intermediate inputs.8 Producers maximize profits subject to their

6See Emerson et al. (2006) and Ecorys and CASE-Ukraine (2007).7See Rutherford (1999) and Boehringer et al. (2003).8See equation (4).

2


Figure 3.1: Model structure

Domestic demand

- indirect taxes on private, public and investment demand

- indirect taxes on intermediate demand

b b b

b b b

Region 1 Region 9

Domestic salesExports

Imports

- import tariffs by

Region 9Region 1Domestic output

Value Added

- direct tax on labor and capital

Intermediate demand

CapitalSector-specific

capital

(in a04 and a24P)

Unskilled

labor

Skilled

labor

esdm=5

etreg=3

etdx=5

esreg=3

s = o

s_VA=1

goods and regions

Source: own illustration

3


production technology.Each sector is assumed to produce a single homogeneous product, which can be sold

on domestic(Hi) or foreign(EXi) markets according to the constant elasticity of trans-formation (CET) function:

Yi = ψi(γHρEX

i + (1− γ)EXρEX

i )1

ρEX , ψi > 0, 0 < γ < 1, (3)

with ρEX = (σEX − 1)/σEX , whereσEX defines the elasticity of transformation betweendomestic output and exports (in GAMS-Code: etdx=5). Producers regard sales on domes-tic markets and exports as imperfect alternatives.9 The output price index of each sectori is determined by both domestic(pH,i) and export prices(pEX,i): p̂i = f(pH,i, pEX,i),and the export price is defined as the FOB world market price(pEX,i) multiplied with theprice of foreign exchange(pfx): pEX,i = pEX,ipfx. Demanded goods are either imported(IMj) or produced domestically(DJ) so that domestic supply(DSj) is described by theconstant elasticity of substitution (CES) function:

DSj = ψj(βDρIMj + (1− β)IMρIM

j )1

ρIM , ψj > 0, 0 < β < 1, (4)

with ρIM = (σIM − 1)/σIM , whereσIM defines the elasticity of substitution betweenimports and domestic goods (in GAMS-Code: esdm=5). This means that consumer pref-erences are modeled as Armington-style product differentiation. The domestic price indexof each goodj is determined by the domestic sales price(pD,j), the import price(pIM,j)

and the import tariff(τIM,j): pj = f(pD,j, pIM,j(1 + τIM,j)). The import price equalsthe CIF world market price(pIM,j) multiplied with the price of foreign exchange(pfx):pIM,j = pIM,jpfx.

The consumption side is represented by public consumption,investment and intermedi-ate consumption as well as by final consumption of households. A representative house-hold derives utility from consumption of goods and servicesand finances its total con-sumption by income from labor (

∑

s wsL) and capital endowments (rK) and by receivedtransfers from the government (TG

hh) and from abroad (T ahh). This means that the value of

total consumption of a representative household (ΣjCjpj(1 + τj))10 does not exceed theincome multiplied with the total share of consumption (θ, 0 < θ < 1):

∑

j

Cjpj(1 + τj) ≤ θ

[

∑

s

wsL+ rK + TGhh + T a

hh

]

(5)

The representative household of the model is disaggregatedinto four types according tothe poverty line and the place of residence11: non-poor urban and rural households, poor

9This assumption is based on Armington (1969). See also Dervis et al. (1982), p. 221-223, 226-227.

10Cj is the consumption of good j and τj represents the rate of consumption tax for goodj.

11The poverty line is calculated following the methodology of the Ministryof Economy of Ukraine (available in Ukrainian at http://zakon.rada.gov.ua/cgi-bin/laws/main.cgi?nreg=z0401-02).

4


urban and rural households. Non-poor households are endowed with both capital andlabor (skilled and unskilled) whereas poor households are only endowed with unskilledlabor. All households receive transfers from the government and pay taxes and socialcontributions. But only non-poor households receive transfers from abroad and save aconstant share of their income.

The government receives income from public capital endowments12 (rKp + rspKsp),revenue from direct (

∑

i τi(rki +∑

sws,ili)) and indirect taxes (∑

j τjpj(Cj + INVj +

IDj + Gj + EXj)), from import tariffs (∑

j,r τIM,j,rpIM,jIMj,r), transfers from abroad(T a

G) and from households (T hhG ). Direct taxes are modeled as sector-specific taxes on the

use of production factors (capital and labor). Indirect taxes, in contrast, are modeled asproduct-specific taxes on private (Cj), investment (INVj), intermediate (IDj) and public(Gj) demand as well as on exports (EXj). Import tariffs (τIM,j,r) are product-specific anddistinguished by region. Government’s income is used for savings (pinvSAV G), transfersto households (TG

hh) and to abroad (TGa ), and to provide public services13 (

∑

j pjGj). Thepublic budget constraint is given by:

rKp + rspKsp +∑

i τi(rki +∑

s ws,ili) +∑

j τjpj(Cj + INVj + IDj +Gj + EXj)

+∑

j,r τIM,j,rpIM,jIMj,r + T aG + T hh

G (6)

= pinvSAVG + TG

hh + TGa +

∑

j pjGj.

Aggregate investment is modeled as a Cobb-Douglas composite over all goodsj:

INV = ψ∏

j

INVφj

j , φj ≥ 0,∑

j

φj = 1, ψ > 0. (7)

The price index for one unit of the aggregate investment goodis given by:pinv = f(pj(1+

τj)). The sum of public (SAV G) and private savings (SAV hh) minus the current account(CA) balance equals the aggregate investment:

pinv(SAVG + SAV hh)− CA = pinvINV. (8)

Equilibrium is defined by zero profits for producers, balanced budgets for householdsand the government, and by market clearing for all goods and factor markets. For equal-ization of the balance of payments, it must be valid that the CIF value of imports togetherwith the current account balance and transfers from the government to abroad (TG

a ) areequal to the FOB value of exports plus transfers from abroad to households (T a

hh) and tothe government (T a

G):∑

j

pIM,jIMj + CA+ TGa =

∑

i

pEX,iEXi + T ahh + T a

G. (9)

The price of foreign exchange(pfx) is chosen as the numeraire.

12Including capital income in state-owned sectors with sector-specific capital (rspKsp):mining and pipeline transportation (a04 and a24P).

13Consumption levels of public services are determined by a Cobb-Douglas function.

5


4 Data and policy experiment

The base year of our analysis is 2007. The backbone of the model is formed by a So-cial Accounting Matrix (SAM)14 with 38 sectors. It was constructed with the data of theUkrainian National Accounts and Input-Output Tables for 2007 at basic and consumerprices (publications of the State Statistics Committee of Ukraine).15 A SAM must bea balanced matrix so that the row sums equal the corresponding column sums. As theempirically estimated SAM for Ukraine was not balanced in the first version (due to in-consistency of data sources), we used a few balancing items in order to match all rowsand columns.

Additional information on indirect taxes, subsidies and imports (separately for inter-mediate, private, public and investment demand) as well as information on services tradeflows are also taken from the publication of the State Statistics Committee of Ukraine.Labor remuneration is disaggregated with data from this source as well.

The consumption shares per household type and sector are calculated from the Derzhkom-stat16 household budget survey for 2007 covering more than 10,000 Ukrainian householdsand over 200 different commodity groups (COICOP classification). Using this data theshares of payments from households to government as well as the shares of transfers fromgovernment to poor households in their total expenditures are computed. The respectivefigures are listed in Table 4.1.

Table 4.1: Shares for household disaggregation (in %)type of household (h) non-poor urban non-poor rural poor urban poor ruraldivision of transfers from house-holds to government∗

74 14 2 10

shares of transfers from govern-ment in total expenditures

35 35

Source: own calculations∗Transfers include taxes and social contributions.

All elasticities of substitution and transformation are borrowed from Pavel et al. (2004)and presented in Table 4.2. Data on Ukrainian commodity trade flows are drawn from theUnited Nations Commodity Trade Statistics Database (Comtrade). These data were ag-gregated into 17 (b01-b017) commodity groups. We used different correspondence tablesto convert the data from the HS96 into the KVED classification(Ukrainian classifica-tion which is based on NACE Rev.1). Ukraine’s exports and imports were grouped intothe following nine trading regions: EU-15, EU-12, other Europe, Asia, Africa, America,Commonwealth of Independent States (CIS), Russia and the rest of the world (ROW).

14See Pyatt and Round (1985).15Concerning the sectoral structure two changes were made in the SAM compared tothe original Input-Output Table. The heat supply sector was added to the electric en-ergy sector (a17) and the pipeline transit of oil and gas (a24P) was separated from thetransportation sector.

16The State Statistics Committee of Ukraine.

6


Table 4.2: Model elasticitiesParameter Value Descriptions 0 Elasticity of substitution between value added and intermediate inputss_VA 1 Elasticity of substitution between primary factors: capital and labouresdm 5 Armington elasticity of substitution between imports and domestic goodsetdx 5 Elasticity of transformation between domestic production and exportsesreg 3 Elasticity of substitution between import originetreg 3 Elasticity of transformation between export destination

Source: Pavel (2004), p. 4.

The first eight regions include countries representing the key trading partners of Ukrainewith all other countries being summarized as the rest of the world.17 Figure 4.1 illustratesthe trade structure of Ukraine in 2007 and a detailed description of countries’ aggregationinto trading regions is given in Appendix A1.

Figure 4.1: Structure of Ukrainian commodity trade

0 4 8 12 16 20

EU15

EU12

Other Europe

CIS

Russia

Asia

Africa

America

ROW

Exports in billions of US-$Imports in billions of US-$

Source: own calculations

Information on import tariffs is taken from the Law of Ukraine “About the CustomsTariff of Ukraine” including all amendments made due to Ukraine’s accession to the WTOin 2008. The law includes three types of tariff rates (ad valorem, specific and mixed).First, the ad valorem equivalents of the specific and mixed tariffs were calculated.18 Theresulting tariff rates were transformed from the HS2000 into the KVED classificationusing again correspondence tables and applying different averages (simple, weighted,

17Exports and imports for the ROW region are obtained as a residual.18Following WTO et al. (2007), p.187-188.

7


Table 4.3: Calculated import tariffsSector SAM code Import-weighted MFN tariff∗

Agriculture b01 5,63Forestry, logging and related service activities b02 1,71Fishing b03 5,00Mining of coal and peat b04 0,00Production of hydrocarbons b05 0,50Mining and quarrying b06 2,23Food-processing b07 13,66Textile industry b08 8,06Wood industry b09 0,98Manufacture of coke products b10 1,61Petroleum refinement b11 1,64Chemical industry b12 3,71Other non-metallic products b13 7,07Metallurgy, metal processing b14 1,93Machine-building b15 3,09Other production b16 1,85Electric energy b17 3,50

Source: own calculations∗These tariff rates apply to all trading regions except for Russia and CIS.

import-weighted). Table 4.3 shows the calculated import tariffs. With an import-weightedMFN tariff rate of 13.66 percent the food-processing, beverages and tobacco sector is themost protected one.

Different trade regimes are included in the model. Commodity trade with Russia andother CIS countries is classified as free trade because of theexisting FTA between Ukraineand the CIS countries.19 The MFN status is applied to trade with all other regions as theincluded countries are either members of the WTO or have bilateral trade agreements withUkraine to establish this trade regime.

As our purpose is to quantify trade liberalization effects between Ukraine and the EUthe counterfactual consists in the elimination of the import tariffs in all commodity groupsfor two regions in the model: EU-12 and EU-15. For all other regions the estimated tariffrates are still valid.

5 The simulation results

The results of our comparative static evaluation of the tariff elimination between Ukraineand the EU describe the full adjustment of the Ukrainian economy after the external shockof tariff elimination. This process is typically understood as a medium-term perspectiveover 7-10 years. Moreover, according to the CGE modeling framework the estimatedresults represent the isolated impacts of the trade liberalization on Ukraine’s economy.The possible effects of all other events affecting the economic development (e.g. changes

19The FTA was established in 1999.

8


Table 5.1: Aggregate resultsBenchmark Simulation of tariff elimination

Welfare per household type (Hicksian welfareindex, change in %):- Urban households - -0.07- Rural households - -0.09- Urban poor households - 0.00- Rural poor households - -0.01Consumption per household type (UAH bn):- Urban households 273.128 272.945- Rural households 96.059 95.971- Urban poor households 33.717 33.717- Rural poor households 26.715 26.712Price index for households’ consumption com-posites (change in %):- Urban households - 0.07- Rural households - 0.10- Urban poor households - 0.05- Rural poor households - 0.08Real GDP (change in %) - 0.00Tariff revenue (share of public budget, in %) 4.03 1.65Indirect tax rate (weighted average, in %) 13.15 13.70Real factor return (change in %):- Return to capital - -0.08- Return to sector-specific capital in mining(a04)

- 0.74

- Return to sector-specific capital in pipelinetransit (a24P)

- 0.00

- Wage rate for unskilled labor - 0.07- Wage rate for skilled labor - 0.08Aggregate exports (UAH bn) 323.205 328.438Aggregate imports (UAH bn) 364.373 369.606Aggregate exports (change in %) - 1.62Aggregate imports (change in %) - 1.44

Source: authors’ simulation

in energy and commodity prices, exchange rates, factor productivity, etc.) are not con-sidered. All results of our policy experiment reflect changes of the respective variablesrelative to the benchmark year 2007.

5.1 Aggregate effects

Economy-wide results of the counterfactual experiment areillustrated in Table 5.1. Theelimination of tariffs causes a fall of relative import prices and a reduction of the tariffrevenue from 4.03% to 1.65% of the Ukrainian public budget. In order to compensatefor these revenue losses the government raises the indirecttax rate from an average levelof 13.15% to 13.70%. The last effect outweighs the decline ofrelative import pricesand leads to a minor rise of the consumer prices as the price indices for households’

9


Figure 5.1: Regional structure of Ukrainian foreign trade


consumption composites increase.Tariff elimination causes a reallocation of the productionfactors across the sectors and

an appropriate shift in production levels20 while the real GDP remains broadly unchanged.A shift in the production factors demand leads to a decrease of the return to capital by0.08%21, while the resulting labor remuneration grows slightly by 0.07% for unskilledand by 0.08% for skilled labor which means that capital losesfrom the tariff elimination.These results confirm the deepening of Ukraine’s specialization in production of labor-intensive goods after the trade liberalization.22

Concerning welfare we have two opposite effects. The increase in labor remuneration isexpected to stimulate consumption. In contrast, the higherconsumer prices and reducedreturn to capital should have a negative impact. In fact, thenegative effects predominateand our simulation suggests no change (for urban poor households) or even a small reduc-tion of consumption by nearly 0.08% for non-poor and by approximately 0.01% for ruralpoor households. The stronger negative welfare effect for non-poor households is causedby their higher tax burden compared to the poor household types23.

The strongest effect of the tariff elimination occurs on foreign trade flows of Ukraine.The country’s imports rise by 1.44% and stimulate an increase of exports by 1.62% asadditional imports need to be purchased with foreign currency. Despite of changes inaggregate imports and exports, the fundamental trade structure of Ukraine with the model-

20See Table 5.2.21The return to sector-specific capital in mining (a04) grows by 0.74%.22Following the Heckscher-Ohlin and Stolper-Samuelson theorems, see Feenstra (2004),p. 15, 32, 174.

23See Table 4.1.

10


specific regions remains almost unchanged as illustrated inFigure 5.1. Nevertheless, theremoval of import tariffs between Ukraine and the EU leads toan increase of imports fromthe EU member countries (EU-15 and EU-12) by 1.37 percentagepoints (from 38.44% to39.81%) while the import shares of all other regions decline. The strongest fall in importshares is observed for Russia (by 0.52 percentage points). The results for export structuresuggest a slight decrease of exports to the EU and CIS countries and demonstrate a smallrise of export shares for other regions except for Africa andROW.

5.2 Disaggregate results

Table 5.2 illustrates the sector-specific results of the EU-Ukraine trade liberalization sim-ulation. The elimination of import tariffs causes a reduction of unit costs in almost allproduction sectors.24 It strongly favoures Ukraine’s chemical and textile industries, met-allurgy, mining and quarrying and manufacture of coke products. These activities expe-rience the strongest output increase while a noticeable rise of production is observed insuch sectors as wood industry, mining of coal and peat and other production. The outputincrease occurs in these sectors because they are relatively unprotected in the benchmark(Table 4.3.) and benefit from lower prices for intermediate goods which take over 50% oftheir total inputs (see Appendix A2). Moreover, these winning sectors (except for manu-facture of coke products and mining of coal) are export-oriented (see Appendix A2) andgain additionally from trade liberalization because the tariff-elimination-induced demandfor imports leads to a devaluation pressure and, consequently, to a stimulation of exports.In addition to the aforementioned activities, hotels and restaurants benefit mostly amongthe services because this sector is initially unprotected,exports nearly 51% of its outputand gains from the elimination of the highest import tariff (13.66%) in the food industry(i.e. cheaper intermediate inputs). On the other hand, food-processing and production ofnon-metallic mineral products, agriculture, fishery and petroleum refinement reduce theiroutput because of a high initial level of protection and low export shares.

The results on factor and intermediate demand are consistent with the output changes.The sectors with extended production after the simulation raise their factor and intermedi-ate demand as the rise of output needs an increased factor andintermediate input. On thecontrary, demand for production factors and intermediate products declines in the sectorslosing from trade liberalization25.

The developments of exports and imports also reflect the results for the output changes.Tariff removal leads to a rise of imports in all sectors (up toelectric energy, but except pro-

24A minor increase of unit costs is observed in such sectors as forestry, mining of coaland peat, manufacture of coke products and petroleum refinement. This arises becauseof increased factor returns (e.g. labor or sector-specific capital) and increased indirecttaxation of intermediate inputs.

25The strongest fall of factor and intermediate demand is observed in food-processingand production of non-metallic mineral products, agriculture, fishery and petroleum re-finement.

11


duction of hydrocarbons26) with initially high protected food-processing, production ofnon-metallic mineral products and agriculture on the top. This increase of import demandis accompanied by a rise of exports in chemical and textile industry, metallurgy, miningand quarrying, machine building, wood industry and other production. In contrast, sectorsas petroleum refinement, food-processing, production of non-metallic mineral products,agriculture and fishery reduce their exports. Concerning foreign trade in services, thechanges in imports and exports are small as all service activities are unprotected in thebenchmark equilibrium. Nevertheless, hotels and restaurants, and construction27 consti-tute exceptions with a rise of exports by 1.34% and 1.10%, respectively.

26We observe a slight decrease of imports in the sector of hydrocarbons (a05) becauseof price changes in this sector: the relative import price of hydrocarbons remains un-changed while the relative domestic supply price decreases.

27Construction gains from the elimination of import tariff for non-metallic mineral products(initial value 7,07%) which allows for higher output and exports.

12

Effects

ofTradeLiberalization

between

theE

Uand

Ukraine

ina

CG

EM

odel

Table 5.2: Disaggregate results

Sector

Changes relative to benchmark (in %)Capital Skilled Unskilled Intermediate Unit

Output Exports Imports demand labor labor demand costsdemand demand

a01 Agriculture -1,34 -0,78 6,28 -1,31 -1,46 -1,45 -1,34 -0,11a02 Forestry 0,14 0,07 2,99 0,27 0,12 0,13 0,14 0,02a03 Fishing -1,22 -0,48 0,02 -1,12 -1,28 -1,27 -1,22 -0,15a04 Mining of coal and peat 0,49 0,05 1,00 0,66 0,67 0,49 0,09a05 Production of hydrocarbons 0,13 0,38 -0,13 0,19 0,03 0,04 0,13 -0,05a06 Mining and quarrying 1,25 1,83 0,86 1,32 1,16 1,17 1,25 -0,11a07 Food-processing -2,49 -1,32 12,58 -2,41 -2,57 -2,56 -2,49 -0,24a08 Textile industry 2,85 5,24 2,48 2,93 2,77 2,78 2,85 -0,46a09 Wood industry 0,47 1,99 0,06 0,57 0,42 0,43 0,47 -0,30a10 Manufacture of coke products 1,17 1,11 2,15 1,22 1,06 1,07 1,17 0,01a11 Petroleum refinement -1,15 -1,52 0,91 -1,05 -1,21 -1,20 -1,15 0,08a12 Chemical industry 2,80 5,53 1,37 2,88 2,72 2,73 2,80 -0,52a13 Other non-metallic products -1,95 -1,11 9,10 -1,86 -2,01 -2,00 -1,95 -0,17a14 Metallurgy, metal processing 1,47 2,11 1,11 1,56 1,41 1,42 1,47 -0,13a15 Machine building 0,09 1,94 1,01 0,19 0,03 0,04 0,09 -0,37a16 Other production 0,49 1,90 1,27 0,61 0,45 0,46 0,49 -0,28a17 Electric energy, heat supply 0,21 0,12 0,30 0,29 0,14 0,15 0,21 0,02a18 Gas supply 0,19 0,24 0,14 0,32 0,16 0,17 0,19 -0,01a20 Water supply -0,10 -0,58 0,40 0,05 -0,10 -0,09 -0,10 0,10a21 Construction 0,21 1,10 -0,69 0,32 0,16 0,17 0,21 -0,18a22 Trade and repair activities -0,02 0,18 -0,26 0,04 -0,12 -0,11 -0,02 -0,04a23 Hotels and restaurants 0,83 1,34 -0,75 0,89 0,73 0,74 0,83 -0,10a24 Transport -0,03 -0,03 -0,05 0,06 -0,09 -0,09 -0,03 0,00a25 Post and telecommunications 0,16 1,00 -0,87 0,22 0,06 0,07 0,16 -0,17a26 Financial activities 0,04 0,13 -0,06 0,12 -0,04 -0,03 0,04 -0,02a27 Real estate activities 0,10 0,12 0,08 0,14 -0,02 -0,01 0,10 0,00a28 Renting 0,14 0,42 -0,19 0,17 0,01 0,02 0,14 -0,06a29 Computer and related activities 0,28 0,87 -0,52 0,35 0,19 0,20 0,28 -0,12a30 Research and development 0,17 0,28 0,01 0,30 0,15 0,16 0,17 -0,02a31 Other business activities 0,03 0,43 -0,42 0,12 -0,03 -0,02 0,03 -0,08a32 Public administration -0,04 -0,60 0,53 0,11 -0,05 -0,04 -0,04 0,11a33 Education -0,11 -0,81 0,60 0,03 -0,13 -0,12 -0,11 0,14a34 Health care and social assistance 0,04 0,05 0,03 0,17 0,02 0,03 0,04 0,00a35 Streets cleaning, other utilities -0,04 -0,46 0,38 0,08 -0,07 -0,06 -0,04 0,08a36 Social activities -0,12 0,03 -0,13 -0,12 -0,12 0,13a37 Leisure activities -0,17 -0,36 0,07 -0,06 -0,22 -0,21 -0,17 0,04a38 Other activities 0,02 0,16 -0,17 0,06 -0,09 -0,08 0,02 -0,03

a24P Pipeline transit -0,03 -0,03 -0,08 -0,07 -0,043 0,00negative changes strong positive changes

∗a35: sewage, refuse disposal; a37 includes recreational, entertainment, cultural and sporting activities.


13


5.3 Robustness and sensitivity analysis

To check the robustness of our results with respect to the underlying data and elasticityvalues we repeated our simulation with some changes. First of all, we conducted thecounterfactual experiment with the data for 2004 examiningwhether the benchmark year2007 was a «representative» year and whether the choice of another base year before theworld economic crisis would have led to significantly different results. This experimentshows that the difference between the results is slight or even negligible. In particular, thesimulation with the benchmark year 2004 suggests a reduction of consumption by nearly0.13% for both non-poor household types and by 0.07% for urban poor households.28

This confirms the robustness of our results and supports the general experience in theCGE modeling that the choice of the base year has a minor impact on the robustness ofsimulation results.29

For examining the sensitivity of welfare changes with respect to the elasticities of sub-stitution and transformation we ran simulations varying the values of elasticities from 0.5(weak substitutability/transformability) to 5 (high substitutability/transformability). Theresults of this analysis are illustrated in Figure 5.230 and suggest that the magnitude ofwelfare effects is sensitive to the different values of elasticities. Nevertheless, the choiceof different parameter values cannot lead to fundamentallydifferent results as the welfarechanges remain negative (up to -0.16%).31

5.4 Comparison with other studies and critique

The results of our comparative static evaluation of Ukraine’s trade liberalization are con-trary to the trade theory which predicts welfare benefits in case of import tariffs removal.Moreover, the represented results are not consistent with most studies32 on trade liberal-ization in Ukraine and other Former Soviet Union (FSU) countries as they predict positive

28The only difference occurs for rural poor households which increase their consumptionby 0.01% in comparison with the reduction by 0.01% before. The reason of such differ-ence is the rise of output in the sector of streets cleaning and other utilities (+0.07%) inthe simulation with data for 2004 comparing with the fall by 0.04% in the initial modelingexercise. The rural poor households gain from the output increase in this sector becauseit takes almost 30% of their consumption.

29See Jensen et al. (2005).30The results for different values of the elasticity of transformation between export desti-nations (etreg) are not illustrated in Figure 5.2 because they do not influence the welfarechanges of all household types.

31The only exception is the welfare effect for urban poor households which is positive(up to 0.08%) if the elasticity of transformation between domestic output and exports(etdx) equals 4.5 or less and if the elasticity of substitution between imports origin (esreg)equals 3 or more. The reason of such difference lies in the dissenting consumptionpatterns and factor endowments of urban poor households. In particular, the agriculturalgoods and food products take almost 62% of their aggregate consumption while unskilledlabor represents the only income source of this household type.

32See Pavel et al. (2004), Jensen et al. (2005), Harbuzyuk and Lutz (2008), Maliszewskaet al. (2009), Ecorys and CASE-Ukraine (2007).

14


Figure 5.2: Sensitivity analysis of welfare changes to the elasticities of substitution andtransformation

-.16

-.12

-.08

-.04

.00

.04

.08

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

for rural households

for rural poor households

for urban households

for urban poor households

-.16

-.12

-.08

-.04

.00

.04

.08

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5





-.16

-.12

-.08

-.04

.00

.04

.08

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5





Elasticity of substitution between imports and domestic goods

Elasticity of substitution between imports origin

Elasticity of transformation between domestic output and exports

Hic

ksia

n w

elfa

re in

de

x, ch

an

ge

in %

Hic

ksia

n w

elfa

re in

de

x, ch

an

ge

in %

Hic

ksia

n w

elfa

re in

de

x, ch

an

ge

in %

Source: authors’ simulations

15


welfare effects (3-5%) for the countries.33

This discrepancy between our research and other studies occurs because all previousevaluations of Ukraine’s trade liberalization are made with the data on import tariffs be-fore the country’s WTO accession. Though, simulating tradeliberalization through theelimination of strongly reduced tariff rates after Ukraine’s WTO accession causes no wel-fare increase as the protection level is already low. Hence,the unilateral tariff eliminationbetween Ukraine and the EU has no significant impact on the county’s welfare consider-ing the current constellation of import tariffs.

Despite of the mentioned discrepancy, our results of the almost unchanged or evenslightly fallen welfare level of Ukraine are supported by Emerson et al. (2006), Manchin(2004) and Francois and Manchin (2009). Emerson et al. (2006) analyzed the implicationsof the EU-Ukraine free trade using a multi-regional CGE model34 (without intermediateinputs) and also found in the simulation of the classic FTA35 a negative welfare change of-0.06%. Francois and Manchin (2009) found that the classical form of free trade leads toa decrease of Ukrainian real income by 2.12%. Moreover, a reduction of the Russian wel-fare level by nearly 1500 Mill. US-Dollars is predicted by Manchin (2004) for the simpleRussia-EU FTA incorporating tariffs removal on industrialgoods (excluding agriculture).

However, the results of our policy experiment reflect the typical «gains from trade» dueto improved allocation of production factors among sectorsafter the tariff eliminationwithout a consideration of additional gains that can accessorily raise the welfare level ofthe country. Such gains arise from the impact of tariff elimination on investments andinvestment climate, competition, product variety and productivity. Modeling of dynamiccapital adjustment and additional investments can increase the welfare effects of tradeliberalization and constitutes the subject of further research. Other criticism refers to notconsidered particularities of the Ukrainian economy such as unemployment, rigidities onfactor markets and economies of scale (e.g. in metallurgy, manufacture of machinery andequipment, etc.).

6 Summary and policy implications

The simulation of trade liberalization between Ukraine andthe EU shows that it has nosignificant impact on the county’s welfare as the poor and non-poor households slightlyreduce their consumption by 0.02% and 0.10%, respectively,and real GDP remains ba-sically unchanged. This means that elimination of stronglyreduced tariff rates afterUkraine’s WTO accession generates no welfare gains becauseof the already low levelof protection. Nevertheless, the import tariffs removal has a relatively strong impact onthe foreign trade flows of Ukraine. The country’s imports rise by 1.43% and exports by

33See section 2.34A modified model of Brenton and Whalley (1999).35The simulation of the simple FTA between the EU-15, Central and Eastern EuropeanCountries (CEEC), South-Eastern European Countries (SEEC) and Ukraine implies theelimination of all tariffs and the steel quota on Ukrainian exports.

16


1.63%. The removal of import tariffs between the partners leads to an increase of im-ports from the EU member countries by 1.37 percentage pointswhile the import sharesof all other regions decline. Industry-specific results show that chemical and textile in-dustry, metallurgy, mining and quarrying, manufacture of coke products, wood industryand other production strongly benefit from tariff elimination and increase their output andexports. On the other hand, initially high protected sectors such as food-processing andproduction of non-metallic mineral products, agriculture, fishery and petroleum refine-ment reduce their production and lose from the trade liberalization.

Our results confirm the conclusions of Ecorys and CASE-Ukraine (2007), Emersonet al. (2006), Maliszewska et al. (2009), Shumylo (2007) that tariff elimination as theclassical form of free trade is not profitable for Ukraine. The country should strive for adeeper level of integration into the EU incorporating not only the removal of tariff barriersbut also the harmonization of the Ukrainian regulatory norms, technical standards andthe rules of origin with the European ones. It ought also include the simplification ofthe customs procedures and the abatement of corruption related to Ukrainian Customs.Such a deep and comprehensive FTA, which is actually negotiated between the partners,should be more beneficial for Ukraine as the welfare increasecan reach 6% according tothe aforementioned studies.

17


References

Armington, P. (1969). A Theory of Demand for Products Distuinguished by Place ofProduction.Internationally Monetary Fund Staff Paper 16, 159–176.

Boehringer, C., T. Rutherford, and W. Wiegard (2003). Computable General EquilibriumAnalysis: Opening a Black Box.ZEW Discussion Paper No. 03-56.

Brenton, P. and J. Whalley (1999). Evaluating a Ukraine-EU Free Trade Agreement usinga Numerical General Equilibrium Model.Centre for European Policy Studies (CEPS).

Dervis, K., J. De Melo, and S. Robinson (1982).General Equilibrium Models for Devel-opment Policy. Cambridge.

Ecorys and CASE-Ukraine (2007). Global Analysis Report forthe EU-Ukraine TSIA,Ref. TRADE06/D01, DG-Trade.European Commission.

Emerson, M., T. Edwards, I. Gazizulin, M. Luecke, D. Mueller-Jentsch, V. Nanviska,V. Pyatnytskiy, A. Schneider, R. Schweikert, O. Shevtsov, and O. Shumylo (2006).The Prospect of Deep Free Trade between the European Union and Ukraine. Centrefor European Policy Studies (CEPS), Institut fuer Weltwirtschaft (IFW), InternationalCentre for Policy Studies (ICPS).

Feenstra, R. (2004).Advanced International Trade: Theory and Evidence. Princeton.

Francois, J. and M. Manchin (2009). Economic Impact of a Potential Free Trade Agree-ment (FTA) between the European Union and the Commonwealth of the IndependentStates.CASE Network Report No. 84.

Harbuzyuk, O. and S. Lutz (2008). Analyzing Trade Opening inUkraine: Effects of aCustoms Union with the EU.Econ Change Restruct 41, 221–238.

Jensen, J., P. Svensson, F. Pavel, L. Handrich, V. Movchan, and O. Betily (2005).Analysisof Economic Impacts of Ukraine’s Accession to the WTO: Overall Impact Assessment.Kyiv, Munic, Copenhagen.

Jones, R. W. (1965). The Structure of Simple General Equilibrium Models.The Journalof Politival Economy Vol.73, No.6, 557–572.

Kosse, I. (2002). Using a CGE Model to Evaluate Impact TariffReductions in Ukraine.National University of Kyiv Mohyla Academy.

Maliszewska, M., I. Orlova, and S. Taran (2009). Deep Integration with the EU and itsLikely Impact on Selected ENP Countries and Russia.CASE Network Reports.

Manchin, M. (2004). The Economic Effects of a Russia-EU FTA.Tinbergen InstituteDiscussion Paper TI 2004-131 /2.

18


Pavel, F., I. Burakovsky, N. Selitska, and V. Movchan (2004). Economic Impact ofUkraine’s WTO Accession: First results from a Computable General EquilibriumModel. Institute for Economic Research and Policy Consulting Working Paper, No.30.

Pyatt, G. and J. Round (1985). Social Accounting Matrices: ABasis for Planning.theWorld Bank.

Rutherford, T. (1999). Applied General Equilibrium Modeling with MPSGE as a GAMSSubsystem: An Overview of the Modeling Framework and Syntax. ComputationalEconomics 14 (1/2).

Shoven, J. and J. Whalley (1992). Applying General Equilibrium. Cambridge UniversityPress.

Shumylo, O. e. a. (2007).Free Trade between Ukraine and the EU: An impact assessment.Kyiv.

WTO, UNCTAD, and ITC (2007).World Tariff Profiles 2006. Switzerland.

19


Appendix

A1: Countries’ aggregation into trading regions

Region Country Region Country

1. EU-15 5. AsiaAustria ChinaBelgium IndiaDenmark IndonesiaFinland IranFrance IsraelGermany JapanGreece LebanonIreland South KoreaItaly SyriaLuxembourg TurkeyNetherlands United Arab EmiratesPortugal VietnamSpain JordanSweden MalaysiaUK Pakistan

SingaporeThailandSaudi Arabia

2. EU-12 6. AfricaEstonia AlgeriaLatvia EgyptLithuania TunisiaCzech Republic LibyaHungary GhanaPoland MaoroccoSlovakia NigeriaSloveniaCyprusMaltaBulgariaRomania

3. Other Europe 7. AmericaBosnia and Herzegovina ArgentinaCroatia BrazilMacedonia CanadaSerbia USASwitzerland MexicoNorway Br. Virgin IslandsAlbania

4. CIS 8. RussiaArmenia Russian FederationAzerbaijanBelarus 9. Rest of the worldGeorgia All other countriesKazakhstanKyrgyzstanMoldovaTajikistanTuekmenistanUzbekistan

Source: authors’ calculations

20

Effects

ofTradeLiberalization

between

theE

Uand

Ukraine

ina

CG

EM

odelA2: Initial input and output structure of production sectors

SectorInput (in %) Output (in %)

Intermediate Capital Sector-specific Skilled Unskilled Depreciation Total Domestic Exports Totaldemand demand capital demand labor demand labor demand sales

a01 Agriculture 58,23 31,63 0,00 3,55 5,68 0,92 100 92,45 7,55 100a02 Forestry 39,34 7,27 0,00 27,10 26,28 0,00 100 67,63 32,37 100a03 Fishing 72,08 9,94 0,00 8,22 9,76 0,00 100 97,22 2,78 100a04 Mining of coal and peat 38,85 0,00 15,98 25,71 19,27 0,18 100 93,48 6,52 100a05 Production of hydrocarbons 25,24 46,57 0,00 16,11 12,08 0,00 100 92,99 7,01 100a06 Mining and quarrying 51,85 27,40 0,00 10,62 7,96 2,16 100 73,09 26,91 100a07 Food-processing 73,86 10,95 0,00 7,35 5,51 2,34 100 78,33 21,67 100a08 Textile industry 47,51 20,68 0,00 12,07 9,05 10,69 100 28,80 71,20 100a09 Wood industry 73,36 9,03 0,00 10,07 7,54 0,00 100 71,20 28,80 100a10 Manufacture of coke products 77,11 13,87 0,00 4,00 3,00 2,02 100 95,79 4,21 100a11 Petroleum refinement 87,90 4,46 0,00 4,37 3,27 0,00 100 78,97 21,03 100a12 Chemical industry 78,83 9,92 0,00 6,43 4,82 0,00 100 46,33 53,67 100a13 Other non-metallic products 71,04 11,06 0,00 10,23 7,67 0,00 100 91,39 8,61 100a14 Metallurgy, metal processing 77,76 8,55 0,00 7,61 5,71 0,36 100 39,24 60,76 100a15 Machine building 73,47 9,43 0,00 9,77 7,32 0,00 100 55,08 44,92 100a16 Other production 70,34 6,69 0,00 10,27 7,70 4,99 100 76,60 23,40 100a17 Electric energy, heat supply 64,04 15,02 0,00 13,72 7,22 0,00 100 95,91 4,09 100a18 Gas supply 45,76 9,08 0,00 29,59 15,57 0,00 100 99,93 0,07 100a20 Water supply 60,50 0,68 0,00 25,43 13,39 0,00 100 99,61 0,39 100a21 Construction 68,41 8,89 0,00 11,67 11,04 0,00 100 99,22 0,78 100a22 Trade and repair activities 72,46 17,46 0,00 7,07 3,01 0,00 100 99,65 0,35 100a23 Hotels and restaurants 55,74 26,38 0,00 11,21 6,67 0,00 100 49,08 50,92 100a24 Transport 56,25 18,11 0,00 13,73 11,91 0,00 100 96,23 3,77 100a25 Post and telecommunications 47,58 30,69 0,00 14,08 7,65 0,00 100 90,10 9,90 100a26 Financial activities 31,18 36,05 0,00 29,96 2,81 0,00 100 96,10 3,90 100a27 Real estate activities 42,89 42,22 0,00 10,20 4,69 0,00 100 97,15 2,85 100a28 Renting 36,76 51,16 0,00 8,28 3,81 0,00 100 91,90 8,10 100a29 Computer and related activities 53,43 23,93 0,00 15,51 7,13 0,00 100 84,67 15,33 100a30 Research and development 22,43 9,26 0,00 53,82 8,18 6,32 100 79,97 20,03 100a31 Other business activities 51,90 18,89 0,00 20,01 9,20 0,00 100 93,41 6,59 100a32 Public administration 26,09 3,84 0,00 64,17 5,90 0,00 100 99,77 0,23 100a33 Education 26,91 7,25 0,00 54,39 11,46 0,00 100 99,46 0,54 100a34 Health care and social assistance 35,73 8,37 0,00 42,73 13,17 0,00 100 98,59 1,41 100a35 Streets cleaning, other utilities 55,59 7,38 0,00 20,22 16,81 0,00 100 99,50 0,50 100a36 Social activities 46,27 0,71 0,00 28,96 24,07 0,00 100 100,00 0,00 100a37 Leisure activities 51,07 15,93 0,00 26,53 6,47 0,00 100 89,84 10,16 100a38 Other activities 34,24 45,17 0,00 16,56 4,03 0,00 100 87,52 12,48 100

a24P Pipeline transit 81,24 0,00 9,93 4,73 4,10 0,00 100 0,00 100,00 100

Source: authors’ calculations

21

Documents

European Trade Study Group - ETSG - Effects of Trade Liberalization between … · 2013. 4. 9. · Effects of Trade Liberalization between the EU and Ukraine in a CGE Model are small