Study: Transatlantic Trade and Investment Partnership (TTIP) Who benefits from a free trade deal?...

Transatlantic Trade and Investment Partnership (TTIP)Who benefits from a free trade deal?

Part 1: Macroeconomic Effects

3

Contents

1. Introduction 4

2. Commentsonthestudymethod 5

3. TTIP-Scenarios 9

4. Wheredoestradegrow,wheredoesitshrinkandbyhowmuch? 13

4.1 EffectsonGermanforeigntrade 13

4.2 EffectsonEUtrade 16

4.3 EffectsinNorthAmerica 19

5. Howdorealpercapitaincomeschange? 21

5.1 EffectsintheEU 21

5.2 EffectsontheUSAandthirdcountries 26

6. Whathappensinthelabormarkets? 31

6.1 Searchunemploymentandforeigntrade 31

6.2 Twomillionnewjobs 34

7. Summary 42

Literature 45

Figures-andTables 47

GlobalEconomicDynamics(GED) 48

Imprint 50

Contents

Transatlantic Trade and Investment Partnership (TTIP)Who benefits from a free trade deal?

Part 1: Macroeconomic Effects

Prof. Gabriel Felbermayr, Ph.D.Benedikt HeidSybille Lehwald

4

Introduction

1. Introduction

ThisfirstpartofthestudyisdevotedtothemacroeconomiceffectsofaTransatlanticTradeand

Investment Partnership (TTIP) between the European Union and the USA. For this purpose,

empirical models of the world economy were used to clarify two essential scenarios: (i) the

eliminationoftariffsintradebetweentheEUandtheUSA,and(ii)aliberalizationthatgoesfar

beyondjusteliminatingtariffsatascalethatcanbemeasuredforcomparableandactuallyexisting

free-tradeagreements.Ourapproachdiffersfromthetraditionalwayofdealingwithsubjectinthat

thecomprehensivescenario(ii)isdefinedusingmeasurableexperiencewithsimilaragreements,

notguessesaboutwhatisconsideredpoliticallypossible.Ourresultsthusshowpotentialsagainst

whichthesuccessoftheagreementcanbemeasuredexpostfacto.

Theapproachusedinthisfirstpartofthestudyismacro-economicinnature,meaningthatthe

analysisusesaggregateddatabycountry.Itlooksatthechangesinworld-widetradeflows,real

percapita incomeandunemployment. Insodoing, it identifiescountriesthatwouldgainfrom

TTIPand those thatwould lose.Thesecondpartof the study ismicroeconomic innatureand

examinesthedisaggregatedeffectsindetailforGermany.Forexample,estimatesoftheTTIPeffect

onindividualregions,industries,educationallevelsandoccupationalgroupsarepresented.

After a discussion of the methods used and a more detailed description of the scenarios, we

examinetheTTIPeffectsonthestructureofworldtrade.Weanswerthequestion“Wheredoes

tradegrow,wheredoesitshrinkandbyhowmuch?”WethenturnourattentiontoTTIPeffectson

realpercapitaincome,i.e.,percapitaGrossDomesticProduct(GDP),whichenablesglobalwelfare

effectstobequantified.Finally,wereportontheresultsofamodelsimulationthatallowsusto

quantifytheeffectsonunemploymentrates.

5

Comments on the study method

2. Comments on the study method

ToobtainreliablenumbersfortheGDPeffectofatransatlanticfree-tradeagreementonallEU

member states essentially requires (a) an appropriate theoretical model framework and (b)

reasonablescenarios.Forthispurpose,weuseacomputablegeneralequilibriummodelthatwas

developedat the ifo Institut to analyze the free-tradeagreement.1The fundamental innovation

of the model is that it combines econometric and simulation-based methods. This enables us

to ensure that the trade costs, whose reduction are the ultimate concern of every free-trade

agreement,areproperlyestimatedandconsistentlyappliedwhencalculatingmeasurementsof

welfare.To thatend,observedreal-world trade flowsandempiricalproxies for the factors that

determinetradecostsareused.Thestrengthofthemodelliesinthefactthatitsbasicscenario

isexactlyreproduced:Thisappliesessentiallytobilateraltradeflowsbetweenalargenumberof

countries,realincomeandunemploymentrates.Inthesimulationoftheliberalizationscenarios,

specificelementsoftheestimatedcostsoftradearechangedandtheeffectsontradeflows,real

incomeandunemploymentarethencalculated.

TheapproachdiffersfromcommerciallyavailableCGE(computablegeneralequilibrium)models

frequentlyusedtoestimatetheeffectsoftradeliberalization.Infact,mostexistingTTIPstudies

alsouseCGEmethods.2Insuchstudies,tradecostsaretypicallynotestimatedconsistentlyfrom

modeltomodel.Othermodelingdifferencesconsistinthecountrybreakdownandthetreatment

of unemployment. In our approach, we consider all countries for which bilateral trade data is

available.Thereisnoregionalaggregationoritisonlyperformedafterwards,tobetterclarifythe

results.Inmanymodelvariants,wealsoassumefrictionalunemployment,whicharisesfromthe

explicitmodelingofthejobsearchprocessbyemployeesandemployers.Thismodelingapproach

wasrecognizedin2010withtheNobelPrizeandofferstheadvantagethatitreflectsthemost

importantcharacteristicsofnationallabormarketinstitutionsquitewell.

On thisbasis, the initial equilibrium for126countrieshasbeencalibrated insuchaway that

the model reproduces the bilateral trade flows observed in 2007 (before the economic crisis)

betweencountrypairs(ourdatabaseincludes15,750suchpairs)andallGDPstatisticsforthat

year.3Themodelisalsoexactly“calibrated”forthereferenceyear,anditincludesnotonlythe27

EUmembercountriesbutvirtuallyalleconomicallyrelevantcountriesintheworld.

1 ThismodelisexplainedindetailinFelbermayretal.(2013).

2 ThisappliestotheCGEstudiesonTTIPforSweden(Kommerskollegium,2013),Austria(FrancoisandPindyuk,2013),France(Fontagne and Gourdon, 2013), and for the study by the European Commission (Francois et al., 2013). These studies werefrequentlycriticized in thepastbecause theexante forecasts regarding tradeandwelfareeffects typically turnedout tobesubstantiallytoolowintheexpostevaluation.Hosny(2013)describesthedesignoftheregularCGEmodels;Ackermannofferscriticism(2006).

3 Unfortunately the modeling of frictional unemployment is only possible for those countries that have comparable data onunemploymentratesandforregulatinglabormarkets;seealsosection6.

6


Afurtheradvantageofthisstructuraleconomicmethod,i.e.,onethatassumesanexpliciteconomic

theory, is that unlike the usual econometric methods, which don’t bother to have a grounded

theoretical modeling (called reduced form estimates), it offers the possibility of analyzing the

consequencesofacontrafactualimplementation,suchasaTTIPagreement,beforeoneactually

takesplace.Thisismadepossiblebytheadditionalstructureoftheunderlyingeconomicmodel.

Reducedformestimationgenerallyallowsretroactiveanalysis,i.e.,afterimplementation,sothat

economicmeasurescanonlybedesignedafterwards,inordertobeabletoreacttospecificpartly

negative accompanying effects that have already occurred. Finally, structural modeling allows

TTIPanalysiswithouthavingtouseacentralassumptionofalmostallreducedformestimates:

that thegeneral equilibriumeffects canbe considered ignorable.4That iswhy for agreements

likeTTIP,whoserealgoalistochangethegeneralequilibrium(changethetradeflowsbetween

all participating partners, raising welfare in the participating countries), it is essential to use

structuralmethods.

Understandinghowtomodelbilateraltradecostsisveryimportantforthediscussionthatfollows

inthisstudy.Themodernliteraturemakesaveryclearandimportantpoint:Tariffsandothertrade-

policybarriers,incomparisontonaturalfrictionslikethesheergeographicaldistancebetweentwo

tradingparties,areofrelativelysmallimportance.5Andthequantitativeroleoftariffbarriersin

transatlantic trade is really negligible, compared to other trade costs.6 An exact measurement

ofexistingtradecostsishoweverofgreatimportanceincalculatingtheeffectsofliberalization

scenarios: Felbermayr et al. (2013) show that the welfare gains from trade liberalization are

disproportionatelyhigherifthetradecostsarealreadylowintheinitialequilibrium.

Figure 1 shows schematically how the trade costs between two countries can be allocated to

individualcategories.First therearethetariffswhich,asalreadymentioned,playonlyasmall

role relative to the other trade barriers. The remaining trade costs are then classified under

theheading“non-tariff”tradebarriers.Theseincludealargenumberofpotentiallyveryvaried

elements. First there are protectionist trade policy measures, which make access to domestic

markets more difficult for foreign suppliers. It includes such classic instruments as import

quotas, but also administrative and regulatoryhurdles thatdiscriminate against foreign firms.

It can also include the necessity to obtain approval for products separately for both markets,

done frequently by applying different approval conditions and procedures, different standards

toenvironmental,healthorconsumerprotectionpolicy;differentindustrystandards,packaging

regulationsandinformationrequirements;regulatingaccesstopublicprocurementoreconomic

policydevelopmentprograms,suchasgovernmentexportcreditinsurance;andsoon.

4 ThisassumptioniscalledSUTVA(StableUnitTreatmentAssumption)ineconomicliterature.Itisacomponentoftheestablishedreducedformmethods,whichareappliedtopolicyevaluationasstandards.

5 SeeBaierandBergstrand(2001).

6 Seethestudiesmentionedinnote2andFelbermayretal.(2013).

7


Besides non-tariff barriers based on trade policy, there are additional trade costs that, while

politically induced, have nothing to do with trade policy in the strict sense. For example,

instrumentsofinfrastructure,educationorlabormarketpolicyorindustrialpolicymeasurescan

beusedtochangetradecosts.Someexamplesaretheexpansionoftransportationoptionsorthe

prominenceofforeignlanguagesinschoolcurricula.

The last cost category in Figure 2 describes the so-called “natural” barriers. These are due to

geographicdistance,lack(orpresence)ofacommonlanguage,sharedcolonialpast,commonlegal

traditions,sharedcurrency,etc.

Empirically,alltradecostsbetween126countriesareestimatedinsuchawaythatatradeflow

equation (gravitation equation) is structurally derived from the simulation model that is used

later.Bydeterminingtradeelasticities,thematrixofbilateraltradecosts(126countriestimes126

countries)canbeestimated.Togetherwithmultilateraltradecostvariablesandgrossdomestic

production of the countries, this matrix replicates the expectation value of the trade actually

observedbetweenallcountries..

Source: Schematic representation by ifo Institut

Figure 1: Modeling trade costs

„Natural“barriers

Other policies

Trade policy

Non-tariffs barriers

TariffsTrade costs overall

8


Adetailedrepresentationoftheempiricalmodelwouldoverwhelmthisstudy.Insteadletussimply

mentionherethateconometricestimationofthegravitationequationhasmadegreatprogressin

recentyears, includingtheground-breakingworkofAndersonandvanWincoop(2003;2004).

Theyshowthatthetradecostswithinotherpairsareimportantformakinganaccurateestimate

oftradecostswithinacountrypair.Forexample,howmuchgeographicaldistancerestrainstrade

between twocountriesalsodependson theaveragedistanceof these twocountries fromtheir

othertradingpartners.Anadditionalproblem,forwhichtherearenowgoodsolutionmethods,

is thepresenceof countrypairs inwhich there isno trade. In theolder literature, thiswasa

problem.Eggeretal.(2011)andFelbermayretal.(2013)accordinglyusenon-linearmethodsfor

estimatingthegravitationequation.Theseeconometricproceduresalsounderlietheestimatesin

thisstudy.Finally,itisimportanttoconsiderthattradecostsarethemselvesdependentontrade

volumes.Thisappliesespeciallytothelikelihoodthatafree-tradezonewillbecreatedbetween

twospecificcountries.Toobtainaccurateestimatesanyway,instrumentvariableshavetobeused.7

Theestimatethendeliverstwoimportantresults:first,amatrixoftradecostsbetweenallcountry

pairsandsecond,theaverageeffectofafree-tradezoneonbilateraltrade.

7 Whentradecostsarethemselvesdrivenbytradevolumes,theycannotbetreatedas„exogenous“.Thishasimplicationsfortheeconometricstrategy;see,forexample,theworkofEggeretal.(2011).

9

TTIP-Scenarios

3. TTIP-Scenarios

How does a free-trade agreement work in general, and a Transatlantic Trade and Investment

Partnership(TTIP)inparticular?Suchagreementshaveadirecteffectontradecostsbetweenthe

twocountries involvedbecausetheyreducethetariffbarriersandcertainnon-tariffbarriersto

trade.Becausethescaleoftheeffectsdependsontherelatedmacro-economicvariablesonwhich

theamountoftotaltradecostsaredependent,itisimportanttoestimatetheunderlyingtradecost

matrixaccurately.Indirecteffectsthenresultinthegeneralequilibriumthroughfeedbackfrom

pricesandincome.Thecentralissueinthiscontextcanbestatedthisway:owisthetradecost

matrixchangedbyTTIP?

We consider two scenarios:

• Tariffscenario:Inthiscase,weassumethatthetradecostmatrixdescribedaboveischanged

sothatthetradecostsbetweentheUSAandEuropeancountriesarereducedbytheextent

ofthetariffsnowbeingappliedinthesebilateralrelationships.Certainexceptionsforspecial

productsmayremaininforce,butonaverage,areductiontozeroorvirtuallyzeroforallgoods

seemsreasonable.Theaveragetariffrateintransatlantictradeamountstoabout3.5%.Thatis

howmuchtradecostsbetweentheEUandUSAarereducedinthetariffscenario,whiletrade

costswithothercountrypairsremainunaffected.

• Comprehensiveliberalizationscenario:Inthiscase,weadaptthetradecostmatrixsothatthe

resultingsimulatedchangeintradeflowscorrespondstotheeconometricallymeasuredtrade

creationfromobservedfree-tradeagreements.

Thetariffscenariorequiresnoadditionaldescription.Thecomprehensiveliberalizationscenario,

ontheotherhand, issubstantiallymorecomplex.Thereductionof thestatedbarriersreleases

realresourcesthatcanbeusedforusefulactivities,whichresultdirectlyinwelfaregainsforthe

economiesaffected.Thatappliesregardlessofhowthenon-tariffbarriersareultimatelyreduced:

bymutualrecognitionofdifferentstandards,byharmonizationorbyeliminationofmeasuresthat

arepurelydiscriminatoryinnature.

Incomparisonwiththecomprehensive liberalizationscenario, theeliminationof tariffsmainly

meansthatthetransferofincomefromtheconsumertothestateisreversed.Whiletariffs,besides

havingthisdistributioneffect,alsocausealossofwelfare,thelossisverylowwhenthetariffsare

aslowastheyareinthiscase.8

8 SeeFelbermayr,JungandLarch(2013).

10

TTIP-Scenarios

Inordertodistinguishourapproachfromotherstudies,lookatFigure2.Thisshowsthroughthe

blankareasinwhichtradecategoriesthetransatlanticagreementwillleadtoareductionoftrade

costs.

First, the import tariffs appliedbetween the EU andUSAwill disappear, just like in the tariff

scenario.Howcanthechangeinnon-tariffbarriersbemodeled?Thereisaproblemwiththisin

thescientificliterature,becausethedefinitionandquantificationofnon-tariffbarrierscontinues

tobedisputed.9Oneparticularcircumstancecreatesspecialdifficulties:Evenwhenitispossible

toclearlydistinguishnon-tariffbarriersfromtariffbarriers,itremainsunclearwhichcomponents

ofthenon-tariffbarrierscaninfactbeinfluencedbyfree-tradeagreements.Inthiscontext,the

literaturespeaksof“actionability”andseeksto identifythosewhich, inthe jungleof themost

variedtradepolicymeasures,canbechangedinsomecircumstances.Thereisnosystematicand

generallyrecognizedwayofdoingthat.Inasecondstep,theremustbeawayofclarifyingthe

extenttowhichafree-tradeagreementcouldlowerthenon-tariffbarriers.Thereisnorecognized

methodofestimationforthiseither;thestudiesuseestimatesbyexperts.

Ourapproachisdifferentandavoidsallspeculation.Fromthegravitationequationthatweare

alreadyusingtoquantifythetradecostmatrix,weobtainaneconometricestimateforthetrade

creation effects of existing free-trade agreements, such as the European Union or the North

American Free-trade agreement (NAFTA). For all sectors and on average for all participating

countries,thedatashowthattheexistingagreementsincreasetradeinaggregatebyabout80%.10

Thisnumberreflectsthestatusquoof2007,butisstableovertime.Itreflectsthefactthattrade

agreementsarenotreachedbetweenrandompairsofcountriesorregions.Instead,theprobability

ofhavinganagreementishigherifthereisalreadyarelativelylargeamountoftradebetweena

countrypair.11Inaddition,itconsidersthird-countryeffects,aswellastradecreationthatresults

indirectlyfromachangeintheGDPsofallcountries.

9 SeeAndersonetal.(2008).

10 This value deviates from the one used in the industry analysis. That is due to the choice of a larger aggregation level formacroeconomicconsideration (theaggregatevalue isnot thesameas theaverageof the industryvalues)and isbasedonadifferenteconometricstrategy(solvingthe“endogeneityproblem”).

11 Itispossiblethatagreementsareonlysignediftradestimulationbetweentwocountriesisexpectedforreasonsthathavenothingtodowithtradepolicy(e.g.,economicgrowth).Inthatcase,apositivecorrelationwouldexistbetweenthetradevolumeandthepresenceofanagreement,butitshouldnotbeinterpretedasacausalrelationship.Inthiscontext,itistermeda“spurious”correlation.

11

TTIP-Scenarios

Theeconometricestimatenotonlyallowsaccurateparameterizationofabaseequilibriumfora

largenumberofcountries,italsooffersareasonablescenario.Thetradecreationof80%byalready

existingdeepagreementsjustmentionedinourviewalsorepresentsarealisticorderofmagnitude

for theaggregateeffectofa transatlantic free-tradeagreement. It canagainbeconcluded that

mereeliminationoftheremainingtariffsbetweentheEUandtheUSAwouldneversufficetobring

about tradecreationof thismagnitude. Instead, the figuresshowthat the lion’sshareof trade

creationmustcomefromloweringnon-tariffbarriers.

Thiscanbeunderstoodusingasimpleexample:Inthegravitationmodel,thereisapartialanalytic

multiplierconnectionbetweenthechangeinbilateraltradeandthechangeinallvariabletrade

costs,inwhichthemultiplieristheelasticityoftrade.Iftradethenincreasesby80%andtrade

elasticityis5,12thentradecostsmusthavefallenby80%:5=16%.Thetariffsoutsidethefree-

trade agreement amount to 3.5%. That means that the non-tariff barriers must have fallen by

16%–3.5%=12.5%.

Whilethisexampleisnaturallyhighlysimplified,itneverthelessmakesclearthatourapproach

requiresnodeterminationofthe“actionability“ofbarriersandnospeculationonthescaleofcost

reductions.Instead,weassumethatTTIPwillhaveapproximatelythesameeffectsontradecosts

thatotherfree-tradeagreementshavehad.Inthatway,thoseeffectsthatareespeciallydifficult

toquantifyarealsotakenintoaccount,reflectedundertheheading“otherpolicies.”Moreover,

our scenario definition reflects the fact that the stimulation of trade by reducing trade policy

12 AndersonandvanWincoop(2004)provideanoverviewofestimatesoftradeelasticity.

Source: Schematic representation by ifo Institut

Figure 2: Change in trade costs from TTIP

„Natural“barriers

Other policies

Trade policy

Non-tariffs barriers

TariffsTrade costs overall

12

TTIP-Scenarios

barriers leads to secondary and tertiary effects from greater public and private investment in

furtherimprovementsintradeinfrastructure.

Theaveragetrade-creatingeffectoffree-tradeagreementsinourstudyismeasuredatabout80%.

Thisissubstantiallyhigherthancalculatedinotherstudies.13Thedifferenceisduemainlytothe

differentscenariodefinition:Becausewebaseourdeepliberalizationscenarioonthemeasured

effectsinrealdata,ourresultsarehigher.Twootherdifferencesarealsoimportant,however:First,

thefocusontradecostsallowsustoreproducethestatusquooftheinitialequilibriumwithavery

terseformulationofthedemandsideofthemodel.Butbecausetradecostchangescontributein

anon-linearwaytochangesintradeflows,ourmodelismorereactivethanothersthatachieve

theresultingequilibriumbyfixingconsumptionweightsspecifictoeachcountry.14Second,our

approachmakesitpossibletoactuallyinclude126countriesintheanalysis.Thatmeansthatwe

don’thavetoperformanyregionalaggregationsthatassumeafrictionlessexchangeofgoods.For

example,tradewithintheEUisaffectedbytradecostsinourmethod,whichreflectsreality.This

alsoresultsinastrongreactionbythemodelstotransatlantictradeliberalization.

Inallscenarios,weadoptthefollowingthoughtexperiment.Weknowtheworldasitisrepresented

intheinitialequilibriuminthedata.Nowweadaptthetradeflowsinsuchawaythatthetrade

costsbetweentheEUstatesandtheUSAdecreasetotheextentthatthetradegainsoinduced

between the participating countries corresponds to the empirically measured effect of already

existingagreements.Because theseagreementshaveexistedonaverage forquiteawhile,our

approachresultsinthesimulatedeffectsofTTIPreflectingthecompleteexpressionofalldirect

andindirecteffects.

Inthesimulationmodelused,thereisarelativelysimpleconnectionbetweenthewelfaregainsof

acountryfromthenewagreement,thechangesinacountry’sopennessingeneralandso-called

tradeelasticity.Themodelfollowscurrentresearchindoingthis.15Tosummarize,itcanbesaid

thatthegreaterthetrade-creatingeffectofanagreement,themorerealpercapitaincomerises;

thegreater the tradeelasticity, themore thenewrelativecapacity forcompetingonpricewill

changetradeflowsbutthelowerthepercapitaincomegainswillbe.Theleveloftradeelasticity

actuallyreflectshoweasilythegoodsofdifferentcountriescanbesubstitutedforeachother.The

easierthesubstitution,theless“valuable”internationaltradewillbe.

13 Cf.note2.

14 Usingsuchweighting,thetraditionalmodelsrepresentthetradeflowsobserved.Inourapproach,thisisachievedbythetradecostmatrix.

15 Arkolakis et al. (2012) present a simple formula that can be used to quantify the trade gains and has minimal informationrequirements.

13

Where does trade grow, where does it shrink and by how much?

4. Where does trade grow, where does it shrink and by how much?

TTIPwillleadtoanoticeablechangeinglobaltradestructure.Weassumeinourambitiousscenario

thattradebetweentheUSAandtheEUmemberstatesincreasesonaverageexactlytotheextent

wehavemeasuredinthedataoncomparableexistingagreements.

The intertwining of countries through trade is an important indicator of cooperation, not only

in trade policy but also in many other political areas. For example, a weakening of the trade

relationshipsamongEUmemberstatescouldbeconsideredcriticalbecauseitcoulddiminishthe

interestofindividualcountriesintheEuropeanunificationproject.

4.1 Effects on German foreign trade

Table1showsthecalculatedratesofchangeofbilateral tradevolumesand itsmost important

tradingpartnersinEuropeandtheworld.Thefiguresindicatehowtradevolumeswouldappear

incomparisonwithactuallyobservedvolumesiftherehadalreadybeenanagreementbetween

the EU and USA in 2010. We are thus comparing the factual, observed reality of 2010 with a

counterfactualreality inwhichweassumethat theagreementalreadyexisted.Bydoingsowe

avoidhavingtoproposeforecastsoffuturetradevolumes(i.e.,quantitiesandprices),whichwould

beassociatedwithforecastingerrors.Andweavoidmodelingspecificphase-inpaths,sincedoing

sowouldalsobeextremelyspeculative.

ThefirstlineofTable1showsthatinthetariffscenario,GermanexportstotheUSAwouldrise

by1.13%andimportsby1.65%.ThatmeansthatGermany’stradesurplusofaboutUSD30billion

woulddeclineslightly.Lookingatthedeepliberalizationscenario instead,exportsandimports

showverysimilaradaptationsthatineachcasearemorethan90%.Wefindasaresultastrong

increaseintradeflowsbetweenGermanyandtheUSA.Thisstronggrowthshouldbeunderstood

inthecontextoftheassumedscenarios.Asexplainedabove,itincludesboththeeliminationof

tariffsandthereductionofnon-tariffbarriersthatcanbedirectlyinfluencedbypolicyaswellas

thosewhichresultfromtheeffectsofloweringpoliticalbarriersonothertrade-creatingvariables,

suchasdirectinvestments.

14


TheGermany-USAexampleshowsquiteclearlythatanagreementthatresultsonlyinlowering

tariffscannottriggeranysignificanttrade-creatingeffects.Ontheotherhand,acomprehensive

anddeepagreementhasmuchstrongereffects.

Table 1 also shows very forcefully that Germany’s trade with its other trading partners could

decline sharply in someareas. Theportents of these trade effects are consistentlynegative in

boththetariffandcomprehensiveliberalizationscenarios.Inprinciple,TTIPwouldsignificantly

alterthetradediversioneffectscurrentlyinforceintheEU.Suchalterationsoccurredinthepast

becausetradeamongtheEUcountriesthemselvesisbarrier-free,whiletradebytheEUcountries

withtheUSAissubjecttobothtariffandnon-tariffbarriers.Thelevelingofthebarrierswiththe

USAthusleadstoadeclineintradethatcameaboutasaresultofpreferentialtreatmentofintra-

Europeantradeflows.ThisismosttrueforGreatBritain,whichhasrelativelylownaturaltrade

barriers(language,culture)withtheUSA,sothat loweringpolicyfrictionleadstostrongtrade

creationanddiversion.

Interestingly, TTIP leads to an expansion of Germany’s exports to Japan, and does so in both

assumedscenarios.This is related to the fact that thecompetitivenessofGermanexporters is

increasedbytheavailabilityofcheaperintermediateproductsfromtheUSA.Moreover,thereisa

reductionofcompetitivepressureontheJapanesemarketbecauseitsexportsgobacktotheUSA.

TheeffectonGermany’stradewithChinaisnegative,however,atleastinthedeepliberalization

scenario.Adeclineinbothexportsandimportsofabout13%istobeexpectedineachcase.

Table 1: Change in German foreign trade with traditional partner countries

Exporter Importer 2010* Trade volumes (in USD millions)

Comprehensive liberalization %

Tariff scenario %

GER USA 83,553 93.54 1.13

USA GER 51,645 93.56 1.65

GER GBR 72,052 –40.91 –0.70

GBR GER 43,583 –40.93 –0.57

GER FRA 109,223 –23.34 –0.38

FRA GER 76,518 –23.34 –0.24

GER ITA 74,245 –29.45 –0.37

ITA GER 52,687 –29.45 –0.55

GER JPN 17,487 4.81 2.40

JPN GER 24,891 4.76 –1.68

GER CHN 67,728 –12.68 2.19

CHN GER 92,536 –12.71 –2.94

GER = Germany, GBR = Great Britain, FRA = France, ITA = Italy, JPN = Japan, CHN = China.

Source: ifo Institut

15


NextweexamineGermanforeigntradewiththeGIIPScountries(Greece,Ireland,Italy,Portugal

andSpain).Thisisofinterest,giventhecurrentcrisisintheEuropeanUnion.ThedatainTable

2confirmwhatwasalreadyevidentinTable1:theGIIPScountriesarenoexception–theirtrade

withGermanywilldropduetoTTIP.Theeffectsareverysmall,ifyoulookatthetariffscenario,

butaresubstantialifthemoreambitiousscenarioisassumed.

Finally,weconsiderthechangeinGermany’stradewiththeBRICScountries(Brazil,Russia,India,

China,SouthAfrica).Asshown inTable3, theeliminationof tariffsbetween theEUandUSA

causesGermany’sexportstotheBRICSgenerallytoriseslightly.Thereasonliesintheincreased

pricecompetitivenessofGermanyduetocheaperimportsofintermediateproductsfromtheUSA.

ImportsfromtheBRICS,ontheotherhand,declinebecausethecompetitivepressureinGermany

increasesfromtheintensifiedpresenceofAmericancompanies.Ifadeepliberalizationscenario

isassumed,bothexportsand importsundergo increased tradediversion:Germanexports that

previouslywenttotheBRICSnowgototheUSA;andGermanyreplacesimportsfromtheBRICS

withimportsfromtheUSA.ThisoccurseventhoughthetradebarrierswiththeBRICShavenot

changednominally.Tradediversionisdrivensolelybythechangeinrelativetradecosts.

Table 2: Changes in German foreign trade with the Peripheral EU (GIIPS)



Tariff scenario %

GER GRC 6,655 –29.94 –0.14

GRC GER 2,322 –29.93 –0.95

GER IRL 5,195 –34.87 –0.64

IRL GER 10,662 –34.85 –0.16

GER ITA 74,245 –29.45 –0.37

ITA GER 52,687 –29.45 –0.55

GER PRT 10,306 –29.90 –0.31

PRT GER 5,385 –29.88 –0.55

GER ESP 39,590 –33.71 –0.47

ESP GER 26,142 –33.71 –0.57

GER GIIPS 135,991 –30.96 –0.39

GIIPS GER 97,197 –31.22 –0.52

GER = Germany, GRC = Greece, IRL = Ireland, ITA = Italy, PRT = Portugal, ESP = Spain, GIIPS = Greece, Ireland, Italy, Portugal, Spain.


16


4.2. Effects on EU trade

Next, we look at the change in EU trade with countries in their geographic vicinity. We look

especiallyatthecountriesoftheMaghreb,withwhichtheEUhasafree-tradeagreement(Euro-

MediterraneanAgreement),andthetwosuccessorstatestotheSovietUnionforwhichwehave

data(RussiaandBelarus).

Table 4 shows that in the tariff scenario, EU exports to the countries bordering on the

Mediterranean(Morocco,Tunisia,AlgeriaandEgypt)orthesuccessorstatestotheSovietUnion

would increase, while imports from these countries would decrease. This in turn reflects the

increasedcompetitivenessofEuropeancompanies.Ifthecomprehensiveliberalizationscenario

isconsidered,however,traditionaltradediversioneffectspredominate:EUexportstoandimports

fromtheMaghrebcountriesorEasternEuropedecline.Giventhemassivedropinbarriersbetween

theEUandUSAinourscenario,theeffectsarenoteasilypredictable.

Table 3: Change in German foreign trade with the BRICS countries



Tariff scenario %

GER BRA 12,951 –7.58 2.41

BRA GER 8,844 –7.92 –3.67

GER RUS 32,512 –7.51 1.64

RUS GER 20,362 –7.17 –1.20

GER IND 10,866 –8.96 2.36

IND GER 7,859 –9.26 –2.30

GER CHN 67,728 –12.68 2.19

CHN GER 92,536 –12.71 –2.94

GER SAF 8,274 –3.96 1.24

SAF GER 6,719 –3.73 –1.20

GER BRICS 132,331 –10.06 2.03

BRICS GER 136,320 –10.93 –2.61

GER = Germany, BRA = Brazil, CHN = China, RUS = Russia, IND = India, SAF = South Africa, BRICS = Brazil, Russia, India, China, South Africa.


17


WethenshowthattradebytheGIIPSstateswiththeUSAincreasessharplyinallcases.Table5

showsthatthisisespeciallytruewithcomprehensiveliberalization.Thegrowthrates,withthe

exceptionofIrelandwheretransatlantictradeisalreadyhighlydeveloped,areonthesamescale

asforthebilateralrelationshipoftheUSAwithGermany.

Table 4: Changes in EU trade with EU neighboring countries



Tariff scenario %

EU MAR 18,028 –5.14 0.94

MAR EU 11,775 –5.39 –0.38

EU TUN 14,818 –4.75 1.33

TUN EU 13,287 –4.41 –0.82

EU DZA 21,656 –4.33 4.11

DZA EU 28,641 –0.98 –0.98

EU EGY 19,851 –5.81 1.35

EGY EU 10,584 –7.67 –0.15

EU RUS 109,586 –7.83 1.57

RUS EU 195,846 –7.75 –1.15

EU BLR 8,641 –5.91 2.59

BLR EU 7,662 –14.34 –0.42

EU = European Union, MAR = Morocco, TUN = Tunisia, DZA = Algeria, EGY = Egypt, RUS = Russia, BLR = Belarus.


Table 5: Change in trade of GIIPS countries with the USA


Comprehensive liberali-zation %

Tariff scenario %

USA GRC 1,559 90.43 1.95

GRC USA 917 90.45 0.60

USA IRL 8,022 77.03 1.44

IRL USA 28,424 77.06 1.40

USA ITA 13,254 91.77 1.71

ITA USA 28,151 91.75 1.00

USA PRT 1,068 90.56 1.78

PRT USA 2,053 90.59 1.01

USA ESP 11,575 80.18 1.62

ESP USA 8,724 80.16 0.99

GRC = Greece, IRL = Ireland, ITA = Italy, PRT = Portugal, ESP = Spain.


18


Finally, we examine the change in the trade relationships of Great Britain. This is of special

interestbecauseGreatBritain,unlikeothermembersoftheEuropeanUnion,alreadyhasavery

highdegreeofeconomicintegrationwiththeUSA,andthenewdefinitionofGreatBritain’srolein

theEUisoftendiscussedwithreferencetoitsstrongrelationshipswiththeUSA.

Table6showsthatinthecaseofatransatlanticagreement,bothexportstotheUSAandimportsby

GreatBritainfromtheUSAwouldgrowsharply.Thisisespeciallyevidentforthecomprehensive

liberalization scenario. Trade with Canada would be only slightly affected. In comparison, it

appearsthatintegrationofGreatBritainintheEUwouldbenoticeablyweakerinsomeareas.In

otherwords,throughthetransatlanticagreement,havingGreatBritainremainintheEuropean

CustomsUnionwouldbe lessvaluable forbothGreatBritainandtheotherEUmemberstates.

Giventhisbackground,thediscussionofGreatBritain’sexitfromtheEUcouldtakeonadditional

energy.

Table 6: Change in Great Britain’s trading relationships



Tariff scenario %

GBR USA 49,347 60.56 0.98

USA GBR 42,184 60.61 1.38

GBR CAN 9,400 –1.45 2.67

CAN GBR 15,929 –1.41 –0.15

GBR GER 43,583 –40.93 –0.57

GER GBR 72,052 –40.91 –0.70

GBR FRA 26,610 –36.41 –0.52

FRA GBR 34,002 –36.39 –0.52

GBR ITA 15,268 –41.47 –0.51

ITA GBR 23,191 –41.46 –0.82

GBR ESP 13,710 –45.01 –0.61

ESP GBR 15,690 –45.00 –0.84

GBR IRL 21,788 –45.97 –0.78

IRL GBR 19,420 –45.94 –0.43

GBR JPN 6,277 –13.05 2.25

JPN GBR 13,243 –13.07 –1.95

GBR CHN 9,545 –27.56 2.04

CHN GBR 48,619 –27.57 –3.21

GBR = Great Britain, CAN = Canada, GER = Germany, FRA = France, ITA = Italy, ESP = Spain, IRL = Ireland, JPN = Japan, CHN = China.


19


4.3. Effects in North America

Table7examinesthechangesintradeinNorthAmericaandbetweentheUSAandtheBRICS.A

fewimportantinsightsarestriking.First,TTIPleadstotradediversioneffectswithintheNorth

American Free Trade Area (NAFTA) between USA, Mexico and Canada. In the comprehensive

liberalizationscenario,bothexportsandimportsdeclineforNAFTApartnercountrieswithinthe

region.ThetwoNAFTAcountrieswhosepositionisnotimprovedbyTTIP,MexicoandCanada,

intensifytheirtrade.Thatisanimpressiveexampleoftradediversioneffectsbetweencountries

notdirectlyaffectedinanywaybyTTIP:TheaccessofthesecountriesespeciallytotheUSmarket

becomeslessattractivedueto increasedcompetitionfromtheEU, leadingtoasubstantialrise

intradebetweenthem.Whatmakesthiseffectsostrongisthatthetradebarriers,asweknow,

betweenMexicoandCanadahavealreadybeeneliminated.

Interestingly, TTIP leads to an expansion of trade between the EU and Canada. Geographic

circumstancesaredecisiveforthisresult.BecauseofitsclosenesstotheUSA,Canadaisespecially

affected by trade diversion effects involving the USA. This effect leads to creating trade with

theEUcountriesthataregeographicallyfartheraway,sothattransportcostsarelower,andthe

changeintherelativecoststructuresleadstoreplacementoftheAmericanmarketwiththeEU.

ThiscircumstancemeansthatfinalizationofanagreementbetweentheEUandCanada,currently

undernegotiation,wouldstrengthenthetradeofthecountriesinvolvedwitheachotherbutnot

eliminatethenegativetradediversioneffects.

20


Table 7: Change in trade between USA and Canada



Tariff scenario %

EU CAN 43,565 14.53 2.82

CAN EU 34,965 10.07 0.00

USA MEX 142,763 –15.99 –0.82

MEX USA 221,803 –16.04 1.33

USA CAN 193,554 –9.32 –0.55

CAN USA 271,268 –9.32 1.61

MEX CAN 18,965 83.53 3.10

CAN MEX 6,692 83.63 3.11

USA BRA 26,762 –29.45 –0.95

BRA USA 20,116 –29.72 –2.24

USA RUS 7,878 –29.40 –1.69

RUS USA 16,674 –29.16 0.28

USA IND 15,174 –30.51 –0.99

IND USA 29,214 –30.75 –0.84

USA CHN 83,873 –33.35 –1.16

CHN USA 327,554 –33.38 –1.50

USA SAF 5,993 –26.69 –2.07

SAF USA 8,331 –26.53 0.27

USA BRICS 139,681 –31.78 –1.17

BRICS USA 401,889 –32.69 –1.38

EU = European Union, CAN = Canada, MEX = Mexico, BRA = Brazil, RUS = Russia, IND = India, CHN = China, SAF = South Africa.


21

How do real per capita incomes change?

5. How do real per capita incomes change?

TheTTIPeffectsontradeflowspresentedthusfaressentiallyresult,ashasalreadybeenpointed

outabove,fromtheadaptationoftradecostsandtheresultingchangeinproductioncosts,input

pricesandincomepositionsofthecountries.

5.1 Effects in the EU

Whataretheeffectsofeliminatingthetransatlantictradebarriersonrealpercapitaincomeinthe

EU?DoesTTIPresultingreaterdiscrepanciesinlivingconditionsinEuropeordoesitpromote

convergence?Toanswerthisquestion,wefirstlookattheunambitioustariffscenarioandthen

describetheeffectsofdeepliberalization.

Tariff scenario

Themodelsimulationsnowshowthatatransatlanticfree-tradeagreementwouldnotcontribute

tocausinggreaterdivergenceinlivingconditionsinEurope.Infact,theoppositeismorelikely.

Figure3showsthechangeinrealpercapitaincome(whichisidenticaltorealGDPpercapita)in

allEUmembercountriesiftariffsbetweentheEUandtheUSAareentirelyeliminated.

Thechangeinrealpercapitaincomerangesbetween0.03%(Luxemburg)and0.58%(Lithuania).

For Germany, the value is 0.24%. In the non-weighted EU27 mean, the gain is 0.27% and the

standarddeviationis0.13%.Theseeffectsaresmallbecausethetariffsthatstillexistbetweenthe

EUandUSAarealreadylow(theweightedaverageis2.8%).

Theillustrationabovemakesthefollowingpointsclear:First,allmembercountriesoftheEUbenefit

fromtradeliberalization.Inthetariffscenario,thatwasnotnecessarilythecasebecausetheEUhad

togiveuptariffrevenues.Thatmeans,inthetariffscenario,lessincomeisavailablefordistribution.16

Second,howmucheconomiesbenefitdependsontherealtradestructureoftheindividualcountries,

theirsizeandtheirgeographicalposition.Thosethatexportarelativelylargeamountobtainrelatively

highergains.ThisaccountsforthesomewhathigherbenefitsinGreatBritain,forexample.Higher

incomesinthesecountriesagainresultinincreaseddemandforgoodsfromothercountriesinthe

EU.HowtheseeffectsaredistributedinEuropedependsessentiallyoninnerEuropeantransport

routesand thus thegeographical locationof thecountries.Finally, it also turnsout that smaller

countriestendtogainmorethanlargeones.TheexampleoftheBalticcountriesisespeciallytelling

inthisregard.Smallercountriesaremoreinvolvedintheinternationaldivisionoflaborandtherefore

gainmorefromlowertradecosts.Theeffectsillustratedincludetheseaspects.

16 ItisassumedthatthetariffsrevenueoftheEUisdistributeduniformlytoallEUcountries.Thatisnotactuallythecase,butalltariffrevenuesfromtradewiththeUSAaccountforlessthan0.1%ofEUGDPsothatasubstantialdistortionoftheresultisnotlikely.

22


Figure4correlatespercapitagrowthwiththe levelofrealpercapitaGDPfor2011.Thecloud

diagramshowsanegativecorrelationthatisalsoconfirmedbystatisticalanalysis.Theredline

isalinearregressionequation.Itcanbereadasfollows:thelogofFrenchpercapitaincomeis

about 10.5. The statistical model forecasts an increase in real per capita GDP for this country

of0.23%(whichishigherthanthespecificcalculationsforFranceor0.17%;i.e.,Franceprofits

belowaverage).Foracountrywitha50%lowerpercapitaincomethanFrance(suchasPoland

orHungary,forexample,wherethelogofpercapitaincomeisabout10.0),thestatisticalmodel

forecasts growth of 0.33%.17 This shows that the transatlantic agreement would contribute to

convergenceinEurope:Countriesthatcurrentlyhavealowerpercapitaincome(suchasRomania)

gainmorethanthosethathavehigherincomes(suchasLuxembourg).Thereisalargedegreeof

scatterbutthestatisticalfindingisunambiguous.

17 Thecoefficientofthelogofpercapitaincomefor2011intheregressionequationis–0.21;therelevantstandarddeviationis–0.05.Thusthecorrelationisstatisticallysignificantatthe1%level.Thisfindingdoesnotdependonthedataweighting.


Figure 3: Change in real per capita income in the EU27, tariff scenario

0.00–0.20 0.21–0.50 0.51–1.70

0.17

0.24

0.31

0.310.22

0.27

0.11

0.16

0.19

0.28

0.12

0.17

0.15

0.17

0.260.38

0.23

0.37

0.24

0.30

0.30

0.58

0.49

0.50

0.40

0.33

0,03

23


Deep liberalization

Ifwemovefromthetariffscenariotoanambitiousscenariothatbesideseliminatingtariffs,also

includesa reductionofnon-tariff barriers, the resultingpicture isdifferent.Asalsoexplained

above,itisimportanttorealizeherethatthedeepscenario,inadditiontotrade-policyliberalization

steps,alsoincludesinducedeffects(e.g.,fromtheexpansionofdirectinvestmentsorthereduction

ofeconomicpolicyuncertainty).

Source: Caculations: ifo Institut

Figure 4: Trade gains and status quo per capita income, tariff scenario

SVK

AUT

FIN

EST

LUX

MLTPRT

BEL

BGR

CYP

CZEDNK

HUN

POL

ROM GBR

SWE

NLD

ESP

SVN

DEU

GRC

LTU

IRL

ITA

LVA

FRA

9,5 10,0 10,5 11,0 11,50,0

0,1

0,2

0,3

0,4

0,5

0,6

Basisjahr 2011; Line shows linear regression.

trad

ing

prof

it pe

rcen

tage

logarithmical per capita income

24


Figure5reproducesFigure1forthecaseofadeepagreement.Onaverage,thegainsaresome

23timeshigherthaninthetariffscenario.MoststrikingisthedeviationinLuxembourg,which

profitsagreatdeal fromincreasedtradingbyitsEuropeantradingpartnerswithUSAwithout

havinglargetradeflowswithAmericaitself.Germanyalsoshowsgainsthatare20timeshigher

thaninthetariffscenario.TheEuropeanaveragegainis4.95%withastandarddeviationof1.58%.

OnecountrythatbenefitsrelativelylittleisFrance(2.64%).ThereasonisthatFrancehasrelatively

littletradeingoodswiththeUSA.Evenso,thegainsareintheclearlyvisiblerange.

Thecountriesthatprofitmostaregenerallythosethatalreadyhavehighrelativetradevolumes

with the USA. This explains the large gain by Great Britain. The Scandinavian countries and

Spainalsogetaboveaveragegains.However,inthelatterinstance,itisduetothefactthatSpain

replacesrelativelyexpensiveEuropeanimportswithimportsfromtheUSA,whichhasawelfare-

increasingeffect.


Figure 5: Change in real per capita income in the EU27, deep liberalization

0.00–3.00 3.01–6.00 6.01–10.00

2.64

4.68

6.55

9.706.93

5.03

3.63

4.43

5.28

4.92

2.71

2.58

3.31

4.21

4.434.61

4.83

5.13

6.24

3.73

7.30

5.05

5.44

5.72

6.24

5.03

0,03

25


Figure5suggeststhataboveall,thesmallerperipheralcountriesobtainaboveaveragegainsfrom

tradeliberalization.Thereasonforthisisthesameasinthetariffscenario.However,withthenon-

tariffbarrierseliminated,thestimulationofinner-Europeandemandfromsupplierrelationships

withlargecountriesthatobtainrelativelystrongbenefitsplaysasubstantiallylargerrole.

Figure6againcorrelatestheprojectedgrowthinpercapitaincomewithitscurrentlevel.Again,

the correlation is negative. The minus sign is driven entirely by Luxembourg, however. With

or without Luxembourg, no statistically significant connection between the two values can be

demonstrated.18

18 Theregressioncoefficientofthelogofpercapitaincomein2011amountsto–0.80forastandarddeviationofthesamesize.

Source: Calculations: ifo Institut.

Figure 6: Trade gains and Status Quo per capita income, deep liberalization

SVK

AUT

FIN

EST

LUX

MLT

PRT

BEL

BGRCYP

CZE

DNK

HUN

POL

ROM

GBR

SWE

NLD

ESP

SVN

DEU

GRCLTU

IRL

ITA

LVA

FRA

9,5 10,0 10,5 11,0 11,52

4

6

8

10

Basisjahr 2011; Line shows linear regression.

trad

ing

prof

it pe

rcen

tage

logarithmical per capita income

26


Conclusion

Itcanbeconcludedthatthetransatlanticfreetradeinitiativedoesnotexpandtheincomegap

within Europe. The modest tariff scenario actually shows that the agreement leads to more

convergence:i.e.,thepoorer,oftenperipheralcountriesprofitmorethanthericher,centralones.

However,insuchascenario,theaveragegainsfromtheagreementarelow.Anagreementthat

alsosignificantlylowersnon-tariffbarriersleadstomuchhigherwelfaregains.Alsointhiscase,

thereisanegativecorrelationbetweenthestatusquoincomeofEUmembersandtheirgains,so

thatadditionalconvergenceresults.However,thecorrelationhereisnotareliableoneunderthe

usualstatisticalassumptions.

5.2 Effects on the USA and third countries

Onecentralpointofcriticisminthedebateaboutanyfree-tradeagreementinvolvesitseffect

onthirdcountries. Ifa fewcountries teardownthetradebarriersamongthembutmaintain

tariffandnon-tariffbarriersagainstcountriesoutsidetheagreement,thereistradecreationthat

supportswelfareamongthepartnersbutatthesametime,tradewiththirdpartiesisdiverted.

Inthetariffscenario,itcanintheoryactuallyleadtothepartnercountriesobtainingabsolutely

nobenefit fromtheagreement:The lost tariff income is larger (inmonetary terms) than the

advantagesofimprovedmarketaccess.Andtypically,thosecountriesthatdonotparticipatein

theagreementlose.Infact,itiseventheoreticallypossibleforrealglobaltotalincometofall,

ifthegainsobtainedbyparticipantsintheagreementaresmallerthanthelossesofthosewho

remainoutside.

Theeffectsoflowernon-tariffbarriersaredifferentthanthoseoflowertariffs.Thereareseveral

reasons for this: Tariffs distribute income — essentially from consumers to producers. Their

harmful side effect is how they distort consumption and production decisions. This causes

damagetotheeconomythatrisestothesquareofthetariffs,butwhenthetariffsareverylow

(closetozero),theyarenegligible.Non-tariffbarriersdonotresultinanyincomedistribution;

instead they generate direct economic costs. To make products fit for a foreign market,

bureaucratic, regulatoryandadministrative ruleshave tobe respected,delaysoccurand the

marketriskrises.Thesecostsareassociatedwiththeuseofresourcesbutareofverylittleor

nousetotheconsumer.Inthisway,evenverysmallnon-tariffbarriersreducethepurchasing

powerofconsumersandthustheirrealincome.

27


Anadditionaldifferencebetweentariffandnon-tariffbarriersisimportantinthisconnection.

Non-tariffbarriersassumevariousforms,butoneimportantwaytoliberalizethemistounify

productstandardsorallowautomaticdomesticacceptanceofproductsthatareallowedforuse

abroad.Thatcanalsoassistthirdcountries:Ifaproductsatisfiesthestandardsofonemember

countryinafreetradezone,itmaythenbeallowedforsaleinallcountriesofthezone,evenif

itcomesfromathirdcountry.Withtheadoptionofstandards,thirdcountriescanminimizethe

tradediversioneffectsthatareharmfultothem.

Inordertoquantifytheglobaleffects,werefertothemodelusedintheprevioussection.This

modelwascalibratedandsimulatedfor126countries,sothat itcanbeusedforanalyzingthe

effects.Firstletusturntothetariffscenario.

Tariff scenario

Figure7showsthechangeinrealpercapitaincomeinallcountriesconsidered.Themodelcovers

virtuallyallthecountriesoftheworld,exceptforsomegapsmainlyinAfrica.Countriesthatprofit

fromthetransatlanticagreementareshowninblue,whilethosethatlosefromitareshownin

beige.OnecountrycoloreddarkblueistheUSA.There,realpercapitaincomerisesby0.8%just

fromloweringtariffs.Comparedtoothercountries,thetotaltradebarriersoftheUSAarerelatively

low.Thatisduetolanguage,currencyandageneralpolicyofopennesstoforeigntrade,andthe

resultisthattariffcutscanhaveastrongpositiveeffect.

ThefigureshowsthatthewinnersinthefreetradezoneareessentiallylimitedtotheUSAand

EUmemberstates.Otherthanthose,thereareonlyisolatedcountriesinwhichtheaveragereal

incomerises.Thesearecountriesthatbenefitdisproportionatelyfromadditionalexportsbecause

ofanimprovedeconomyintheEUorUSA.ExamplesareBrazil,KazakhstanandIndonesia,which

areimportantsuppliersofrawmaterialstoEuropeandtheUSA.Thesecountriesproducegoods

likenaturalgasorcottonforwhichthereareveryfewgoodsubstitutes.Interestingly,thegainsfor

KazakhstanorBrazilarehigherthantheaveragerealincomegainsinEurope.Thatshowsthatthe

complicatedinternationalinterweavingoftheflowofgoodscanalsoresultinsurprisingeffects.

CountrieslikeNorwayorJapanseenonoticeablechangesintheirpercapitaincome.

28


Themainlosersfromeliminatingtariffsarethedevelopingcountries.Theyexperiencedramatic

lossesinmarketsharefromintensifiedcompetitionontheEUorUSmarkets.Alternativemarkets

withsimilarmarketpotentialaregeographicallyfarapart.Thisisaproblemespeciallyforcountries

inNorthandWestAfrica,whichtraditionallytradeintensivelywithEurope,especiallyFranceand

Belgium.ThelistoflosersisledbyIvoryCoastandGuinea.TheirexportstoEuropearepushed

outbygoodsfromtheUSA.EastAfricacomesoutalittlebetter,mainlyduetoitsproximityto

otherlargemarketslikeChinaorAustralia/NewZealand.Buttheretoo,significantlossescanbe

experiencedbycountriessuchasUgandaandTanzania.

Overall,itshowswhatwastobefeared:IftariffsbetweentheUSAandEUfall,therelativebarriers

tomarketentryfacedbydevelopingcountriesbecomeonaveragehigher.Itisexactlythepoorer

countriesthatsuffer,someofthemtoaremarkableextent.EuropeandtheUSAwouldhaveto

moderate these negative effects through the quick signing of a “Doha light” compromise. The

resourcesfordoingsoarealreadyavailableinprinciple:ThroughtheEU-USAagreement,despite

lossesinmanythirdcountries,theworldbecomesonaveragericherbysome0.1%.


Figure 7: Change in global real per capita income, tariff scenario

–7.5 bis –4.1 –4.0 bis –1.9 –1.8 bis –0.1 0.0 bis 0.2 0.3 bis 0.5 0.6 bis 1.7

–0.7

0.8

–1.1

–1.4

–1.2–2.2 –1.0

–1.1

–1.5–0.2

–1.1–0.1

–0.9

–0.4 –2.0

0.5

–0.3

0.0

–0.2–2.1

0.30.0 0.4

0.30.2

0.40.2

0.30.2

–2.5–2.3

–3.3

0.7

1.10.3 –1.3

–0.9–1.7

–1.8 –0.9

–0.2

–0.2

–3.10.1

–2.8

–0.4 0.0

–0.6

–0.8

–1.6–2.4

–2.3

–1.1

–1.9–4.1

–4.4–7.4

–6.4 –3.0

–4.4

–4.1

–1.7

–3.3

–1.0–1.8

–2.2–4.4–2.3

–2.1

–0.2

1.7–2.2

–0.9–0.3

0.20.2

0.3 0.30.20.2

0.40.1

0.3

–1.3

0.50.5

0.0

29


Deep liberalization

Finally,letustakealookattheeffectsofdeepliberalizationbetweentheEUandUSA.Itstrade-

creatingeffectsare,asdescribedintheEU27example,severalordersofmagnitudegreaterthan

thosecreatedonlybyeliminatingtariffs.Thisnecessarilymeansthatthetradediversioneffects

arealsomoreintense.Thetrendisforcountriesthatremainoutsidetoloseevenmoresignificantly.

Allthesame,itisclearthatthestrongerrecoveryinthetransatlanticeconomycanalsomakethe

demandeffectinthirdcountriesevenmorepronounced.

Figure 8 shows the calculated effects. As in the simulations already discussed, this is an “all

otherthingsbeingequal”effect,i.e.,itincludesonlythoseeffectsthatcanbetracedbacktothe

transatlantic agreement. In the caseofnon-tariff barriers,however, it is very likely thatmany

countries,especiallythetraditionaltradingpartnersoftheEUandtheUSA,willadoptthesame

standards and regulations. Under certain circumstances, it is even realistic for countries that

alreadyhavefree-tradeagreementswiththeEUorUSAtoindirectlyparticipateinnegotiations

betweentheEUandUSA,sothattheirconcernsaretakenintoaccount.Thisdoesnotshowupin

thecalculations,sothenegativewelfareeffectsmaybeexaggerated.Theydoclearlyreflectthe

trend,however,andindicatewherethereisaneedtoact.

ItisevenmoreobviousthaninthetariffscenariothatthetraditionaltradingpartnersofEurope

andtheUSAarehurtbytheagreement.ThelossesthatwouldbeexperiencedbyCanada,Mexico,

Japan, Australia, Chile or Norway are substantial in this scenario. These countries are highly

motivatedto imitate theeliminationofnon-tariffbarriersbetweentheEUandUSAor improve

theirpartiallyexistingbilateralagreementswiththeUSAandEU,ortoenterintosuchagreements.

Therearemanysignsthatexactlysucheffortsarenowunderway.

Fortheworldingeneral,deepliberalizationbetweentheEUandUSAmeansariseinaverage

realincomeof3.27%.Thatputsenoughmoneyonthetabletocompensatethelosers.Itcanbe

hopedthattheagreementincreasesthewillingnessofdevelopingandemergingcountriestoenter

intocompromisesintheDohaDevelopmentAgenda.Atthesametime,theindustrialcountries

shouldalsobereadytomakecompromises,becauseasubstantialintensificationoftheeconomic

relationshipsbetweentheUSAandEUwouldmakethenecessaryresourcesavailable.

30



Figure 8: Change in global real per capita income, deep liberalization

–9.5 bis –6.1 –6.0 bis –3.1 –3.0 bis 0.0 0.1 bis 3.0 3.1 bis 6.0 6.1 bis 13.4

–9.5

13.4

–7.2

–4.2

–4.7–4.4 –3.4

–2.7–2.6

–2.2–1.7

–1.6

–5.6 –1.8

2.1

–3.2

–3.9–2.1

7.3–3.9 6.2

3.74.7

9.76.9

6.6

2.6

–1.7–0.6

–0.4

–0.5

–1.65.0 –2.2

–1.1–1.4

–0.8 –0.3

–0.4

–0.2

–1.3–0.2

–2.2

0.7 –5.9

–7.4

–2.0

–2.8–3.5

–2.7

–4.0

–3.0–2.8

–2.6–2.6 –2.6

–3.1

–2.6

–2.8

–1.5–2.2

–1.5

–4.0

–1.2–0.6

–4.1

–3.2

–3.3–2.5

2.62.6

4.9 4.42.64.2

4.62.7

4.8

–2.6

5.75.4

–1.3

–0.8–0.5

–0.7

5.1

31

What happens in the labor markets?

6. What happens in the labor markets?

Whiletheprevioussectionofthestudyilluminatedthetradeandwelfareeffectsofbothscenarios,

and the labor market effects were intentionally abstracted, the following section devotes its

attentiontothedecidedlyaggregatedemploymenteffectsofeliminatingtariffsaswellasofadeep

liberalizationscenario.Forthis,themethodologyofourstructuralmodelofforeigntrademustbe

expandedwithanexplicitmodelforthelabormarketsinthecountriesaffected.

6.1. Search unemployment and foreign trade

Themodelusedsofardoesnotexplicitlymodeltheeffectsonthelabormarketbecauseitwas

developed in order to analyze trade flows. In the academic literature about evaluating trade

liberalizationefforts, there isa long traditionof refraining fromexplicitmodelingof the labor

market.Onereasonmaybethelonglackofagenerallyacceptedmodelforlabormarketsandtheir

underlying institutions and frictions. However, at the latest when Christopher Pissarides, Dale

MortensenandPeterDiamondwerehonoredwiththe2010NobelPrizeinEconomicSciencesfor

theirground-breakinginsightsinmodelingsearchunemployment,alargersegmentofthepublic

finallyrealizedthatagenerallyacceptedmodelhasbeenfound.Whatdifferentiatessuchmodels

isthattheydepartfromtheassumptionappliedinclassicaleconomicsthatthewagesofworkers

adaptaslongasittakesforallthoseseekingworktohaveajobofferedbyacompany.19Obviously,

this assumption contradicts the empirical fact of the existence of involuntary unemployment.

Instead, these models assume that the unemployed first have to look for vacant positions and

only find a job with a degree of probability. Likewise, employers must assume costs in order

tofindanemployee.Theserangefromthecostsofanadinanewspaperoron-lineplatformto

theassessmentcenters,selectioninterviewsandhiringandtrainingcoststhatnewemployees

generate. A company’s unfilled positions are likewise only filled with a degree of probability.

These frictions associated with the search lead to search (or frictional) unemployment; even

duringaneconomicboom,therewillalwaysbeacertainnumberofunemployedwhocontinue

tolookforajob.Moreover,labormarketinstitutionsnaturallyaffecttheunemploymentrate.Ifa

countryhasagoodemploymentagency,moreunemployedwillfindjobsandunemploymentwill

be lower.Andwage replacementpayments such asunemployment benefits have an influence

ontheunemploymentrate.Amajoradvantageofthismodelingapproachisthatalltheselabor

marketinstitutionscanbereproducedwithinthemodelframework.Moreover,themodelexplains

thesimultaneousexistenceofjobopeningsandunemployedpeople.

19 Foranintroductiontothesemodels,seePissarides(2000).

32


Recent globalization research has taken advantage of these modeling approaches, in order to

introducethemintoexistingapproachestotheoreticalmodelingoftradeflowsbetweencountries.

AmongthepioneersinthisareawereDavidsonandMatusz(1999).Themostrecentcontributions

inthisfieldarebyHelpmanandItskhoki(2010)aswellasFelbermayretal.(2011).HeidandLarch

(2013)havetakenupthesetheoreticalstudiesandmadethemaccessibletoempiricalquantitative

analysis.20Atthecore,thisempiricalstructuralmodelconsistsofanexpansionofthemodelswe

usedtoanalyzetradeflowsforthetariffandTTIPscenariosintheprevioussectionsofthestudy

forexplicitmodelingof the labormarket.Theresults that followarebasedspecificallyonthis

methodologicalapproach.Giventheirextensivesimilarity,therewillbeonlyashortdiscussionof

theadaptedmethodologyinthepagesthatfollow.

Theempiricalstrategyisanalogoustothepreviousone:wesetthemodelusingtheobserveddata

insuchawaythatthetradeflowsobservedinthebasescenarioarereplicatedintheexpected

value.Thedifferenceisthatwenowsetthemodelexplicitlyfortheobservedunemploymentrates.

Moreover,wetakeintoaccountthefactthatindifferentcountries,therearedifferentincentives

to taking a job, which result from differences in wage replacement payments (the average

unemploymentbenefits,measuredasapercentageoftheaveragewage).21Thenweanalyzethe

effectsofthetariffanddeepliberalizationscenarios.Now,however,wecanexplicitlyinvestigate

thenetemploymenteffectsof these scenarios, andcalculate the firmnumberof jobs that are

createdorlost.Moreover,wecanalsoobservethechangeinrealwages.

From this sketch of themethodology it becomes clear thatwhile the analysis frameworknow

allowsexplicitinvestigationofthelabormarket,itis“bought”atthepriceofasubstantiallyhigher

requirement for data. We additionally need comparable information on unemployment rates,

employmentlevelsandthewagereplacementpayments.Concretely,weusethelabormarketdata

from2010;thewagereplacementratesarehoweverfrom2009,becausetheyarecompiledonly

everytwoyears.Theseareunfortunatelyavailableonlyforadatasetof28OECDcountries.22Being

limited to these industrialized countries guarantees that the labor markets are comparable, at

leasttotheextentthatotherfactors,suchasinformalemploymentorsubsistencefarming,asare

commoninLatinAmerica,AfricaandlargepartsofAsia,forexample,donotdistorttheanalysis.

Arepresentationofthelabormarketstructuresofthesecountriesisunfortunatelynotpossible,

giventhe(lackof)availabledata.

20 Theworkscitedabovearenotacompleteoverviewoftheliterature;forthat,seeHeidandLarch(2013).Thereisalsoanexactdescriptionofthemodelframeworkused.

21 Calculationofwagereplacementpaymentsismorecomplicatedthandescribedhere.WeareusingOECDdatathatcalculatesanaveragewagereplacementratefromacombinationofdifferentworkerlifesituations(familystatus,numberofchildren,etc.).Thesefiguresprovideagoodapproximationoftheaveragelevelofunemploymentbenefitsforacross-sectionofcountries;fordetailsseeOECD(2010).

22 Thecountriesare:Australia,Austria,Belgium,Canada,CzechRepublic,Denmark,Finland,France,Germany,Greece,Hungary,Iceland, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Poland, Portugal, Slovakia, South Korea, Spain, Sweden,Switzerland,Turkey,UnitedKingdomandUnitedStates.

33


Table8showstheunemploymentratesused. Inthedata for2010, theeffectsof thecontinuing

financial,governmentdebtandeurocrisisarequiteevident:Spainhadthehighestunemployment

intheOECDat20%.Germany,intheEuropeancomparison,hadarelativelylowunemployment

rateof7.46%,whiletheunemploymentrateintheUSA,atjustbelow10%,reachedahistoricalpeak.

Table 8: 2010 unemployment rates

Country 2010 unemployment rate in percent

Australia 5.23

Austria 4.39

Belgium 8.29

Canada 8.01

Czech Republic 7.28

Denmark 7.46

Finland 8.40

France 9.36

Germany 7.06

Greece 12.53

Hungary 11.16

Iceland 7.56

Ireland 13.64

Italy 8.42

Japan 5.03

Netherlands 4.45

New Zealand 6.53

Norway 3.61

Poland 9.62

Portugal 10.79

Slovakia 14.37

South Korea 3.72

Spain 20.06

Sweden 8.37

Switzerland 4.54

Turkey 11.88

United Kingdom 7.75

United States 9.63

Source: OECD Labour Force Statistics (2010)

34


Toensurethecomparabilityoftheresultswiththoseintheprevioussection,despitehavingfewer

countriescovered,weconsidered,whenparameterizingthemodel thatgrossomodo, thesame

aggregatedtradecreationeffectsbetweentheEUandUSAareachievedinbothscenariosasinthe

modelwiththe126countrieswithouttakinglabormarketfrictionsintoaccount.

6.2 Two million new jobs

Whateffectdoeseliminationofthetradebarriersintransatlantictradehaveontheunemployment

rate?Howmany jobsarecreated?Andhowdorealwageschange?Toanswer thesequestions,

weconsider first theunambitious tariff scenarioandafterwardsdescribe theeffectsofadeep

liberalization.Atthispointwewouldliketopointoutthatourmodelhasintentionallyremoved

theeconomy-inducedchangesintheunemploymentrateandonlyinvestigatesthelong-termor

accumulatedeffectsoftradeliberalizationontheunemploymentrate.Thatmeansthatthenumbers

presentedbelowaretobeconsideredlong-termresultsorequivalenttochangesinemployment

independentoftheeconomiccycle.Thatmeans,forexample,thata1percentagepointdropinthe

unemploymentratereducestheunemploymentrateduringbothanupswingandadownswingof

theeconomyby1percent.

Tariff scenario

Table9showstheresultsofeliminatingtheexistingtariffs.Thesecondcolumnshowsthepercentage

change in employment. Inversely, column 3 shows the changes in the unemployment rate.

However,thisisthechangeinpercentagepoints.Concretelythismeansthatthetariffelimination

wouldlowertheunemploymentrateinGermanyby0.11percentagepoints.Itbecomesclearthat

employmentrisesinallEUcountriesaswellastheUSA.However,thechangesareverysmalland

fortheEUstatesrangearound0.1percentagepoints.AnexceptionistheUnitedKingdom,which,

becauseofitsspecialclosenesstotheUSAbasedonlanguageandculture,asmentionedabove,

profitsinparticularwithareductionofitsunemploymentrateby0.34percentagepoints.Onthe

otherhand,inthecountriesthatdonotgainfromtariffreduction,unemploymentrisesslightly.

ThiscanbeexplainedbyrelativehighertradecostsbetweenthesecountriesandtheEUandUSA.

Theserelativelyhighertradecostsleadtoasmallertradingvolume,i.e.,lessdemandforproducts

fromthesecountries,whichreducesproductionintheaffectedcountries.Thisweakeneddemand

translatesdirectly intoa loweremploymentdemand fromcompanies in theaffectedcountries,

whichleadstoincreasedunemployment.Conversely,thenowrelativelylowertradecostsinthe

USAandEUmeangreaterdemandforgoodsfromtheEUandUSA,whichtranslatesintonew

hiresandultimatelyintoalowerunemploymentrate.

35


Whenweweighttheindividualchangeswiththegrossdomesticproduct,thereisareductionin

theunemploymentratesinthe28countriesof0.11percentagepoints.

Inatrademodelwithoutaggregatedemploymenteffects,alladaptationsoccurinpricesandwages.

Inourmodel framework,aportionof theseeffects isdiverted intoquantities, i.e.,employment

effects;howevereveninthismodelframework,theyalsoresultinchangesin(real)wages.Real

wagesinthismodelframeworkarealsoanadequatemeasurementofwelfare.Hereweseethat

inGermany,realwagesrise0.54%,intheUSA0.93%.Here,too,theUnitedKingdomprofitsmost

witharealwageincreaseof1.72%.Thecountriesnotaffectedbythetariffreductionmust,onthe

otherhand,acceptsmallerrealwagecuts.Overall,however,realwagesrise0.59%intheOECD,so

thatevenconsideringaggregatedunemploymenteffects,enoughprofitsareachievedinprinciple

tooffsetthedisadvantagedcountriesthroughtransferpayments,forexample.

Evenwithconsiderationoftheemploymenteffects,itappearsthatthetotaleffectsofonlytariff

eliminationareveryslight,andcomparedtotheeconomicfluctuationsoftheunemploymentrate,

shouldbeconsiderednegligible.

Figure10finallyofferssomeinsightintotheeffectsofthetariffreductiononconvergencewithin

theEU.Itshowsonthexaxistheunemploymentratesin2010andontheyaxisthereduction

ofunemployment rates through tariff reduction. It is immediatelyclear that inallEUmember

countries, theunemploymentratesinks, i.e.,allcountriesprofitdirectlyfromarevivalof their

employmentmarkets.Thestraightlinedrawnisaregressionline,anditshowsapositiverise.23

This should be interpreted as follows: the higher the unemployment rate in an EU member

country, thegreater thereductionof theunemployment rate fromtariff reduction.Thismeans

thattariffeliminationcontributestoaconvergenceofthelabormarketsituationwithintheEU:

Thecountrieswiththemostprecariouslabormarketsituationsprofitmost.Thegraphshowsthe

exceptionalsituationoftheUnitedKingdom,which,despiteitsrelativelylowunemploymentrate,

benefitsthemost.

23 Theincreasecoefficentisnotsignificantof0differentbasedontheUnitedKingdomoutlier.Aregressionlinewithoutthatoutlierisatthe5%significancelevelof0differentandalsoshowsaclearlypositiverise.

36


Table 9: Change in employment, unemployment rates and real wages, tariff scenario

Country Percentage rise in employment

Change in unemployment rate in percentage points

Percentage change in real wages

Australia –0.12 0.11 –0.56

Austria 0.07 –0.07 0.32

Belgium 0.02 –0.02 0.09

Canada –0.15 0.15 –0.71

Czech Republic 0.11 –0.10 0.53

Denmark 0.13 –0.12 0.63

Finland 0.21 –0.19 0.97

France 0.12 –0.11 0.54

Germany 0.12 –0.11 0.54

Greece 0.20 –0.17 0.93

Hungary 0.15 –0.13 0.70

Iceland –0.12 0.11 –0.56

Ireland 0.24 –0.21 1.14

Italy 0.16 –0.15 0.72

Japan –0.03 0.03 –0.14

Netherlands 0.09 –0.08 0.40

New Zealand –0.08 0.07 –0.37

Norway –0.12 0.12 –0.55

Poland 0.15 –0.13 0.69

Portugal 0.22 –0.19 1.02

Slovakia 0.14 –0.12 0.66

South Korea –0.03 0.03 –0.15

Spain 0.20 –0.16 0.92

Sweden 0.18 –0.16 0.85

Switzerland –0.11 0.10 –0.50

Turkey –0.11 0.10 –0.51

United Kingdom 0.37 –0.34 1.72

United States 0.20 –0.18 0.93

Average (GDP–weighted) 0.13 –0.11 0.59

Source: Calculations: ifo Institut

37


Deep liberalization

Letusnowlookatthedeepliberalizationscenario.Table10reportsthecorrespondingresults.It

isimmediatelyclearthattheeffectsforEUcountriesareaboutfourtimesaslargeasinthetariff

reductionscenario.Forexample,inGermanytheunemploymentratedropsby0.43percentage

points, which corresponds to an increase in employment of 0.47%. In the USA, the decline in

unemploymentis0.71percentagepointsandintheUnitedKingdom,even1.27percentagepoints.

Countriesaffectedbythebankandgovernmentdebtcrisisalsodoespeciallywell:Unemployment

inSpaindrops0.62percentagepoints,inGreeceandPortugalbyabout0.7percentagepointsand

inIrelandevenby0.84percentagepoints.TheeffectforIrelanddespitethecommonlanguage

withtheUSAissmallerthanintheUnitedKingdomisduetotherelativelymarginalpositionof

IrelandincomparisontotheUnitedKingdom.


Figure 9: Reduction of the unemployment rate and 2010 unemployment rate, tariff scenario

Line shows linear regression.

Austria

Belgium

United Kingdom

Sweden Spain

Netherlands

Slovakia

Greece

Portugal

Poland Hungary

Ireland

Italy

Germany FranceDenmark

Finland

Czech Republic

0 5 10 15 20 250,00

0,05

0,10

0,15

0,20

0,25

0,30

0,35

Unemployment rate in 2010

Redu

ctio

n of

the

unem

ploy

men

t rat

e pe

rcen

tage

38


Evenwithconsiderationofthequantitative(oremployment)effects,theimpactonprices,here

realwages,andthusonconsumerwelfare,issubstantiallylarger:InGermanyrealwagesare2.19%

higherandinthecountriesseverelyaffectedbythecrisisjustmentioned,atleast3%higher.It

isthusapparentthatdeepliberalization,besidesprovidinghigherpositiveemploymenteffects,

offerssubstantiallyhigherwelfareeffectsthandoes justeliminatingtariffs.However, thereare

alsolosersinthisscenariowithintheOECD:It isthosecountriesthatarenotpartoftheTTIP

agreement.However,evenintheircase,onaverage,therealGDP-weightedwageincreaseis2.34%,

sothattherearesufficientprofitsavailableforthethirdcountriesthatdon’tparticipatetobeable

tobecompensatedfortheirrealwagelosses.Onaverage,theOECDunemploymentratefallsby

0.45percentagepoints.

This is especially remarkable: The implementation of TTIP is thus not a zero-sum game but

generates real welfare gains from the elimination of real trade costs, so that (in principle) all

countriescanbenefitfromthisreduction.

Theunderlyingmechanismisthesameasinthetariffscenario:Theincreaseinexportdemand

triggered by the reduction of trade costs leads to more hiring by companies, which directly

causestheunemploymentratetofall.Atthesametime,thiscausesconsumptionofgoodstorise

domestically,basedonthehighernumberofpeopleworking,whichagaincausesmoredemand

for imports from other TTIP member countries. These positive spillover effects in the general

equilibrium with consideration for trade intertwining between countries reinforce the pure

reductionoftradecosts.

Figure11,analogoustoFigure10,providesacleardescriptionof theconvergencebetweenEU

membercountries:Thegraphslookverysimilarbutthevaluesontheyaxisaremuchhigher.It

isalsoevidentherethatthehighertheunemploymentrateinanEUmembercountrybeforeTTIP,

theloweritwillbeafterimplementationofadeepTTIPagreement.24

24 Theslopecoefficientforthisgraphislikewisenotsignificantfrom0differentbecauseoftheUnitedKingdomoutlier.Aregressionlinewithoutthisoutlierisatthe5%significancelevelof0differentandalsoshowsaclearpositiveslope.

39


Table 10: Change in employment, unemployment rate and real wages, deep liberalization

Country Percentage rise in employment

Change in unemployment rate in percentage points

Percentage change in real wages

Australia –0.47 0.44 –2.14

Austria 0.28 –0.27 1.33

Belgium 0.09 –0.08 0.42

Canada –0.60 0.56 –2.75

Czech Republic 0.46 –0.42 2.14

Denmark 0.54 –0.50 2.54

Finland 0.81 –0.75 3.84

France 0.47 –0.43 2.22

Germany 0.47 –0.43 2.19

Greece 0.78 –0.68 3.68

Hungary 0.60 –0.53 2.81

Iceland –0.46 0.42 –2.12

Ireland 0.97 –0.84 4.61

Italy 0.62 –0.57 2.90

Japan –0.11 0.11 –0.53

Netherlands 0.35 –0.34 1.65

New Zealand –0.30 0.28 –1.40

Norway –0.46 0.44 –2.12

Poland 0.58 –0.53 2.75

Portugal 0.85 –0.76 4.03

Slovakia 0.56 –0.48 2.63

South Korea –0.13 0.12 –0.58

Spain 0.78 –0.62 3.65

Sweden 0.72 –0.65 3.37

Switzerland –0.43 0.41 –1.96

Turkey –0.42 0.38 –1.94

United Kingdom 1.38 –1.27 6.60

United States 0.78 –0.71 3.68

Average (GDP–weighted) 0.50 –0.45 2.34


40


Conclusion – 2 million jobs with deep liberalization

Generally,theviewexpressedinthepreviousanalysesisconfirmed:Whileapuretariffelimination

wouldhavepositivewelfareandemploymenteffectsfortheEUandUSA,thepositiveeffectsof

deepliberalizationaremanytimesgreater.ItisalsoclearthatanyfearsaboutEUcountriesdrifting

fartherapartasaresultoftheliberalizationeffortsareunfounded;onthecontrary,theagreement

contributestowardharmonizingthelabormarketsituationsandlivingconditionswithintheEU.

Thisemphasizesthespecialsignificanceofacomprehensiveliberalizationingivingadetectible

stimulustotheeconomiesonbothsidesoftheAtlantic.

Finally,inTable11weprovideanoverviewofthenetadditionaljobscreatedinbothscenariosin

theOECD,i.e.,conversionofthechangesintheunemploymentrateintojobs.Adeepliberalization

willcreateabout181,000newjobs inGermany,andmorethanamillion in theUSA.Thetotal

amount showsagrowth in employment in allOECDcountries ofmore than2million jobs; in

the lessambitioustariffscenario,abouthalfamillion.Thesenumbersmakeitespeciallyclear

thatthedeepliberalizationgeneratesimportantemploymentstimuli.Itshouldalsobeconsidered

in this context that positive spillover effects based on psychological factors (boom mood after

Source: Calculations: ifo Institut.

Figure 10: Drop in the unemployment rate and the 2010 unemployment rate, deep liberalization

Line shows linear regression.Sweden

Austria

Belgium

United Kingdom

Spain

Netherlands

Slovakia

Greece

Portugal

Poland Hungary

Ireland

Italy

Germany FranceDenmark

Finland

Czech Republic

0 5 10 15 20 250,0

0,3

0,6

0,9

1,2

1,5

Unemployment rate in 2010

Redu

ctio

n of

the

unem

ploy

men

t rat

e pe

rcen

tage

41


comprehensiveliberalization)werenotincludedinourmodeling.Itisprobablethattheseeffects

wouldmakethefiguresevenmorepositive.

Table 11: Number of jobs created in both scenarios

Country Deep liberalization Tariff scenario

Australia –52,332 –13,591

Austria 11,638 2,828

Belgium 4,062 873

Canada –101,854 –26,176

Czech Republic 22,278 5,527

Denmark 14,623 3,646

Finland 20,066 5,134

France 121,566 29,921

Germany 181,092 44,831

Greece 34,277 8,766

Hungary 22,613 5,691

Iceland –769 –201

Ireland 18,115 4,549

Italy 140,979 35,538

Japan –71,833 –19,030

Netherlands 29,535 7,121

New Zealand –6,606 –1,748

Norway –11,541 –3,001

Poland 93,333 23,466

Portugal 42,521 10,878

Slovakia 12,995 3,259

South Korea –29,841 –7,912

Spain 143,098 36,457

Sweden 32,515 8,241

Switzerland –18,224 –4,640

Turkey –94,831 –24,625

United Kingdom 400,203 106,134

United States 1,085,501 276,623

Jobs created in the OECD 2,043,178 518,558


42

Summary

7. Summary

In this study we examined the macroeconomic effects of a transatlantic trade and investment

partnership(TTIP)betweentheEUandUSA.Twoscenarioswereexamined:(i)theeliminationof

tariffsintransatlantictrade,(ii)adeep,comprehensiveliberalizationoftradeinwhichregulatory

barriers to market access are also reduced. To do so, an empirical approach was chosen that

applies to the agreement of those trade-creating effects that can be measured in comparable,

alreadyexistingfree-tradeagreements.Bydoingso,thedirectquantificationofnon-tarifftrade

barriersandspeculationabouttheirremovalaspartoftheagreementcanbeavoided.

Themostimportantresultscanbesummarizedasfollows:

1. Trade between the USA and Germany is not significantly strengthened by eliminating

tariffs.Reductionofnon-tariffbarriersaboveandbeyondtariffsaspartofacomprehensive

liberalization scenario has much bigger effects. The growth to be expected is in the 90%

range.

2. ForotherEUcountries,tradegrowthofasimilarscalecanbeexpected.Inallcasesitwastrue

thatgrowthfromsimplyeliminatingtariffsisnegligible.

3. Germany’stradewithitstraditionaltradepartnersinEuropedeclinessharplyinsomeareas

inthecomprehensivescenario(e.g.,withFrance:–23%).Thisisduetothereversalofthetrade

diversioncausedintheEuropeanCustomsUnionanddomesticmarket.Similareffectsexist

alsofortheotherEUcountries,forexampleforGreatBritain.Thetradepolicyintertwiningof

EUcountriesamongthemselvesdeclines.

4. Germany’stradewiththeBRICScountries(Brazil,Russia,India,China,SouthAfrica)would

dropbecauseofthecomprehensiveagreementbyabout10%relativetotheinitialequilibrium.

Giventhemassiveexpansionoftransatlantictrade,thisisaslighteffect.USAtradewiththe

BRICScountrieswouldhoweverdeclinemoresharply(30%).

5. EU trade with neighboring states in North Africa or Eastern Europe would decline by an

averageof5%fromthecomprehensiveagreement.Thisresultsfromthecircumstancethatthe

TTIPpartiallydevaluesexistingpreferenceagreements.

6. Afree-tradeagreementbetweentheUSAandEUhasimportantwelfareeffectsonthecountries

directlyinvolvedandoncountriesthatareonlyindirectlyaffectedbytheagreement.Within

theEUaswelltherearedifferencescuttingacrossthecountries.WithinEurope,theBaltic

statesbenefitmostfromeliminatingtariffsintradewiththeUSA.Relativehighgainsarise

alsoinGreatBritainandinthecountriesborderingtheMediterranean.Germanycanexpectan

43

Summary

increaseinrealpercapitaincomeof0.24%.LocatedattheotherendareFrance,theBenelux

countriesandAustriawithitsneighbors.Theaverageis0.27%.

7. Reducingnon-tariffbarriershasclearlygreatereffectsonrealpercapitaincomesinEurope

than justeliminatingtariffs. ItnowcanbeseenthatGreatBritainwouldespeciallybenefit

fromtheinitiative(growthof9.70%).TheScandinavianmemberstates,theBalticcountries

andSpainseeabove-averageincreases.Germanyprofitsat4.68%abitlessthantheaverage,

whichis4.95%.France,with2.64%,gainsrelativelylittleincomparison.

8. DroppingtransatlantictariffstozerogivesthemosthelptothepoorerEUmembercountries.

Memberstateswhosepercapitaincomein2011was50%belowthatofFrancecouldexpect

welfaregainsthatareabout0.1%higherthanthoseofFrance.Inthecaseofreducingthenon-

tariffbarriers,noconvergenceeffectcanbestatisticallyproven.

9. TheUSAgainssubstantiallymorethantheEU.Inthetariffscenario,therealpercapitaincome

risesabout0.8%;inadeepliberalizationofnon-tariffbarriers,thegainsrisetoabout13.4%.

ThesehighgainsresultfromthefactofalreadylowtradebarrierswiththelargeEuropean

countrieslikeGreatBritainbutalsoGermany.

10.LiberalizationoftradebetweentheEUandUSAleadstotradecreationbetweenthepartners

buttoevidenttradediversionintradewiththirdcountries.Withpuretariffelimination,the

countries of West Africa, which traditionally trade a lot with Europe, lose up to about 7%.

However,therearealsowinnersamongthethirdcountries:Brazil,KazakhstanandIndonesia

havehigherpercentagegainsthanEuropefrompositiveindirecteffects.Welfarerisesbyabout

0.1%.

11. If thenon-tariffbarriersbetweenEUandUSAare liberalized inaddition to the tariffs, the

percapitaincomeinaglobalaveragerisesagood3.27%.Thetradediversioneffectsare,in

comparisontothecustomsscenario,onlyinsignificantlylarger,butdistributedamongthird

countriesinadifferentway.NowthetraditionaltradingpartnersoftheUSA,suchasMexico,

Canada and Chile experience substantial losses; likewise Australia, Japan and Israel each

losebetween9.5and5.5%.Thesecountriesthereforehavestrongincentivestoparticipatein

negotiatingtheliberalizationofnon-tariffbarriers.

12.Ifsearchunemploymentisconsideredinthemodelcalculations,itappearspossible,atleast

fordatareasonsonlyfortheOECDcountries,thatTTIPleadstoariseinemploymentanda

declineinunemploymentintheUSA,EUandonaverageallOECDstates.Inthetariffscenario,

the effects are small; with comprehensive liberalization, they are noticeably larger. In the

OECDaverage,theunemploymentratefallsbyabout0.5percentagepoints.

44

Summary

13. TTIPleadsinsomecountriestojoblosses.Intheambitiousscenario,theseamounttoupto

100,000jobs(inCanada).IntheOECDaverage,however,atotalof2millionadditionaljobs

arecreated.Eveninthetariffscenario,thegrowthinjobsamountstohalfamillionjobs.

14. Realwagesinthedirectlyaffectedcountriesriseonaverage;intheOECDaverage,through

adeepTTIP,theyrisebyabout2.3%,withthegrowthinGreatBritain,IrelandorUSAhigher

than,forexample,inGermany.

15. Incountrieswheretheunemploymentrateintheinitialequilibriumishigherthanaverage,

TTIP leads to an above-average decline in unemployment. This applies both to a simple

loweringoftariffsaswellastodeepliberalization.TTIPthereforealsoleadstoconvergence

onthelabormarketswithintheOECD.

45

Literature

Literature

[1] Ackerman,Frank,2006.“AssessingtheEffectsofTradeLiberalization:ACritical

Examination”[inItalian],Rivistadell’AssociazoneRossi-Doriano.3,2006.

[2] Anderson,JamesE.,andvanWincoop,Eric,2003.“GravitywithGravitas:ASolutiontothe

BorderPuzzle.”,AmericanEconomicReview,93(1),pages170–92.

[3] Anderson,JamesE.,andvanWincoop,Eric,2004.“TradeCosts.”JournalofEconomic

Literature,42(3),691–751.

[4] Anderson,JamesE.,JeffreyBergstrand,PeterEggerandJosephFrancois,2008.“Non-Tariff

BarrierStudyMethodology”,mimeo,ifoInstitut.

[5] Arkolakis,Costas,CostinotArnaudandAndresRodriguez-Clare,2012.“NewTradeModels,

SameOldGains”,AmericanEconomicReview102(1):94–130.

[6] Baier,ScottL.,andBergstrand,JeffreyH.,2001.“Thegrowthofworldtrade:tariffs,

transportcosts,andincomesimilarity.”JournalofInternationalEconomics53(1),1–27.

[7] Baier,ScottL.,andBergstrand,JeffreyH.,2009.“EstimatingtheEffectsofFree-trade

agreementsonInternationalTradeFlowsUsingMatchingEconometrics”.Journalof

InternationalEconomics,77(1),63–76.

[8] Broda,ChristianandWeinstein,David,2006.“GlobalizationandtheGainsfromVariety.”

QuarterlyJournalofEconomics,121(2),541–585.

[9] Davidson,Carl,Martin,LawrenceandMatusz,Steven(1999).“TradeandSearchGenerated

Unemployment”,JournalofInternationalEconomics,48(2),271–299

[10] Ecorys,2013.“StudyonprioritiesinthecontextofEU-UStraderelations”,Ecorys

NederlandB.V.

[11] Egger,Peter,Larch,Mario,Staub,KevinE.,andWinkelmann,Rainer,2011.“TheTrade

EffectsofEndogenousPreferentialTradeAgreements”.AmericanEconomicJournal:

EconomicPolicy3(3):113–143.

[12] Felbermayr,Gabriel,Hans-JörgSchmererandJulienPrat,2011.„Globalizationandlabor

marketoutcomes:Wagebargaining,searchfrictions,andfirmheterogeneity’’,Journalof

EconomicTheory146(1),2011,39–73.

[13] Felbermayr,Gabriel,BenjaminJungandMarioLarch,2013.“IcebergsversusTariffs:A

QuantitativePerspectiveontheGainsfromTrade”,CESifoWorkingPaperNo.4175.

[14] Felbermayr,Gabriel,MarioLarch,LisandraFlach,ErdalYalcinandSebastianBenz,2013.

„DimensionenundAuswirkungeneinesFreihandelsabkommenszwischenderEUundden

USA“,ifoStudie.

[15] Fontagné,LionelandJulienGourdon,2013.Evaluationofabilateraltradeagreement

TAFTAbetweenEUandUS.CEPII,Paris.

[16] Francois,Joseph,MiriamManchin,HannaNorberg,OlgaPindyukandPatrickTomberger,

2013.„ReducingTransatlanticBarrierstoTradeandInvestment:AnEconomic

Assessment“,StudyfortheECCommission.

46

Literature

[17] Francois,JosephandOlgaPindyuk,2013.“ModelingtheEffectsofFree-tradeagreements

betweentheEUandCanada,USAandMoldova/Georgia/ArmeniaontheAustrian

Economy:ModelSimulationsforTradePolicyAnalysis”,FIWResearchReports2012/13.

[18] Heid,BenediktandMarioLarch,2013.“InternationalTradeandUnemployment:

AQuantitativeFramework”,CESifoWorkingPaperNo.4013.

[19] Helpman,ElhananandItskhoki,Oleg,2010.“LabourMarketRigidities,Tradeand

Unemployment”,ReviewofEconomicStudies,77(3),1100–1137.

[20] Hosny,AmrSadek,2013.„SurveyofrecentliteratureonCGEtrademodelswithspecial

referencetothecaseofEgypt,JournalofWorldEconomicResearch2(1):9–19.

[21] Kommerskolegium,2013.“PotentialEffectsfromandEU-USFree-tradeagreement–

SwedeninFocus”,SwedishNationalBoardofTrade.

[22] OECD,2010.BenefitsandWages,http://www.oecd.org/els/

benefitsandwagesoecdindicators.htmAccessed:20.05.2013

[23] OECD,2010.LabourForceStatistics,Accessed20.05.2013

[24] Piermartini,RobertaandRobertTeh,2005.“DemystifyingModelingMethodsforTrade

Policy”,WTODiscussionPaper10.

[25] Pissarides,Christopher,2000.EquilibriumUnemploymentTheory,Cambridge,MA;

London:MITPress.

47

Figures and Tables

Figures and Tables

Figures

Figure1: Modelingtradecosts 7

Figure2: ChangeintradecostsfromTTIP 11

Figure3: ChangeinrealpercapitaincomeintheEU27,tariffscenario 22

Figure4: Tradegainsandstatusquopercapitaincome,tariffscenario 23

Figure5: ChangeinrealpercapitaincomeintheEU27,deepliberalization 24

Figure6: TradegainsandStatusQuopercapitaincome,deepliberalization 25

Figure7: Changeinglobalrealpercapitaincome,tariffscenario 28

Figure8: Changeinglobalrealpercapitaincome,deepliberalization 30

Figure9: Reductionoftheunemploymentrateand2010unemploymentrate,

tariffscenario 37

Figure10: Dropintheunemploymentrateandthe2010unemploymentrate,

deepliberalization 40

Tables

Table1: ChangeinGermanforeigntradewithtraditionalpartnercountries 14

Table2: ChangesinGermanforeigntradewiththePeripheralEU(GIIPS) 15

Table3: ChangeinGermanforeigntradewiththeBRICScountries 16

Table4: ChangesinEUtradewithEUneighboringcountries 17

Table5: ChangeintradeofGIIPScountrieswiththeUSA 17

Table6: ChangeinGreatBritain’stradingrelationships 18

Table7: ChangeintradebetweenUSAandCanada 20

Table8: 2010unemploymentrates 33

Table9: Changeinemployment,unemploymentratesandrealwages,tariffscenario 36

Table10: Changeinemployment,unemploymentrateandrealwages,

deepliberalization 39

Table11: Numberofjobscreatedinbothscenarios 41

48

Global Economic Dynamics (GED)

About the Authors

Prof. Gabriel J. Felbermayr, PhD, LeiterdesForschungsbereichesAußenhandel

ifoInstitut-LeibnizInstituteforEconomicResearchattheUniversityofMunich

SybilleLehwald


BenediktHeid


About the project Global Economic Dynamics (GED)

TheGlobalEconomicDynamics(GED)projectoftheBertelsmannFoundationisintendedto

contributetoabetterunderstandingofthegrowingcomplexityofeconomicdevelopments.

Byusingthemostup-to-datetoolsandmethodsformeasuring,forecastingandmodeling

globaleconomicdynamics,theprojectseekstomaketheireconomiceffectsandtheirpolitical

consequencesmoretransparentandunderstandable.

ContactBertelsmannStiftung

GED-Team

ProgrammNachhaltigWirtschaften

Carl-Bertelsmann-Straße256

D-33311Gütersloh

Phone +49524181-81353

Fax +49524181-681353

[email protected]

GED-Team

DirectorAndreasEsche

DirectorNachhaltigWirtschaften

Phone +49524181-81333

Fax +49524181-681333

[email protected]

49

Global Economic Dynamics (GED)

Projectmanager Dr.JanArpe

ProjectManager

Phone +49524181-81157

Fax +49524181-681157

[email protected]

SamuelGeorge

ProjectManager

Phone +49524181-81661

Fax +1202384-1984

[email protected]

Dr.ThießPetersen

SeniorExpert

Phone +49524181-81218

Fax +49524181-681218

[email protected]

Dr.UlrichSchoof

ProjectManager

Phone +49524181-81384

Fax +49524181-681384

[email protected]

Partners


Poschingerstraße5

81679München

ContactProf.GabrielJ.Felbermayr,PhD

Phone +498992241428|[email protected]|www.cesifo-group.de/felbermayr-g

SybilleLehwald

Phone +498992241250|[email protected]|www.cesifo-group.de/lehwald-s

BenediktHeid

Phone +498992241244|[email protected]|www.cesifo-group.de/heid-b

50

Imprint

Imprint

©2013BertelsmannStiftung

BertelsmannStiftung

Carl-Bertelsmann-Straße256

33311Gütersloh

www.bertelsmann-stiftung.de

Responsible

Dr.UlrichSchoof

Lectorate

RudolfJanGajdacz,München

Design

NicoleMeyerholz,Bielefeld

Images

PeerFrings/Fotolia.com

www.bertelsmann-stiftung.de

Address | Contact

Bertelsmann Stiftung

Carl-Bertelsmann-Straße 256

33311 Gütersloh

GED-Team

Programm Nachhaltig Wirtschaften

Phone +49 5241 81-81353

[email protected]

www.ged-project.de

NEW! GED Shorts

Our GED Shorts are a new

digital format for our research

and analysis work, which

features multimedia, interactive

presentations of our studies,

simulations and forecasts. They

are available for use and re-use

to the general public, including

media representatives.

All the content that you see

in a GED Short is available to

download and share under

Creative Commons. You can

find more information about

GED and GED Shorts at:

www.ged-shorts.org

To receive our free GED Shorts newsletter, please register here: www.ged-project.de/Register

GED Short will be available as app for android and IOS very soon!

Study: Transatlantic Trade and Investment Partnership (TTIP) Who benefits from a free trade deal?...

Government & Nonprofit

Transcript of Study: Transatlantic Trade and Investment Partnership (TTIP) Who benefits from a free trade deal?...