La lecture en ligne est gratuite
Le téléchargement nécessite un accès à la bibliothèque YouScribe
Tout savoir sur nos offres
Télécharger Lire

Federal Tax Competition [Elektronische Ressource] : How a Country's Federal Framework Affects its Position in Competing for Increasingly Mobile Capital / Lisa Tillmann. Betreuer: B. U. Wigger

De
122 pages
Federal Tax Competition- How a Country’s Federal Framework Affects its Position inCompeting for Increasingly Mobile Capital -Zur Erlangung des akademischen Grades einesDoktors der Wirtschaftswissenschaften(Dr. rer. pol.)von der Fakult¨at fur¨Wirtschaftswissenschaftendes Karlsruher Institut fur¨ Technologie (KIT)genehmigteDISSERTATIONvonDiplom-Volkswirtin (Int.) Lisa TillmannTag der mundlic¨ hen Prufung:¨ 25. Mai 2011Referent: Professor Dr. Berthold U. WiggerKorreferent: Dr. Jan Kowalski2011 KarlsruheFederal Tax Competition- How a Country’s Federal Framework Affects its Position inCompeting for Increasingly Mobile Capital -Lisa TillmannDanksagungDiese Arbeit entstand im Rahmen meiner T¨atigkeit als wissenschaftliche Mitarbeiterinam Lehrstuhl fur¨ Finanzwissenschaft und Public Management am Karlsruher Institut fur¨Technologie.Zu allererst danke ich meinem Doktorvater, Herrn Professor Dr. Berthold U. Wigger fur¨die M¨oglichkeit der Anfertigung sowie die wertvolle Betreuung der Arbeit. Herrn Professor¨Dr. JanKowalskibinichfur¨ dieUbernahmedesKorrefereatszuDankverpflichtet,ebensowieHerrn Professor Dr. Martin E. Ruckes fur¨ seine Rolle als Prufer¨ im Rahmen der Disputation.Herrn Dr. Bruno Neibecker m¨ochte ich fur¨ seine Funktion als Prufungsv¨ orsitzenderdanken.Ich danke außerdem meinen Kolleginnen und Kollegen, die mich in den Jahren amLehrstuhl begleitet und unterstutzt¨ haben.
Voir plus Voir moins

FederalTaxCompetition

-HowaCompCounetry’stingFforeIderalFncreasinglyrameworkMobileAffectsCapitalitsP-ositionin

ZurErlangungdesakademischenGradeseines
DoktorsderWirtschaftswissenschaften
l.)oprer.(Dr.vonderFakult¨atf¨ur
haftenaftswissenschWirtscdesKarlsruherInstitutf¨urTechnologie(KIT)
genehmigteDISSERTIONATnvoDiplom-Volkswirtin(Int.)LisaTillmann

Tagderm¨undlichenPr¨ufung:25.Mai2011
Referent:ProfessorDr.BertholdU.Wigger
Korreferent:ProfessorDr.JanKowalski
Karlsruhe2011

-

woH

deraleF

aCompCounetry’sting

axT

eComptition

FforeIderalFncreasinglyrameworkMobileAffectsCapitalitsP-osition

Lisa

Tillmann

in

Danksagung

DieseArbeitentstandimRahmenmeinerT¨atigkeitalswissenschaftlicheMitarbeiterin
amLehrstuhlf¨urFinanzwissenschaftundPublicManagementamKarlsruherInstitutf¨ur
hnologie.ceTZuallererstdankeichmeinemDoktorvater,HerrnProfessorDr.BertholdU.Wiggerf¨ur
dieM¨oglichkeitderAnfertigungsowiediewertvolleBetreuungderArbeit.HerrnProfessor
Dr.JanKowalskibinichf¨urdie¨UbernahmedesKorrefereatszuDankverpflichtet,ebensowie
HerrnProfessorDr.MartinE.Ruckesf¨urseineRollealsPr¨uferimRahmenderDisputation.
HerrnProfessorDr.BrunoNeibeckerm¨ochteichf¨urseineFunktionalsPr¨ufungsvorsitzender
en.dankIchdankeaußerdemmeinenKolleginnenundKollegen,diemichindenJahrenam
Lehrstuhlbegleitetundunterst¨utzthaben.
MeinwichtigsterDankgiltmeinemFreund,MarcoRiva,f¨urseinenBeistand,seinVerst¨andnis
undseineGeduld,insbesondereindenschwierigenPhasenderAnfertigungdieserArbeit.
ZuletztistesmireinbesonderesAnliegen,michbeimeinerFamilieundmeinenguten
Freundenf¨urihrenst¨andigenR¨uckhaltzubedanken.MeinenGeschwisternundFreunden
dankeichdaf¨ur,dasssiejederzeiteinoffenesOhrf¨urmeineSorgenhattenundmichauf
andereGedankengebrachthaben.MeinenElterndankeichdar¨uberhinausf¨urihrest¨andige
bedingungsloseUnterst¨utzungimRahmenmeinerAusbildung,ohnediedasVerfassendieser
Arbeitsichernichtgelungenw¨are.

Karlsruhe,imSommer2011

sttenCon

ductiontronI11.1PurposeoftheStudy...............................
1.2ANoteonFiscalFederalism...........................
1.3AimsandDesignoftheStudy..........................
ReviewiteratureL22.1TheoreticalApproaches..............................
2.1.1InterregionalTaxCompetition......................
2.1.2FederalTaxCompetition.........................
2.1.3InternationalFederalTaxCompetition.................
2.2EmpiricalApproaches..............................
2.2.1EstimatingFederalTaxCompetition..................
2.2.2InterregionalTaxCompetition......................
2.2.3FederalTaxCompetition.........................
2.2.4EfficiencyEnhancingTaxCompetition.................
.....................................Summary2.33FederalTaxCompetitionandCapitalMarketIntegration
3.1TheModel.....................................
3.1.1Households................................
3.1.2Production................................
3.1.3CapitalMobility.............................
3.1.4PolicyImplications............................
3.1.5GovernmentObjectives..........................
3.2HorizontalandVerticalExternalities......................
3.2.1AreStateTaxesTooHighorTooLow?.................
3.2.2AreFederalTaxesTooHighorTooLow?................
3.2.3NumericalExample............................
.....................................Summary3.3

i

11481001217122424262133373412424343474849494355575

TillmannLisa

4TaxCompetitionandFederalDesign
4.1TheModel.....................................
4.2TwoUnitaryCountries..............................
4.2.1CapitalInvestment............................
4.2.2GovernmentObjectives..........................
4.3UnitaryandFederalCountry..........................
4.3.1CapitalInvestment............................
4.3.2GovernmentObjectives..........................
........................Effectsross-JurisdictionalC4.3.34.3.4TotalRevenueEffects..........................
4.4InteractionofTwoFederalStates........................
4.4.1CapitalInvestment............................
4.4.2GovernmentObjectives..........................
........................Effectsross-JurisdictionalC4.4.34.4.4TotalRevenueEffects..........................
.....................................Summary4.5

Conclusion5...............................ResultsofSummary5.15.2PolicyImplications................................

R6eferences

ii

FederalTaxCompetition

59061626364646660747676787184848

878798

39

TillmannLisa

FiguresofList

1.12.1

3.13.23.3

deraleF

TaxesonCorporateIncomeintheEU-15andOECD.............
TaxRevenueStructurewithinOECD......................

CostofInternationalCapitalMobility.....................
............................Equilibriumon-ArbitrageNNon-ArbitrageEquilibrium-BenchmarkCases.................

iii

axT

etitionComp

25

446474

TillmannLisa

ListAbbreviationsof

CEEC

EU

EU-15

IMF

LHS

MCF

ML

OECD

RHS

SMBF

SMCF

TFEU

U.S.

ZMW

CentralandEasternEuropeanCountries

UnionaneEurop

FederalTaxCompetition

FifteenmemberstatescomprisingtheEuropeanUnionuntil2004

InternationalMonetaryFund

Left-handside(ofanequation)

(Perceived)marginalcostofpublicfunds

Maximum-Likelihood

OrganisationofEconomicCooperationandDevelopment

Right-handside(ofanequation)

Socialmarginalbenefitofpublicfunds

Socialmarginalcostofpublicfunds

TreatyontheFunctioningoftheEuropeanUnion

StatesUnited

Zodrow-Mieszkowski-Wilsonmodelofinterregionaltaxcompetition

iv

TillmannLisa

olsbSymofList

bBCc)k(fgGkmn,r¯rRtTuwijy)Gg,Γ(δ

FederalTaxCompetition

asebtaxState/regionalFederaltaxbase
Costofinvestingabroad
Consumptionofaprivategood
Marginalproductofcapital
Consumptionofapublicgoodprovidedatthestate/regionallevel
Consumptionofapublicgoodprovidedatthefederallevel
Amountofcapitalinvestedinonestate/region
Numberofstates/regionsinfederationA,Brespectively
Worldmarketinterestrate
Taxrevenuegeneratedinonestate/region
Taxrevenuegeneratedbythecentralgovernmentofafederation
Capitaltaxrateprevalentinonestate/region
Capitaltaxratechosenbythefederalgovernment
yUtilitWeightassignedtoregionjaccordingtosharedborderswithregioni
Productioninonestate/region
Utilityfunctionfortheconsumptionofpublicgoods
Taxrateonrentalincomeπleviedbythestate/regionalgovernments

v

Δ

θ

π

ρ

σ

τ

Lisa

Tillmann

deraleF

Taxrateonrentalincomeπleviedbythefederalgovernment

Consolidatedtaxrateonrentalincome

πRentalincomeaccruingtoanindividualinonestate/region

Netreturnoncapitalinvestments

Degreeofinternationalcapitalmobility

Consolidatedcapitaltaxrateprevalentinonestate/region

vi

axT

etitionComp

1Chapter

ductiontroIn

1.1PurposeoftheStudy

Thisdissertationintendstoenhancethediscussionofhowacountry’sfederalframework
shapesitspositioninthecontextofinternationalcompetitionforamobilecapitaltaxbase.
Precisely,itexamineshowmulti-levelledgovernmentaffectstheefficiencyofchosencapital
taxrateswithaparticularfocusontheinternationalisationofcapitalmarkets.Itderives
itsraisond’ˆetrefromvariousperspectives.Forone,afederalstructureleadstoanease
ofmobilitybetweenregions,suchthatthesearefacedwithamobiletaxbase,whichmay
notbethecaseforthecentralgovernment.Then,withaworldwidetrendtowardsfiscal
decentralisationandtheintegrationofinternationalcapitalmarkets,competitionformobile
taxbasenotonlywithinthefederation,butbetweencountriesofdifferentfederaldesign,is
aparticularlyrelevantissue.
Thepublicperceptionofanincreasinglyglobalisedeconomyappearstobeshapedby
thepotentialthreatsratherthantheopportunities.Theconnotationofinternationalcapital
investmentactivitieswiththebiblical‘PlagueoftheLocusts’prominentlyshapedbyrepre-
sentativesoftheGermanSocialDemocraticParty,1whichhasevenfounditswayintothe
annualreportoftheGermanCouncilofEconomicExperts,2seemstohavehitanervethen
andhasbeenfurtherfuelledasaconsequenceoftherecentfinancialcrisis.Theperceptionof
internationaltaxcompetition,particularlyformobilecapital,leadingtoa‘racetothebot-
tom’intaxratesculminatinginanattempttoattractmobilecapitalatthecostofcitizens
appearstobewidespreadandresilient.Yetcountriesdonotseemtohavegonebankruptin
droves,suchthattheanalysisoftheexactforcesshapingtheconsequencesofglobalisation
ishighlyrelevant.
Infact,capitalincomeorcorporatetaxratesintheOECDcountrieshavebeenreduced
12SeeSee,Gforeermanxample,CouncilNZZof(2005).EconomicExperts(2005),page35.

1

TillmannLisa

FederalTaxCompetition

overthepastdecades.3Nonetheless,taxrevenueshaveremainedrelativelystable.Figure
1.1showsthedevelopmentofrevenuesfromthetaxationofcorporateincomeforthetime
between1965and2008.TheyaredepictedasapercentageoftotaltaxationandofGDP
fortheEU-15andtheOECD.Overtheconsideredtimeperiod,adropcanbeobservedfor

Figure1.1:TaxesonCorporateIncomeintheEU-15andOECD
Source:BasedonOECD(2010)

theEU-15from2000to2005,whichmaybeindicativeofthe2004enlargementtohavehad
animpactoncorporateincometaxation.Yet,abstractingfromthedipin2008,whichis
likelytohavebeenessentiallydrivenbytheconsequencesofthefinancialcrisis,theoverall
trendshowsthattaxrevenuesrelativetoGDPhavenotbeensubjecttosignificantdeclines.
Likewise,theweightofcorporatetaxreceiptsinthecompositionoftotaltaxrevenuesalso
doesnotappeartohaverecededsignificantly.Figure1.1thusillustratesthatthefearedloss
ofcorporatetaxproceedsdoesnotseemtohavebecomemanifestonaverageintheOECD
andEU-15overthepastdecades.Inthesamedirection,Devereuxetal.(2004)examine
corporatetaxationintheUKtofindthattheobservedreductionoftaxratesdidnothave
animpactontaxrevenues.Auerbach(2007)studiesthetaxationofcorporateincomeinthe
3See,forexample,OECD(2010)orDevereuxetal.(2008)formoredetail.

2

TillmannLisa

FederalTaxCompetition

UnitedStates,wherelikewise,taxrateshavegonedown,whilerevenueshavegoneup.In
thesamedirection,BeckerandFuest(2010)findthatGermancorporatetaxrevenuesatthe
Germanstatelevel(‘L¨ander’)haveinfactrisenwithahigherdegreeofinternationalisation.
Withthesestudiesnotnecessarilysupportingtheviewofthenegativeimpactofglobalisation,
itbecomesanevenmorerelevantquestionastowhichfactorsdrivethedevelopmentoftax
ratesandtaxrevenues.Somepotentialexplanations,suchasabroadeningoftaxbaseorthe
growthoftherespectivesector,aregiveninthecitedstudies.Theexplicitconsiderationof
acountry’sfederalsetupmayyieldfurtherinsights.
TherelevanceofthistopicbecomesapparentbyconsideringthecaseoftheEuropean
Union.Itcurrentlyconsistsof27memberstates,eachofwhichhavetheirownhistorically
grownfederalframework.WithEUenlargementandintegration,capitalflowseasilywithin
Europe4andthenumberofmemberstatesisgrowing.Thathasvariousimplicationsrelevant
inthecontextofthisstudy.Forone,nexttoincreasinglyintenseworldwidecompetitionfor
mobilecapital,EUmemberstatesarefacedwithstrongtaxcompetitionbyothermember
statesduetofreecapitalmobility,whereacommonlabourmarketandincreasinglyuni-
formadjudicationacrosscountriesfurtherraisethepossibilitiesforinvestorstofreelychoose
wheretolocatetheircapital.Nexttotheperceptionofindividualmembercountries,the
EuropeanUnionmaybeseenasaplayerinitsownright.Astherecentfinancialcrisis
demonstrated,itisnotsimplyaloosecollectionofnationssharingopenborders.Onthe
contrary,monetaryandcommoditymarketsareinterdependenttosuchanextentthatpoli-
ciesofonecountryaffecttheentireunion.Theneedforconsolidatedactivitiesmightlead
totheEUbeingjustlyperceivedasasinglefederationofstates.Whilethereisasyetno
genuinesupranationalrevenuesource,itsintroductionhasbeenonthepoliticalagendasince
theearlystagesoftheEuropeanEconomicMonetaryUnion.5Themostrecentadvanceby
theEuropeanCommissioninfavourofagenuineEUtaxaspartofareformofitsfinancing
scheme6may,however,bequiterelevantgiventherecentshifttowardsahigherdegreeof
fiscalresponsibilityattheEUlevel.Tothatend,theEuropeanUnionmaywellbeperceived
asdirectlycompetingwithcountriessuchastheUnitedStatesorChina,whichthemselves
haveamoreorlessfederalstructure.7Hence,theexplicitrecognitionofthefederaldesign
ofcompetingnations,whichmayaswellshapehowtherespectivecountryisaffectedbytax
competition,8isofhighrelevancefortheanalysisoftaxsettingwithintheUnionaswellas
foritspositioncompetingwithotherregions.
4EUmembershipgrantsthefullliberalisationofcapitalflowsbetweencountries,asestablishedinarticles
FEU.T63ff.65SeeWartha(2007)foradetaileddiscussion.
SeeEuropeanCommission(2010).
7Inthatcontext,see,forexample,theempiricalresultsbyAltshulerandGoodspeed(2003)ontax
comp8etitionbetweentheUnitedStatesandtheEuropeanUnion,whichwillbediscussedinsection2.2.2.
2.1.3.sectionSee

3

TillmannLisa

FederalTaxCompetition

BeyondEurope,therehasbeenaworldwidetrendtowardsfiscaldecentralisationoverthe
pastdecades.Nexttothoseestablishedfederalcountriesthathavebeenstudiedextensivelyin
theexistingliterature,9severalemerginganddevelopingcountrieshaveintroducedelements
offiscaldecentralisation.10Thistendencymaybedrivenbysuchpromotionthroughthe
WorldBankortheIMFandmayberestrictedtotheimplementationoftransferschemes
ratherthanactuallocalrevenueautonomy,assuggestedbyBrueckner(2009).Aglobaltrend
towardsfiscaldecentralisationmaynonethelessculminateingrowingnumbersofcountries
actuallyadaptingfederaltaxsystems,suchthattherecognitionoftheconsequencesoftax
baseoverlapbecomesevenmorerelevant.

1.2ANoteonFiscalFederalism

Theconceptoffederalismissubjecttothedynamiccharacterofthepoliticalprocess.A
federationmaybedefinedasacountrycomprisinganumberofindependentstates11under
acommonlegislature,withpublicresponsibilitiesdividedbetweenthecentralandthestate
governmentsasassignedbytheconstitution.12AccordingtoGalligan(2008),afederation
ischaracterisedbythreefeatures:(i)awrittenconstitutionthatcannotbeeasilyamended;
(ii)bicamerallegislaturebythecentralandregionalgovernmentrepresentatives;(iii)judicial
review.Theevolutionoffederalstructuremaybepoliticallymotivatedorpromiseeconomic
advantagesbyprofitingfrompossiblescaleeconomiesorreduceddecision-makingcosts.It
mayariseasanagreementamongformallyindependentstates13orasaconsequenceof
thedecentralisationofaformerunitarystate,as,forinstance,inBelgiumorBrazil.Several
unitarycountries,suchasFranceorItaly,whilenotbeingfederationsperconstitution,infact
haveintroducedelementsofdecentralisation.14China,whichhasevolvedfromacentrally
9Seesection2.2formoredetail.
10AmongtheformerSovietcountries,forexample,nexttoRussia,suchreformshavebeenimplemented
inPoland(see,forexampleShah(2004)formoredetail),theCzechRepublicandSlovakia(Brysonetal.
(2004))orRomania(Sorin-DincaandDinca(2009)).SouthAfrica(seeBahl(2001)forananalysis),Kenia
(Bagaka(2008))orNigeria(Shah(2004))haveintroducedelementsoffiscalfederalismaswellasArgentina
andChile(Shah(2006)),Brazil(deMello(2008)),MexicoandColombia(Shah(2004))orPeru(Ahmadand
Garcia-Escribano(2006)).IntheAsianregion,fiscaldecentralisationwasadriverofreforminChinaand
India(seeMartinez-VazquezandRider(2005)formoredetail),Indonesia(ComolaanddeMello(2010))or
the11OrproPhilippinesvinces,(Llanregions,to(2009)cantons,andUchimdepartmenuratas,ndetc.Suzuki(2009)).
1312See,forexample,SchubertandKlein(2006).
14ThisAmongisothethers,case,Levifore(2009)xample,idiscussesnthetheUnitedrecentrStates,eforminCanada,Italy,Awhoseustralia,fiscalGsermanystemyhorasbeenSwitzerland.characterised
bycentralrevenuecollectionwithsubsequentdistributionofrevenuestotherespectiveregions.Itiscurrently
intheprocessofpassingareformtoresorttosignificantlyhigherdegreesofrevenuedecentralisationwithan
equalisationgrantschemeacrossprovinces.

4

TillmannLisa

FederalTaxCompetition

plannedtradition,isoftenreferredtoasa‘defactofederation’givenitshighdegreeof
state-levelautonomy.15Suchtendenciesareoftenattributedtoanattempttostabiliselarge
andhighly16heterogeneousorfragmentedcountriesbygrantingtheregionsastrongdegree
ofindependence.ThelattermayalsoapplytoSpain,whoseunitarystructureishighly
decentralised.Theresultingdevolutionof(fiscal)responsibilitiesisoftentheconsequenceof
extensivebargainingandinterregionalcompetitionprocessesthatmayresultinasymmetric
intergovernmentalarrangementswithinafederation.17

Figure1.2:TaxRevenueStructurewithinOECD
Source:BasedonOECD(2010)

Thisdifferentiatedviewofdecentralisedfiscalstructureisillustratedbyfigure1.2,which
depictstherevenuestructureofselectedOECDcountriesfor2008.Precisely,itliststheshare
ofgovernmentrevenuesateachjurisdictionallevelintotalgovernmentrevenues.Itincludes
eightofitsmemberstatesasfederationsandanother25unitarycountries.Itbecomes
apparentthattheattributionoftaxrevenuestolowerleveljurisdictionsdoesnothingeona
15See,forexample,Ho(2010)orZheng(2007).Foranoverviewofthespecificissuesconcerningthe
decentralisationofformercentrallyplannedcountriesrefertoRao(2008).
16See,forexample,BirdandEbel(2008).
17Forafurtherdiscussion,seeCongleton(2008).

5

TillmannLisa

FederalTaxCompetition

countrybeingperceivedasunitaryorfederal.Acrossindividualcountries,thecomposition
ofrevenuesdifferssignificantly.OECD(2010)18showsthatfortheincludedfederations,on
average,52%oftotalrevenuesareattributedtothecentralgovernment,while18.7%and
8.3%areattributedtothestate/regionalandlocallevelrespectively.19InSpain,32.4%of
totalgovernmentrevenuesapplyatthecentrallevel,while22.3%(8.9%)applyatthestate
(local)level.Ofthefurtherincluded24unitarycountries,anaverageof62.8%oftotal
governmentrevenuesareattributedtothecentral,while12.4%areattributedtothelocal
level.Theindividualfiguresforeachcountryreflecttherespectivehistoricallygrownfederal
structureandinterjurisdictionalarrangements,suchthattheymaynotbetoousefultobe
compareddirectly.Yetthepointisillustratedthatthedecentralisationofrevenuesdoesnot
necessarilyhingeonwhetheracountryhasaunitaryorafederalstructure.
Theeconomicapproachtofiscalfederalismthusincludestheanalysisofthefiscalrelations
amongjurisdictionsatdifferentlevels,20whileabstractingfromtheconstitutionaldefinition
ofafederalorunitarystate.Itcomprisesthedecentralisationofexpenditureorrevenue
authorityaswellastheexistenceoffiscalequalisationschemes.Theeconomicsliterature
onfiscalfederalismhas,overthepastyears,developedquiteintensivelyanddiversely.21
Earlyapproachesfocusontheefficiency-enhancingeffectsofadecentralisedfederalstruc-
ture.Tiebout(1956)isthefirsttopointtowardshowinterregionalcompetitionformobile
citizenswithinafederationcanraisewelfare.Inhis‘DecentralisationTheorem’Oates(1972)
prominentlyformalisestheargumentthattheuniformsupplyofpublicgoodsatthecentral
levelofacountrycanbereplacedbylocalprovisiontailoredtotheactualpreferencesofthe
citizens,therebyincreasingoverallwelfare.22Variousstrandsofliteraturerootedintheseper-
ceptionssubsequentlyevolved.Theliteratureonenvironmentalfederalismthatemphasises
thebenefits(andrisks)ofdecentraliseddecision-makingwithrespecttoecologicalstandards
buildsonOates’DecentralisationTheorem.23TheideasformulatedbyTiebout(1956)have
substantiallyshapedthemostinfluentialcontributionsoninterregionaltaxcompetition,for
example,byZodrowandMieszkowski(1986)orWilson(1986).Furtherapproaches,which
havetheirrootsinthePigouvianperspectiveontheinternalisationofspillovereffectsacross
18SeeOECD(2010)tableE,page28.
19Percentagepointsmissingto100areassignedtosocialsecurityfundsorsupranational-levelrevenues(as
in20thecaseoftheEUmemberstates).
SeeAhmadandBrosio(2008).
21See,forexample,Oates(2005,2008),forrelativelyrecentsurveysonthedifferentstrandsofresearchin
ederalism.ffiscal22Oates(1972)abstractsfrompossiblescaleeconomiesandrestrictstheviewtowelfare-maximisinggov-
ts.ernmen23See,forexample,Dinan,CropperandPortney(1999),Oates(2002)or,morerecentlyBanzhafandChupp
(2010).

6

TillmannLisa

FederalTaxCompetition

regions,24argueinfavouroftheexistenceofintergovernmentalgrantorfiscalequalisation
schemes.25ExtendingNiskanen’s(1971)viewofpolicymakersasLeviathansratherthan
welfaremaximisers,BrennanandBuchanan(1980)makethepointofthewelfare-increasing
potentialofdecentralisationbyconstrainingtheincumbents’revenue-maximisingambitions.
Otherapproaches,whichOates(2008)referstoasthe‘secondgenerationoffiscalfed-
eralism’incorporateelementsofpublicchoice,contracttheoryorinformationasymmetries
intothefiscalfederalismliteratureandexplicitlytakeintoaccountthetypicalincentive
anddecision-makingstructuresarisingfromdecentralisationaswellasincludingitspossi-
bledistortiveeffects.Suchapproachesalsoanalysethedevolutionoffiscalresponsibilities
withinacountry.26Onestrandofliteraturetakesapoliticaleconomyapproachtofiscal
federalismandexplicitlymodelsthemotivationbehindapolitician’sactionsaswellasthe
structureofdecision-makingandelectoralprocessesthatshapefiscaloutcomesunderde-
centralisedasopposedtounitarystates.27BasedonearlycontributionsbyKornai(1979,
1986),someauthorsexplicitlyconsidertheformationoffiscalbudgetrulesandhowthey
influencegovernmentbehaviour.28Theyardstickcompetitionliterature,asshapedbyBesley
andCase(1995)29furtherconjecturesthatfederalstructureleadstoabettercomparability
betweenregions,suchthatpolicymakersmayhaveincentivestoengageintaxmimicking
behaviour,whileatthesametimevotershaveameanstoevaluatetheperformanceoftheir
incumbents,therebyrestrainingpossiblerent-seekingbehaviourofgovernments.Suchin-
formationalspilloverswithinafederationarealsoincorporatedintotheso-calledlaboratory
federalismapproaches,30inwhichthedirectcomparisonofregionsunderafederalstructure
providesaframeworkthatfavoursinnovationinsuchthatfiscalreformsimplementedby
particularlyinnovativeregionsmaylaterbeadaptedacrossthefederation.
Notwithstandingthevastarrayofresearchonfiscalfederalism,thisdissertationfocuses
ontherevenuesideinthecontextoffederaldecisionmakingwithaparticularfocusoncom-
petitionforamobilecapitaltaxbase.31Inwhatfollows,thedistinctionofafederationin
24(1947).PigouSee,2526SeeSee,Bforoadweayxample,(2006)Cforongletonasurvey(2008)onortheEichenliteraturebergeronandeFreyqualisation(2008)granfortaschemes.detaileddiscussion.
27Someofthewell-knowncontributionsinthatrespectincludeLockwood(2002,2006),BesleyandCoate
(2003)aswellasPerssonandTabellini(1992,2002).Gurievetal.(2010)providearecentcontributionin
that28AmcontextbrosaniodandiscussingBhoordignonwinterest(2008)groupsdevelopshapiefnstrumenederaltspofortlicies.heassignmentofrevenueswithinafederation.
TheconsequencesofsoftorhardbudgetconstraintswithinafederationhavebeenanalysedbyWildasin
(1997),QianandRoland(1998)orGoodspeed(2002a)andlaterBesfamilleandLockwood(2008)orFink
(2009).tratmannSand3029See,Moreforerecentxample,conRtributionsose-Ackeincluderman(1980),KotsogiannisStrumpfandSc(2002),hwKager(otsogiannis2008)orandRincSckehw(2006,ager(2009).2006a,2006b)or
KerberandEckardt(2007).
31Thereare,ofcourse,variousapproacheswithafocusonothertypesoftaxesintheexistingliteratureon

7

TillmannLisa

FederalTaxCompetition

contrasttoaunitarycountrywillthusrefertoacountrywithadecentralisedcapitaltaxrev-
enuestructureirrespectiveofitsconstitutionalspecifications.Theexistingliterature,which
willbediscussedindetailinchapter2,analysestaxationinafederationwithseveralstates
andonefederalgovernment,whichisgenerallycharacterisedbytwoelements:competition
formobiletaxbasesbetweenlowerleveljurisdictionsandconcurrenttaxationofthesame
taxbasesbylowerleveljurisdictionsandthecentralgovernment.Taxcompetitiongives
risetohorizontalexternalities,aseachlowerleveljurisdictionignoresthepositiveeffectofa
higherowntaxrateontaxrevenueoftheotherlowerleveljurisdictions.Taxationwillthen
betoolowinequilibrium.32Concurrenttaxation,ontheotherhand,givesrisetovertical
taxexternalities,aseachlowerleveljurisdictionignoresthenegativeeffectofahigherown
taxrateonthecentralgovernment’staxrevenue,andviceversa.Thisleavestaxratestoo
highinequilibrium.33Giventhatthetwoelementsoffederaltaxationhaveoppositewelfare
effects,thequestionwhichmaybedominantinagivensettingisofhighrelevanceandwill
bediscussedwithreferencetotheinternationalintegrationofcapitalmarketsinthecourse
thesis.thisof

1.3AimsandDesignoftheStudy

Withtheissuesoutlinedsofar,theintentionofthiscontributionistoenhancethediscussion
regardingthefollowingquestions:

1.Whatisthecurrentstateofresearchontheconsequencesofinternationaltaxcompe-
titionwithregardtothedecentralisedrevenuestructureofacountry?

2.Howdoestheincreasingintegrationofinternationalcapitalmarketsaffectthetaxation
capital?of3.Howdoesacountry’sfederalframeworkaffectitspositionwhencompetingwithother
countriesforaninternationallymobilecapitaltaxbase?

Hence,thisdissertationissetupthefollowingway:Thesecondchapterreviewstheexist-
ingempiricalandtheoreticalliteraturerelevantinthecontextofthisstudy.Thetheoretical
focuswillforonepartlieoncontributionsconcerningthetaxationofmobilecapitalwith
federaltaxsetting.Assuch,concurrenttaxationofpersonalincomehasbeenthesubjectofanalysis,among
others,inMintzandTulkens(1986),Boadwayetal.(1998)or,morerecently,Klor(2006)andAronsson
(2010).Excisetaxationofconsumptiongoodsisalsowidelyexamined,as,forexample,inBesleyandRosen
(1998),Devereuxetal.(2007)orRizzo(2010).
32Amongothers,refertoWildasin(1989),ordePaterandMyers(1994).
33See,forinstance,WiggerandWartha(2004),orMakris(2006).

8

TillmannLisa

FederalTaxCompetition

interregionalcompetition,therebycharacterisingtherespectivehorizontaleffects.Subse-
quently,modelsanalysingtaxationinafederationwithanactivefederalandmultiplelower
levelgovernmentsareconsidered.Theseidentifytaxbaseoverlapasasourceofverticalex-
ternalities,whoseinteractionwithhorizontalexternalitiestriggeredbylowerleveltaxcompe-
titionculminatesinthequestionwhichofthesewillbedominantinagivensetting.Finally,
thesmallliteraturecombiningthecontributionsoninternationalhorizontaltaxcompetition
withtheinsightsfromthefederaltaxationliteratureisreviewed.Thesurveyoftheempirical
literaturestartsbydiscussingstudiesaimingtofindevidenceofhorizontaltaxinteraction
betweengovernments.Itincludessuchinterrelationsamongregionalgovernmentswithin
onecountryaswellasinternationalhorizontaltaxinteractions,andfurtherdiscusseshow
theseareaffectedbyincreasedcapitalmarketintegration.Subsequently,thestudiesseeking
evidenceofverticaltaxinteractionwithinafederationarereviewed,followedbyadiscussion
ofcontributionsaimingtoclarifywhethergovernmentsaremoreappropriatelyassumedto
haverevenueorwelfaremaximisingintentions,adistinctionthatisfoundtoberelevantin
thereviewofthetheoreticalliterature.Chapter2concludesbysummarisingtheresultsand
pointingtowardsfurthertopicsofresearch.Itderivesasetofkeyquestionsthatappear
toberelevantfortheanalysisofpolicyimplicationsbyfederaltaxcompetition.Thethus
identifiedkeyaspectsaresoughttobefurtheranalysedinchapters3and4.
Chapter3considersafederationwhosedegreeofintegrationintointernationalcapital
marketsisexplicitlymodelledandanalysestheefficiencyoftheresultingtaxrates.Itis
foundthatstrongintegrationintointernationalcapitalmarketsfunctionsbylimitingthe
extenttowhichhorizontalexternalitiesmaydetertaxsettingwithinafederation,suchthat
theverticalexternalityhasarelativelystrongerimpactontheoveralltaxburden.
Thescopeisextendedtoanalysetwocountriescompetingformobilecapitalwhileallowing
fordifferencesinfederaldesignbyeachofthecountriesinchapter4.Itconsiderstwounitary
countriesasareferencecaseforpurelyhorizontalinteraction.Themodelisthenexpanded
toincludetheinteractionofafederalandaunitarycountry,inordertofinallyexamine
twocompetingfederations.Therespectiveexternalitiesarisingineachofthethreecases
areevaluatedandpolicyimplicationsarederived.Amongotherthings,itturnsoutthat
federalstructuremayhaveapositiveimpactregardingacountry’spositionininternational
tition.eompctaxFinally,chapter5concludesbysummarisingtheresultsandpointingouttheirimplica-
tionswithregardtothekeyquestionsderivedinchapter2.Itfurtherdepictsthepolicy
implicationstobedrawnfromthediscussionoffederaltaxcompetitionaswellasfurther
topicsofresearch.

9

2Chapter

ReviewLiterature

“Competitionbetweengovernmentunitsatthesameoratadifferentlevelex-
plainsfundamentalcharacteristicsoftheworkingofdecentralizedsystems,andis
potentiallyacrucialfactorfortheirefficiency.”

ThisquotefromtheintroductionoftheHandbookofFiscalFederalism1highlightstherele-
vanceofstrategicinteractionbetweenjurisdictionalentitiesanditsinterdependencewitha
country’sfederalorganisation.Giventheabundanceofliteratureonfederaltaxcompetition,
thischapterintendstogiveastructuredoverviewoftheexistingbodyofwork.Itissetupas
follows:Thenextsectionreviewstheexistingtheoreticalliteratureoncapitaltaxationina
federationandtheimplicationsofinterregionaltaxcompetitionwithinsuchasetting.The
existingempiricalliteratureconcerningfederaltaxcompetitionissurveyedinsection2.2,
withaparticularfocusoninterregionaltaxcompetitionanditsimplicationswithrespect
tofederalstructure.Section2.3concludesandderivessomekeyquestionsrelevanttothe
evaluationofpolicymakinginthecontextofinternationalfederaltaxcompetition.

hApproacTheoretical2.1es

Thissectiongivesanoverviewoftheexistingtheoreticalapproachestofederalcapitaltax
competition.Itcanbedividedintothefollowingmajorstrands:Forone,afederationis
seenasabundleofregionsunderthecommonlegislationofonecountry,wherethefederal
leveldoesnotactivelyparticipateindecisions.Hence,taxcompetitioninafederationis
characterisedbyhorizontaleffectsoccuringbetweenjurisdictionsatthesamelevel.2In
1SeeAhmadandBrosio(2008),p.22.
2Inthatrespect,itisequivalenttointernationaltaxcompetitionbetweencountries.Surveysofthe
literatureoninterregionaltaxcompetitionexist,forinstance,byWilson(1999)orWildasinandWilson
(2004).

10

TillmannLisa

FederalTaxCompetition

anotherstrand,thefederalgovernmentisexplicitlyrecognisedasanactiveplayer,which
leadstoverticaltaxcompetitioneffectsbeingtriggeredbyregionalandfederaldecision-
makersinteractingwithinonecountry.Athird,morerecentfieldanalysestaxregimeswhen
federationsareexplicitlymodelledincompetitionwithothercountries.Thissectionwillgive
anoverviewoftheexistingliteratureineachofthesefields.
Thereferencecasetoassesstheefficiencyoftaxsettingisthesolutionreachedbyan
omniscientcentralplannerunderdistortivetaxationtailoredtoeachregion.Theanalysisis
restrictedtoapproachesinwhichthestrategyvariableisthechosentaxrate,despitebeing
perfectlyawarethatothervariables,suchasthelevelofexpenditures,theprovisionofapublic
consumptiongoodorpubliclyprovidedinputsforproductionhavealsobeenconsidered.The
respectiveoutcomesessentiallyhingeonthegovernment’sobjectivefunction.Typically,the
literaturedistinguishesbetweentheperceptionofrevenue-maximisingLeviathangovernments
orbenevolentwelfare-maximisingdecision-makers.3
Inthereferencecase,acentralLeviathangovernmentwillchoosetaxrates,suchthat
thecountryreachesthemaximumoftheLaffercurve4.Bycontrast,inafederationof
Leviathans,theobjectivefunctionsofgovernmentsdifferinsuchthateachpolicymakeronly
caresaboutownrevenues,whichtriggerstherespectiveexternalitiestobecharacterisedin
theproceedingsections.Inefficienttaxationthenimpliesthatrevenuesarenotmaximisedby
choiceofthetaxrate.Withinefficientlylowtaxratesacountryfindsitselfontheupward-
slopingsideoftheLaffer-curve,whilethereverseholdsforinefficientlyhightaxrates.5Since
revenue-maximisationdoesnotimplywelfare-maximisation,6bylimitingtheLeviathaninits
possibilityforwastefulconsumption,welfaremayinfactberaisedthroughlowertaxrates.
Abenevolentgovernmenttypicallyaimstomaximiseitscitizens’utility
u(c,g,G),wherecisprivategoodconsumptionandg(G)istheamountofapublicgood
providedattheregional(federal)level.7Theomniscientcentralplannerwillachieveasecond-
bestoptimumwithdistortivetaxeschoosingtaxrates,suchthatthesocialmarginalcostof
raisingpublicfundsequalstheirsocialmarginalbenefit,thatis,SMCF=SMBF.8That
equalitymaybedistortedbydecentralisationinsuchthatonejurisdiction’sgovernmentmis-
perceivesthemarginalcostofpublicfundsandthusover-orunderprovidestherespective
publicgood.Inefficientlylowtaxratesthenimplythatwelfarecouldberaisedbyanincrease
3Inthatcontext,see,forexample,BuchananandMusgrave(2001).
4SeeLaffer(1979),whichwasearlierreferredtobyWannisky(1978).TheLeviathanapproachisassigned
toBrennanandBuchanan(1977,1978,1980).
5Inthelattercase,revenuescouldberaisedbyadropintaxrates,whichisalsoreferredtoastheLaffer
paradoxon,as,forinstance,inDahlby(1996).
6See,forexample,theseminalcontributionsbyBrennanandBuchanan(1977,1978,1980).
7Utilitymay,ofcourse,bevariedorextendedtoincludeotherelements.Forinstance,seeWrede(2002),
62.p.8Forinstance,seeKeen(1998).

11

TillmannLisa

FederalTaxCompetition

intaxratesandviceversa.Inefficienciesmaystemfromdifferingobjectivefunctionsorfrom
theassumptionofmyopicgovernments.

2.1.1InterregionalTaxCompetition
Afederationconsistingofnautonomousregionsisfirstconsidered,wheredecision-makingis
perfectlydecentralised.Whilethecapitalstockmaybefixedwithinthefederation,capital
istypicallyassumedtobemobileacrossregions.Thatis,eachjurisdictioniperceivesits
capitaltaxbaseaselastic.Uponchoosingitstaxrate,eachregionalgovernmentwilltrigger
apositivehorizontalexternality,insuchthatinvestorswillchoosetorelocatetheircapital,
therebyraisingtheotherregions’taxbasesbj=i,9suchthat
1−nbjti>0.
=1i=jGiventhateachgovernmentonlyaccountsforitsowntaxbaseandneglectstheexpansion
oftheotherregions’taxbase,theresultinghorizontalexternalitypointstowardsinefficiently
lowtaxation.Whetherornotinterregionaltaxcompetitionenhancesordeterswelfare,is
subjecttotheperspectivetaken.Therefore,thoseapproachesassumingforgovernmentsto
bebenevolentwillbereviewedfirst,followedbytheresultsgeneratedforLeviathans.

BenevolentGovernments
Withbenevolentdecision-makers,eachregionalgovernmentchoosestomaximiseitscitizens’
utilitybyprovidingoptimalopportunitiesfortheconsumptionofpublicandprivategoods,
wherepublicgoodprovisionisfinancedfromcapitaltaxation(and,possibly,thetaxation
ofafixedfactor,whichis,however,mostlyassumedtobeexogenous).10Itwillrecognise
hownotonlyitsowndecisions,butalsothosebyeveryotherinstitutionfeedbackintoits
inhabitants’welfarefunction.Itneglects,however,howitsownchoicesaffecttheutilityof
otherregions’inhabitants,therebytriggeringahorizontalexternalitythatpointstowards
anunderprovisionofpublicgoods.Tiebout(1956)establishedtheviewofinterregional
taxcompetitionraisingefficiencyviaso-calledTiebout-sortingprocesses.Theunderlying
reasoningisthatbychoosingtheirpreferredlocation,perfectlymobilehouseholdswillinduce
governmentstointernalisetheseexternalitiesandtoprovideefficientlevelsofpublicgoods
andtaxation.11OneofthecentralassumptionsintheTieboutworldisthatofpublicgood
provisionbeingfinancedbybenefittaxation.Thatassumptionisdroppedinanotherstrand
910FoSee,riforenstance,xample,seeWilsonWilson(1999).(2005).
11Fischel(1976)andWhite(1976)firstintegratedmobileprofit-maximisingenterprisesintoaTiebout-type
setting.MorerecentapproachescanbefoundinRichterandWellisch(1996)orWellisch(2000).

12

TillmannLisa

FederalTaxCompetition

oftaxcompetitionliterature,whichwasmainlyshapedbyZodrowandMieszkowski(1986)or
Wilson(1986),hereafterreferredtoasZMW.12Typically,thesemodelsconsiderbenevolent
regionalgovernmentswhichderivetheirentirebudgetfromthetaxationofcapital13engaging
inNashcompetitionovertaxrateswiththeotherregions.Thefederationconsistsofn
identicalregions,eachinhabitedbyoneindividualendowedwithoneunitofanimmobile
factortobeusedasaninputforproduction(labourintheZMW-approach)andoneunitof
interregionallymobilecapital.Investorswillreacttowardsgiventaxratesbylocatingcapital
whereitsnetreturnishighest,suchthatnon-arbitragewillleadtothelatterbeingequalised
acrossstates.Intheresultingequilibrium,chosentaxratesandpublicgoodprovisioninthe
regionwillbeinefficientlylowduetointerregionaltaxcompetitiontriggeringthedescribed
horizontalexternality.Inneglectingthepositiveimpactitstaxhikehasonotherjurisdictions
byexpandingtheirtaxbase,themarginalcostofpublicfundsperceivedbyeachregionis
greaterthanthe‘true’socialmarginalcostofpublicfunds(MCF>SMCF).14
Thesituationchangeswhenregionsarenolongersmall(suchthatthenetreturnoncapital
equalstheexogenousworldmarketinterestrate),butwhenlargeregionsinfluencethenet
returnoncapitalwithinthefederationbytheirchoiceoftaxrate.15ThebasicZMW-result
stillholds,butataxhikeinoneregionsubsequentlytriggersanegativehorizontalexternality
byreducingthefederation-widenetreturnoncapital.16Thatnegativeexternalityattenuates
thepositivehorizontalexternalitycharacterisedbyZMW,suchthattheMCFperceivedby
oneregiondeviatesfromtheSMCFbyalesseramountthaninthecaseofsmallregions.
AfurthercentralresultoftheZMWmodelisthatwhilethemobilefactorwillbesubject
tolowtaxrates,theresultinglossofrevenueswillbecompensatedbyhighertaxrateson
theimmobilefactor.BondandSamuelson(1989)orRazinandSadka(1991)evenreachthe
solutionthattaxratesonmobilecapitalwillbezeroinequilibrium.17Incontrast,ifone
countryismodelledasaStackelbergleader,Gordon(1992)findsthatpositivecapitaltax
rateswillbechosen.Furthermore,Goodspeed(1998)pointsoutthattheresultofazerotax
12AsimilarapproachcanbefoundinWildasin(1989).
13TheconsiderationofmorethanjustonetypeoftaxratecanbefoundinBucovetskyandWilson(1991)
2005).(1991,Wilsonor14ZodrowandMieszkowski(1986)alsoexaminethelevelofsupplyofaproductivepublicgoodandalso
obtaintheresultofitsunderprovision.Yetthisresultisdrivenbyaspecificassumptionnecessaryforthe
stabilityoftheirderivedequilibrium,whoseeconomicinterpretationwascriticisedbyNoiset(1995).The
resultisnolongerobtainedwhenthatassumptionisdropped,asshownbyDhillonetal.(2007)whofind
thatpublicgoodprovisioncantheneitherbetoolow,toohighorefficient.Alongthesamelines,Keenand
Marc15hand(1997)findindicatorsofoverprovisionofproductivegoodsasopposedtoprivategoods.
Wildasin(1988)orHoyt(1991)pursuesuchanapproach.
16ThathorizontalexternalityisalsocharacterisedbyDePaterandMyers(1994),whospeakofa‘pecuniary
y’.externalit17ThatresultisincorrespondencewithDiamondandMirrlees(1971),accordingtowhomproduction
efficiencyrequiresforthegovernmentstodispensewithitstaxation.

13

TillmannLisa

FederalTaxCompetition

rateisdrivenbytheimplicitassumptionthatcapitaldrawsnoutilityfromtheprovisionof
publicgoodsandwillnolongerensueifthelatterisdropped.
Varioussubsequentpapersdismisstheassumptionofidenticalregionsinfavourofthe
recognitionofasymmetrictaxcompetition.Whileinthoseapproachesconsideringidentical
regions,theresultingsymmetricNashequilibriumleadstoanefficientallocationofcapital,
ifthissymmetryassumptionisdropped,theensuingasymmetricequilibriumwilldistort
capitalallocation.Bucovetsky(1991)orWilson(1991)considerregionsdifferinginsize,as
expressedbythenumberofinhabitantsandtheamountofcapitalavailable.18Theygenerate
anasymmetricequilibrium,inwhichthelargeregionchoosesahighertaxratethanthe
smallregion.Thatresultcanagainbeattributedtothelargeregion’simpactontheoverall
netreturntocapitalwithinthefederation,suchthatitstaxbasewillreactlesselastically
towardsachangeintaxratesthanthatofthesmalljurisdiction.Publicgoodprovisionwill
beinefficientlylow,yetwelfareinthesmallregionwillbehigherthaninthelargeone,given
thattheformerisanetimporterofcapitalandprofitsfromthepositiveexternalitytriggered
byitscompetitor.Cooperationwouldraiseoverallwelfare,yetifdifferencesinpopulation
aresufficientlylarge,thesmallregionwillbeworseoffundercollusion.Inanextention,
DePaterandMyers(1994)examinethestrategicbehaviouroflargeregionsmanipulatingthe
netreturnoncapitalintheirfavour,wheretheresultsinturnhingeonwhetheracountry
isanimporteroranexporterofcapital.Thatis,intheformercase,acountryhasan
incentivetoraiseitstaxrate,suchthattheresultingdropinthenetreturnoncapitalis
partiallyburdenedontheforeigninvestors.Thereverseholdsinthelattercase.Thereby,
thetendencytowardsanunderprovisionofthepublicgoodwillbeattenuatedforcapital
importingregionsandaugmentedforcapitalexportingregions.
InanotherextentiontotheZMW-model,Brueckner(2000,2004)examinestheeffect
ofheterogeneouspreferencesregardingthesuppliedpublicgoods.Heshowsthatwithan
increasingvaluationofthepublicgood,equilibriumtaxratesandlevelsofprovisionwillrise
atthecostofcapitalinvestmentsandwages.Ifoneregionisamonopolistinitsproduction
andexportsittoaneighbour,Noiset(2003)findsthatthemonopolistcanpartiallyimpose
theburdenofacapitaltaxhikeontheconsumersintheotherregionviatheresulting
prices.inmark-upLee(1997)considersatwo-periodmodelinwhichcapitalissubjecttotransactioncostsin
thesecondperiodafteritwasinvestedfreelyinperiodone.Thatlimitationofinterregional
capitalinvestmentmobilitypointstowardsinefficientlyhightaxrates,giventhatataxhike
incursanegativehorizontalexternalitybyreducingthenetreturnoncapital,whiletheloss
ofcapitaltaxbasewillbesignificantlylowerandmayevenapproachzero.
18Bucovetsky(1991)assumesforaquadraticproductionfunction,whileWilson(1991)considersageneral
oneaswellasfurtherincludingthetaxationofmobilelabourtoshowthatBucovetsky’sresultsstillhold.

14

TillmannLisa

FederalTaxCompetition

Cross-regionallandownershipwithinafederationasameanstointernalisetheoccuring
positiveexternalitiesisdiscussedbyLee(2003).Aregion’sMCFwillthenbethecloserto
theSMCF,themoreinterrelatedownershipstructuresintherespectiveregionsare.
Ifcountriescompeteforcapitaltaxbaseviathechoiceofcapitaltaxratesandecological
standards,OatesandSchwab(1988)findthattheprevalenceofataxoncapitalimplies
inefficientlylowecologicalstandardsacrossthefederationinsuchthateachregionneglects
thepositiveeffectittriggersontheothersbyraisingitseco-standardandtherebylosing
ase.btaxcapitalAninterestingrecentcontributionbyKempfandRota-Graziosi(2010)includestheen-
dogenisationofleadershipinafederationofcompetingregions.Theyfindthatifregions
displayasufficientlylargedegreeofasymmetry,taxrateswillbedrivendownwardslessthan
standardtheorypredictsandthatthelargercountrywillnotnecessarilychooseahigherlevel
ates.rtaxofWithintheliterature,remediesfortheresultinginefficiencieshavebeensuggested,for
example,bycooperation,19Pigouviansubsidiesortaxes,20ortheintroductionofafiscal
equalisationscheme.21Thesearediscussedinmoredetailattheendofsection2.1.2.

LeviathanGovernments
Whentheassumptionofbenevolentgovernmentsisdroppedforarevenue-maximisingLeviathan
government,asprominentlysuggestedbyBrennanandBuchanan(1977,1980),eachregional
governmentchoosesataxratesuchthattheLaffercurvewithrespecttoitsregionalbudget
ismaximised.Withinterregionaltaxcompetition,eachdecision-makerneglectstheimpact
itsdecisionshaveonrevenuesintheotherjurisdictions,therebytriggeringapositiveex-
ternalitythatdistortstheLeviathan’staxratechoiceincontrasttothecentrallyplanned
optimum.EachregionalLeviathangovernmentthuschoosesinefficientlylowtaxratesin
suchthatcentrallycoordinatedtaxsettingwouldraiserevenuesforallregions.Brennanand
Buchanan(1980),however,explicitlydismissthatpossibilityofcollusion,accentuatingthe
welfare-enhancingeffectsofdecentralisedtaxationbyreducingthewastefulactivitiesand
thusincreasingcitizens’welfare.22
WhenthegovernmentisassumedtobeaLeviathanonlytoacertaindegreeandbenev-
19BoadwayandWildasin(1984).
20Wildasin(1989)DePaterandMyers(1994)orBucovetskyetal.(1998).
21K¨othenb¨urger(2002).
22Infact,Wilson(2005)questionstheanalysisoftaxratechoicesbyLeviathangovernments.Heargues
thatrevenue-maximisingpolicymakershaveanincentivetochoosethe(lesstransparent)amountofpublic
expendituresratherthantaxratestomaximiserents.Heassertsthatvoterschoosingapositivetaxrateon
interregionallymobilecapitalmaythenincreasepublicsectorefficiency.

15

TillmannLisa

FederalTaxCompetition

olentotherwise,23EdwardsandKeen(1996)examinethepossibilityofwelfare-increasing
cooperationagreements.Theyfindthatthesearedesirableifthemarginalexcessburdenof
taxationexceedstheriseinwastefulexpenditurethegovernmentincursforamarginalrevenue
increase.24Parry(2003)aimstoquantifythewelfarecostofinterregionaltaxcompetition
andconcludesthatitwillbeparticularlylowwhengovernmentsarenolongermodelled
asperfectlybenevolentandwillapproachzeroforpureLeviathans.ZissimosandWooders
(2005)considerLeviathangovernmentsinamodelwherepublicgoodsservetheregionsto
differentiatethemselvesfromotherjurisdictions.Iffirmsdifferintheirappreciationofthe
publicgood,governmentscanusethisfacttomitigatetaxcompetition.Theyshowthat
withinthissetting,suchanalleviationofthedegreeoftaxcompetitionwillreduceefficiency.
K¨othenb¨urger(2005)considersthepossibilityofafiscalequalisationgrantschemeformoder-
ateLeviathangovernments.BrennanandBuchanan(1980)hadarguedthatsuchinstitutions
willprovideopportunitiesforpureLeviathanstocolludeandthusdeterthewelfare-enhancing
effectsoftaxcompetition.K¨othenb¨urger(2005)supportsthisargumentforpureLeviathans,
yetshowsthataredistributionschemewillinduceamoderaterevenue-maximisertochoose
alowertaxratethanitwouldotherwise,ifthenumberofcompetingregionsislargeand/or
iftheincumbentsplaceahighvalueonvoterwelfare.BesleyandSmart(2007)discusshow
taxcompetitionaffectswelfarewhenvotersdonotknowforcertainwhethertheirgovern-
mentsarewelfaremaximisersinasignallingmodelwithelectionsandrent-seekingbehaviour.
Amongotherthings,theyconcludethattaxcompetitionwillmostlikelyimprovewelfareif
policymakersaremorelikelytobebenevolentandviceversa.Thiscounterintuitiveresult
isreferredtoanimprovementofselectionprocessesoftheLeviathanpolicymakerswhen
thesearefewinnumbers.JanebaandSchjelderup(2009)considerregionscompetingfor
mobilecapitalandexplicitlymodeltherespectiveelectionprocesses.Forbenevolentgov-
ernments,theyfindthatwelfarewillbelowerinanopenasopposedtoaclosedeconomy,
whilethereverseholdstrueforLeviathangovernments,unlesstheappreciationofthepublic
goodissufficientlylarge.FormoderateLeviathans,theyfurtherfindthattherespective
legislativebargainingprocessesfurtherinfluenceoutcomesinsuchthatunderapresidential-
congressionalregimetaxcompetitionwillraisevoters’welfare,whileunderaparliamentary
democraticsystemthewelfareofvotersislikelytobereducedduetotaxcompetition,if
voterssufficientlyvaluetherespectivepublicgood.
23ThatassumptionmaybejustifiedbytheintentionofaLeviathangovernmenttobere-elected.
24Inasimilardirection,Rauscher(1998,2000)analyseswhethertaxcompetitionwillraisepublicsector
efficiencybyreducingthewasteoftaxrevenuesandinanextentionexaminestheeffectoftaxcompetition
onpublicsectorinnovation.

16

TillmannLisa

FederalTaxCompetition

2.1.2FederalTaxCompetition
Intheprevioussection,afederationwascharacterisedbyanumberofstatescompetingfor
mobiletaxbasewithoneoverarching,thoughpassive,federalgovernment.Yet,asKeen
(1998)pointedoutinhisseminalpaper,realitycallsfortheexplicitrecognitionofthefederal
levelasaplayerinitsownright.25Theearliestcontributionsinthatrespectcanbetraced
backtoCassingandHillmann(1982),Flowers(1988)orJohnson(1988).Dahlby(1994,
1996),BoadwayandKeen(1996),Boadwayetal.(1998)and,mostprominently,Keen
(1998)broughtthefocusofresearchtowardsanactiveroleofthefederalgovernment.
MostofthemorerecentapproachesextendthestandardZMW-modeltoincludeanother
levelofgovernment,asprominentlydonebyKeenandKotsogiannis(2002,2003).Bycon-
sideringnidenticalregions,symmetricequilibriathatdonotdistortcapitalallocationarise
attheregionallevel,suchthatthefocusofanalysiscanberestrictedtotheefficiencyof
chosentaxrates.Theanalysistypicallyconsidersinterdependenciesresultingfromtaxbase
overlap.26Thatis,upper-andlowerlevelgovernmentseachlevyataxonabasethatdisplays
somedegreeofoverlapbetweenthedifferentlevels.Mostmodelsassumetaxbasecongru-
ency,inwhichcasethenfederaltaxbaseBcanbedenotedasthesumofthenregionaltax
basesbi,thatisB=bi.Whenfederalandstategovernmentseachraisetheirownrevenue
=1ifromtaxation,theaggregatetaxburdenτiprevalentinoneregionisthesumofthefederal
taxrateT27andtherespectiveregionaltaxrateti,suchthatτi=T+ti.
Nexttothepreviouslycharacterisedhorizontalexternalitesattheregionallevel,each
region’schoicewillalsoaffectthefederaltaxbase,whichtheregionalpolicymakerdoesnot
25Thesimultaneoustaxationofthesametaxbasebytwolevelsofgovernmentisrelevanttothetax
systemsofmanycountries.Nexttothefederallevel,Canadianprovinces,SwisscantonsandtheU.S.states
imposetheirowncorporatetax.Theco-existenceofthelocalbusinesstaxandthefederalcorporatetax
inGermanyareanotherexample,althoughthereisnoperfecttaxbaseoverlap.Seealsosection5.2fora
discussion.further26Thissurveywillfocusonsuchapproaches.Othermodelsconsidertheeffectsoftaxdeductionschemes
betweendifferentlevelsofgovernment,inwhichtypicallyasingleregionalgovernmentisincluded,suchthat
onlyverticalexternalitiesoccur.See,forexample,Dahlbyetal.(2000)whodiscusstheoptimaldeductionof
regionalfromfederaltaxesinordertointernalisetheverticalexternalitytriggeredbyaregionaltaxhike.In
anextentionofaspatialcompetitionmodeldevelopedbySalop(1979),FlochelandMadies(2002)findthat
deductibilityoftheregionalfromthenationaltaxloadwillleadtotheverticalexternalitytriggeredatthe
regionallevelbeingintensified:Aregionaltaxhikenotonlytriggersaverticalexternalityatthefederallevel
byindividualsrelocatingtheirinvestments,italsoleadstotheamountsdeductiblefromthefederaltaxload
toincrease.Wrede(2002)analysesthereversecase,butgeneratesanalogousresults.Additionally,Flochel
andMadies(2002)extendthescopetoincludeseveralregionalgovernmentsandfindthatdeductibilityof
regionaltaxesreducesthehorizontalelasticityofthetaxbase.Regionswillthusgaintaxsettingpowersuch
thattheyfurtherincreasetheimpactoftheverticalexternality.
27Itistherebyimplicitlyassumedthatthefederalauthoritiessetoneuniformtaxacrossallregions.

17

TillmannLisa

FederalTaxCompetition

accountfor.Dependingonwhethercapitalsupplyisfixedwithinthefederationorhasthe
opportunityofbeingrelocatedabroad,theimpactofaregionaltaxhikeonfederaltaxbase
willbecharacterisedby
nBti=biti≤0.
=1iThatis,inaclosedeconomy,foragivenregionaltaxhikecapitalwillrelocatebetweenregions
untilitsnetreturnisequalacrossthefederationandnoverticalexternalitywilloccur.Ifcap-
italcanberelocatedabroadorisinsomeotherwayendogenisedasafunctionoftaxrates,28
averticalexternalityistriggeredbynarrowingthefederaltaxbase.Pointingtowardsineffi-
cientlyhighregionaltaxation,itmightattenuatethehorizontalexternalitytriggeredatthe
regionallevelandevenresultininefficientlyhighconsolidatedtaxrates.Thewelfareeffects
againhingeonwhetherthegovernmentsareassumedtobebenevolent29orLeviathans.30
Theadditionalrecognitionoftheverticalexternalitytherebypointstowardsapossiblerise
ofwelfarethroughanincreaseinpublicgoodprovisionforbenevolentgovernmentsoraloss
ofwelfarebyreducingtheextenttowhichtaxcompetitioncurbstherevenue-maximising
Leviathangovernments’intentions.
Theresultsfurtherhingeonthetimingofdecision-making.Thatis,whileregionsare
typicallyassumedtoengageinNash-competition,thefederalgovernmentcaneitheractsi-
multaneouslywiththeregions,asmodelledbyGrazziniandPetretto(2007)orWrede(1996),
oritcanactasaStackelbergleader,asmodelled,forexample,byKeenandKotsogiannis
(2002,2003).IfitplaysNash,itwillchooseitstaxrateefficientlygivenstatetaxation.Ifit
isaStackelbergleader,thefederalgovernmentwillperfectlyanticipatethetaxratechoices
ofthefollowingregions(whotakethefederaltaxrateasgiven)andadaptitstaxchoice
accordingly.ItmaysetitstaxratehigherorlowerthanintheNashequilibriuminorderto
induceregionalpolicymakerstochooseefficientlevelsoftaxation.Thatwillinturnhinge
onthestrategicrelationshipoftaxratesatbothlevels.Iftheyarestrategicsubstitutes,
federaltaxationwilltendtobehigherthanintheNashequilibrium,andlowerinthecase
ofstrategiccomplementarity.Therespectivewelfareeffectsoftheresultingequilibriaagain
dependonwhethergovernmentsareassumedtobebenevolentorLeviathans.
28Forexample,KeenandKotsogiannis(2002,2003)orGrazziniandPetretto(2007)modelthesupplyof
capitalbymeansofthepropensityforsavingsSincontrasttoconsumption,whichhingesonthenetreturn
to29capitalKeenρand,sucKhthatotsogiannisS=(S(ρ2002,),where2004),ρisWilsoninturn(2003)aorfunctionMadiesofthee(2008)ffectivdeveeclopapitalataxpproachesburden.withwelfare-
rs.eecision-makdmaximising30Asmodelled,forexample,byFlowers(1988),Wrede(1996,2000)orKeenandKotsogiannis(2003).
WiggerandWartha(2004)considerthecaseofamoderateLeviathangovernment.

18

TillmannLisa

BenevolentGovernments

FederalTaxCompetition

Withpolicymakersaimingtomaximisevoters’welfare,jurisdictionsateachlevelfinancethe
provisionofpublicgoodsthroughtherevenuesgeneratedfromtaxation.Publicgoodspro-
videdatthefederalorregionallevelaretypicallyassumedtobeperfectsubstitutes.31Federal
governmentsareassumedtoprovidetheirfundsequallyamongtheregions,incorrespondence
withtheshareeachregion’spopulationhasinthecountry’stotalsize.Hence,eachregional
governmentincorporatesonlyn1thoftheimpactitsdecisionhasonfederalbudgets,withthe
resultingverticalexternalitypointingtowardsinefficientlyhightaxation.32Theefficiency
ofregionaltaxratesthendependsonwhicheffect(horizontalorvertical)dominates.An
increaseofoverallwelfarefollowingacoordinatedtaxhikeattheregionallevelimpliesthat
regionaltaxrateswillbechoseninefficientlylowinequilibrium.Thereversepointstowards
inefficientlyhighregionaltaxratesandthedominanceofverticalexternalities.
Themostwidelyrecognisedcontributionconsideringwelfare-maximisinggovernmentsin
thecontextoffederaltaxcompetitionisbyKeenandKotsogiannis(2002)whoconsider
atwo-levelgovernmentwithnjurisdictionseachinhabitedbyasingleindividual.33They
concludethatwhetherornottheverticalorthehorizontalexternalitywilldominate,hinges
ontherelativeappreciationforconsumptionofpublicandprivategoods,whichfollowthe
usualInada-conditions.Furthermore,theresultisdrivenbytheelasticityoftaxbasewith
respecttotheprevalenttaxburden,thatis,theinterestelasticityofsavingsandcapital
demandwithrespecttotheinterestrate.Itisshownthatforinelasticsavings,thehorizontal
externalitywilldominateandviceversa.Wilson(2003)pointedoutthatifonlyoneregion
isconsideredatthelowerlevel,withlocalandfederalpublicgoodsbeingperfectsubstitutes,
thefirst-bestcentrallyplannedoptimumforconsolidatedtaxratescanbeachievedbythe
centralgovernmentchoosingTsuchthatτisefficient.Forthattobethecase,itisirrelevant
whetherthefederalgovernmentactsasaStackelbergleaderorplaysNash.Inavariation
ofthemodelbyKeenandKotsogiannis(2002),Madies(2008)examinescapitaltaxation
ifaproductivepublicgoodisprovided.Hefindsthathorizontalexternalitiesmaythen
resultininefficientlyhighprovision,giventhatnexttothewell-knownpositivehorizontal
externalityataxhikeinoneregionwilltriggeranegativehorizontalexternalityinsuchthat
theresultinghigherlevelofproductivegoodprovisionmayinturnattractcapitalintothat
federationandtherebyreducetheotherregions’taxbase.Ifthatlattereffectdominates,the
horizontalexternalitiespointtowardsinefficientlyhightaxation,suchthathorizontaland
verticalexternalitiespointinthesamedirectionandmayaggravateeachother.
31See,forexample,Keen(1998),p.478.
32Forinstance,bereferredtoKeen(1998)orHindriksandMyles(2006),p.579.
33EarlierapproacheswereformulatedbyDahlby(1996)forcommoditytaxationorSobel(1997)inarather
generalessayontaxbaseoverlap,whereonlyonelower-leveldecision-makerisconsidered.

19

TillmannLisa

FederalTaxCompetition

ThedegreeofinternationalcapitalmarketintegrationisconsideredbyWrede(2002),
whoexaminesthetwopolarcasesofperfectintegrationasopposedtoaclosedeconomy.
Intheperfectlyintegratedsmalleconomy,wherecapitaltaxbaseisperfectlymobileand
thenetreturnoncapitalisdefinedovertheworld-marketinterestrate,strongbottom-up
verticalexternalitiesoccurasopposedtothehorizontalexternalities,whichvanish.Forthe
reversecaseofaclosedeconomy,onlyhorizontalandnoverticalexternalitiesoccur,given
thatthecapitalstockisfixedwithinthefederation.Wilson(2003)drawsalinebetween
theseresultsandthosebyKeenandKotsogiannis(2002),insuchthatthecaseofperfect
capitalmarketintegrationissimilartothatofperfectelasticityofsavingswithrespectto
theinterestrate.KeenandKotsogiannis(2004)themselvesextendtheirpreviousmodelto
analysehowavariationinthenumberofregionswithinthefederationinfluencestheresults.
Theyfindthat,whichevertheequilibriumoutcome(inefficientlyhighorlowtaxrates,that
is),anincreaseinthenumberofcompetingstateswillfurtherintensifythatinefficiency,such
thatwelfarewillbereduced.

LeviathanGovernments
ForLeviathangovernments,policymakersagaincareonlyabouttheirownrevenuesand
therebyneglecttheexternaltaxbaseeffecttriggeredbythechosentaxregimewiththe
resultingexternalitiesagainpointinginoppositedirections.
Flowers(1988)wasthefirsttopointtowardspossibleinefficienciesresultingfromaco-
occupationoftaxbasebetweendifferentlevelsofrevenue-maximisinggovernments.She
statedthatbyabstractingfrompossiblehorizontalexternalities,aggregatetaxationwillbe
toohigh.WiggerandWartha(2004)alsoexcludehorizontalexternalitiesbyconsideringonly
onegovernmentateachlevel.TheyexaminemoderateLeviathansandfindthattaxrates
andpublicgoodprovisionwillbeinefficientlyhigh,suchthatadropintaxrateswillraise
taxrevenues.Theprovisionofpublicgoodswillbedistortedinsuchthatthegovernment
imposingthelowertaxratewillhaveagreaterincentivetooverprovidethepublicgoodgiven
thatitunderratestheMCFmorethantheotherlevelgovernment.Wrede(1996,2002)
examineswhetherafederationofLeviathanswillalwaysfinditselfinanequilibriumwith
inefficientlyhightaxratesiflower-levelinterregionalcompetitionisincluded.Heconcludes
thatahigherelasticityofthetaxbaseandinternationalcompetitionfortaxbasethen
functionbylimitingtheprobabilityofthefederationendinguponthenegativelyslopedside
oftheLaffercurve.Healsoexaminestheextinctionofinterregionaltaxcompetitionthrough
thecreationofataxcartelattheregionallevel.Equilibriumisthencharacterisedbythe
respectiveverticalexternalitiesonly.Wrede(2000)considersLeviathanfederationsinwhich
expendituresforapublicproductiongoodareameanstoexpandthetaxbase.Hefinds
thattaxeswillbetoohigh,whileexpenditureswillbetoolowifbothlevelsofgovernment

20

TillmannLisa

FederalTaxCompetition

providethepublicgood.
Inawidelyrecognisedpaper,KeenandKotsogiannis(2003)characteriseandidentifythe
respectiveexternalitiestriggeredateachlevel.TheyshowthatforLeviathangovernments,
whetherornotconsolidatedtaxrateswillbeinefficientlyhighorlowhingesontheelasticity
ofthetaxbase(whichtheyendogenisewithrespecttotheinterestonsavings)aswellas
therelativestrategicinteractionoffederalandregionaltaxrates.Severalvariationsof
thatapproachconsiderhowtheintensityofinterregionaltaxcompetitioninafederalsetting
affectstheequilibrium.KeenandKotsogiannis(2003)themselvesraisethenumberofregions
inordertointensifytaxcompetition.Referringtoanearlierversionofthatpaper,Wrede
(1996)examineshowtheassumptionofhorizontallyperfectlymobilecapitalchangesthe
results.FlochelandMadies(2002)reduceinternationalmobilitycostsforcompaniesand
furtherdiscusstheeffectofapossibledeductibilityofregionaltaxloadsfromthefederaltax
burdenasalimittointerregionalcapitalmobility.Allstudiesreachthesameconclusion:
Namely,thatanincreaseininternationaltaxcompetitionwillraiseconsolidatedrevenues.
Theintuitionbehindthisresultisthefollowing:Forintensifiedinterregionaltaxcompetition,
eachsingleregion’stax-settingpowerisreducedandconvergestozero,suchthattherelative
impactofthefederalgovernment’sdecisionsrises.Itwillthusapproachtheoutcomereached
byacentralplanner,therebymaximisingconsolidatedrevenuesthroughtheappropriate
choiceofT.

GainsEfficiencyVariouspossiblemeasuresforefficiencygainsinafederalsettinghavebeenexamined.For
one,followingtheresultofinefficientlylowtaxratesattheregionallevel,regionalbudgets
couldberaisedbyacoordinatedtaxhikeofallregions.34Ifthefederalgovernmentcan
perfectlydisposeoffederaltransferschemes,ithasameanstoinstallthesecond-bestefficient
35outcome.Overtaxationofacommontaxbasecanbecomparedtotheexcessiveusageofacom-
mon,36suchthattheapplicationofaPigouviantax37orthedefinitionofpropertyrights38
34ThiswillbeafeasiblesolutionfromtheperspectiveofaLeviathangovernment.Ifgovernmentsare
benevolentandthefederaltaxbaseiselastic,theeffectofcollusionhingesonwhetheroverallwelfarecould
beraisedbyahikeinconsolidatedtaxrates.See,forexample,Wrede(2000)orKeenandKotsogiannis
2003).(2002,35See,forexample,Keen(1998),Wrede(2002)orWilson(2003).BoadwayandKeen(1996)hadpreviously
pointedoutthattheexplicitrecognitionoffederalstructureinfluencesthedesignofatransferschemeby
possiblymakingtransfersfromthestatestothefederallevelefficient,insteadofthe-typicallyassumed-
around.yawother36See,forexample,KeenandKotsogiannis(2002).
37(1947).igouPSee38SeeCoase(1960).

21

TillmannLisa

FederalTaxCompetition

maybejustified.Inthatrespect,Dahlby(1996)recommendstherightforthebenevolent
federalgovernmenttoimposeataxonregionaltaxrevenues,wheretherespectiveratewould
bechoseninorderfortheMCFperceivedbytheregionstoequaltheSMCF.Sobel(1997)
suggestsfortheregionalgovernmentstobegiventheexclusiverighttotax,subjecttothe
conditionthatacertainminimumlevelofrevenuesmustbegenerated.Hetherebyseesthe
regionsengageincompetitionforthetaxbasebychoosingthemostefficienttaxsystem,
ratherthanthelowesttaxrates.Keen(1998)suggeststhatapropertyrightfortaxationof
acommontaxbasecouldbeattributedtoonelevelofgovernment,39withtheobligationto
distributerevenuesacrossthefederation,therebyrealisingthefirst-bestsolutionofacentral
planner.

2.1.3InternationalFederalTaxCompetition
Overtheyears,theliteratureontaxcompetitionhasproducedamultitudeofpaperson
thetheoreticalconsequencesofinternationalcompetitionformobilecapital,whoseprinciples
applyjustaswelltolower-leveltaxcompetitionwithinafederation.40Withtheexplicit
recognitionof(multiple-layer)federalstructure,theinclusionofthosefindingsintothein-
ternationaltaxcompetitionliteraturewasonlyasmallstep.
Inthatrespect,Wrede(2002)41hasdevelopedageneralmodeloftwoidenticalLeviathan
federations42engaginginNashcompetitionforinternationallymobiletaxbasebychoice
ofa(distortive)taxrate.43Alllevelsofgovernmentinbothcountriesaremyopicandthe
worldwideprovisionoftaxbaseiselastic.Withinonefederationthesameverticaland
horizontalexternalitiesariseasdescribedintheprevioussection.Inaddition,positivecross-
nationalhorizontalexternalitiesareidentified.Thatis,eachcountryneglectsthatfora
taxhiketheothercountry’staxbasewillexpandbycapitalrelocatingtowhereitsnet
returnishighest.Wrede(2002)thenexaminestherevenue-effectofdomesticincontrastto
cross-nationalcollusion.Cooperationagreementswithinthefederationwilltheninternalise
theinner-federalexternalities,but,obviously,notthoseacrosstheborders.Therevenue-
effectofacoordinatedreductionintheconsolidateddomestictaxessentiallyhingesonthe
strategicinteractionofdomesticandforeigntaxrates.Itwillbepositiveiftheyarestrategic
substitutesorsufficientlyweakcomplements.44Across-countrycooperationconsistingof
39isTheconsideredcentralandlevnoel,hthatorizonis,talifesevffectseraloccur,regionsthatarepropsubjertectytorightchorizonouldtalbetaxcattributedompetition.toanIfyoofnlytheonelevels.region
4041InSeeassectionubsection,2.1.1.butlessextensive,alsoWrede(1996).
4342ANottsomepwithstandingointoftthatheitanalysis,abstainsforfromsimplificationthecpurponsiderationoses,ofoneaosfptehecificcountax,triestheisaapproacssumedhtcoanbebeunitaryeasily.
appliedtocapitaltaxation.
44Inthelattercase,taxpayerswillalsoprofitfromsuchcollusion.

22

TillmannLisa

FederalTaxCompetition

coordinatedtaxhikeswillberevenue-increasingiftheworldwidetaxbaseissufficiently
45inelastic.Forwelfare-maximisingcompetingfederations,Wrede(1997,2002)analysesanextention
oftheZMW-model.Mutuallymyopicgovernmentsareassumedtochoosetaxratesandthe
amountofexpendituresundertaxbasecongruencyandperfectsubstitutabilityoffederaland
regionalpublicgoods.Theverticalexternalitiesthentendtodominatewithinthefederations.
Thefurtherresultingcross-borderhorizontalexternalitiespointingtowardsinefficientlylow
taxrateswill,however,dominateintotal,suchthatworld-widetaxratesandlevelsofpublic
goodprovisionwillbeinefficientlylow.
GrazziniandPetretto(2007)consideranextentionofthemodeldevelopedbyKeen
andKotsogiannis(2002)46inwhichafederalcountrycompeteswithaunitarycountry.
Theydevelopathree-stagegameinwhichtheupper-levelgovernmentinthefederation
actsasaStackelbergleaderwithrespecttotheregionsandsimultaneouslyengagesinNash-
competitionwiththenationalgovernmentintheunitarycountry.47UnlikeWrede(2002),
theynotonlyidentifycross-nationalhorizontalexternalities,theyalsocharacterisecross-
nationalverticalexternalitiesfromthenationalunitarygovernmenttotheregionsofthe
federation.Theyfindthatunderspecificassumptionsregardingthestrategicacross-country
relationshipofnational(andregional)taxrates,theconsolidatedtaxrateinthefederation
willbeinefficientlylowandtaxratesintheunitarystatewillbeinefficientlyhigh.48
Throughthechosentaxrates,therelativesizeofverticalandhorizontalfiscalexternali-
tiesinafederationisdependentontheshareoffederalandregionalexpendituresforpublic
goods.49WilsonandJaneba(2005)developamodelinwhichthedegreeofdecentralisation
servesasastrategictoolincompetingforinternationallymobilecapital.Thedomesticcoun-
trychoosestheamountofcentrallyprovidedpublicgoods,suchthathorizontalexternalities
aredominatedbytheverticalexternalities.Theconsolidatedtaxratewillthusrise,which
willshiftthedomesticreactioncurveupwards,alongtheforeigncountry’sreactioncurve,
45Iftheworldwidetaxbaseisfixed,theLeviathangovernmentswillcolludetochoosethehighestpossible
tax46rSeeate.section2.1.2.
47Inthefirststageofthegame,nationalgovernmentsdecideupontheirtaxrates.Instagetwo,theregions
choosetheirtaxratesinaNashgame,takingthoseoftheupper-levelgovernmentsasgiven,andinthethird
stagethecitizenschoosehowmuchtosaveandwheretoinvesttheirsavings.Governmentrevenuesarethen
distributedamongthecitizensintheformoflump-sumtransfers.
48Theseinefficienciescanbetracedbacktothedecision-makersmisjudgingthedegreeofcapitalmobility
andthesensitivityofthenetreturnoncapitaltogiventaxchanges.Neitherofthenational-levelgovernments
isassumedtoanticipatethestrategicreactionoftheothernational-levelgovernmentsinresponsetoatax
hike.Thetax-inducedlossofcapitaltaxbaseisthenoverestimatedaswellasthereduction(increase)inthe
othercountry’staxrate,ifthelatterisacapitalimporter(exporter).Theresultsarealsoessentiallydriven
by49theassumptionthattheunitarycountryneglectsthetaxrateresponsesatthefederation’sregionallevel.
See,forexample,KeenandKotsogiannis(2002).

23

TillmannLisa

FederalTaxCompetition

suchthattheforeigncountry’soptimaltaxratewillrise,andcapitalwillagainbeattracted
tothedomesticcountry.Thesamereasoningholdsfortheforeigncountry,anditisshown
thatbothcountries’welfarewillriseasaconsequenceofthemitigatedunderprovisionof
s.dogopublic

eshApproacEmpirical2.2

Intheliteratureonfederaltaxcompetition,empiricalworkhasparticularlyfocusedonthe
identificationofstrategicinteractionbetweenjurisdictions.Whileonestrandfocusesonthe
horizontalinteractionbetweenjurisdictionsatthesameinstitutionallevel,amorerecent
fieldofempiricalworkisdrivenbythetheoreticalimplicationsoftheexplicitrecognitionof
acountry’sfederalstructureandaimstoidentifytheinterdependenceoftaxsettingacross
differentlevelsofgovernment.Thissectionissetupbyfirstgivingabriefoverviewofthe
econometricmethodsandmodelspecificationissuesregardingtheidentificationofstrategic
taxinteraction.Subsequently,theexistingstudiesaimingtoidentifyinterregionalcompeti-
tionbetweenjurisdictionsatthesamelevelarereviewed,followedbyasurveyofstudiesthat
extendtheirscopetoincludeverticalinterdependencies-wherethesemayeitherincludeboth
horizontalandverticalinteractionorallowforonlythelatter.50Thelastsubsectiondeals
withanotherissuerelevantfortheanalysisoffederaltaxcompetition,thatis,thequestion
whethergovernmentscanbeperceivedasLeviathansorasbenevolentbycharacterisingthe
approachestakentoidentifyLeviathanbehaviouranddiscussingthegeneratedresults.

2.2.1EstimatingFederalTaxCompetition
Theempiricalliteratureoninterregionaltaxcompetitionaimstoidentifycorrelationsbetween
thefiscalchoicevariablesofdifferentjurisdictionsapplyingspatialeconometricsmethods.
Theempiricaltaskisthentoestimatereactionfunctionsofastrategyvariable(here:the
chosentaxrate)injurisdictioniwithrespecttothevectoroftaxrateschoseninother
jurisdictions.Itsslopethencharacterisesthestrategicinterdependenceofchoices,witha
positivecoefficientimplyingacomplementaryrelationshipbetweentherespectivetaxrates,
whilethereverseisthecaseforsubstitutability.
Whileasignificantslopecoefficientpointstowardstheexistenceofhorizontalinteraction,
thelattermaybeattributedmainlytotwocauses.Forone,horizontalexternalitiesare
consideredtoariseasaconsequenceoftaxbasemobility.Ajurisdictionisthusaffected
bythechoiceofastrategyvariableinotherjurisdictionsinsuchthatitstaxbasewill
50Giventhattheexistingempiricalliteratureisnotasabundantasthetheoreticalcontributions,the
focusisnotrestrictedtostudiesoncapitaltaxation,butinsteadthoseconsideringexciseorpersonalincome
taxationareincludedaswell.

24

TillmannLisa

FederalTaxCompetition

changeinresponsetothat.Inourcase,capitalinvestmentswillshiftwithrespecttothetax
burdenprevalentinonejurisdiction.Otherapproaches,particularlybasedintheworkby
Salmon(1987)orBesleyandCase(1995),relatetheexistenceofexternalitiestoinformational
asymmetriesinaprincipal-agentsetting.Theyardstickcompetitionliteratureseesthesource
oftaxinteractionsinagovernment’sattempttomaximiseitschancesofbeingre-electedby
copyingsuccessfulneighbours’policies.51Inordertodistinguishthesetwosourcesoftax
mimickingbehaviour,onemightfurthercontrolforwhetherjurisdictioni’staxbasedepends
onthetaxratechoicesmadeiniaswellasinallotherjurisdictions,whichwouldindicate
fortaxcompetitiontotriggertheobservedeffects.Ontheotherhand,bycontrollingfor
animpactoftaxratesonelectoraloutcomes,yardstickcompetitionmaybeidentifiedasthe
sourceofinterdependencies.52Someauthorsspecificallytrytopinpointwhichofthetwo
effectstheirregressionislikelytocapture,whileotherssimplytestfortheexistenceoftax
mimickingbehaviour.Restrictingitsattentiontostrategicinteractionresultingfromtax
basecompetition,thissurveywillthusexcludethosepapersthatspecificallycharacterise
yardstickcompetitionasthedrivingforcebehindtheirresults.
Evidenceofverticalinterdependenceissoughtbyestimatingtheslopeofthereaction
functionbetweenupper-andlower-leveltaxrates.Thatrelationshiptendstobeofrelevance
particularlyfortheinterpretationoftheoreticalresults,inwhichthequestionwhetherornot
taxratesaresetefficientlyoftenhingesonwhetherfederalandstatetaxratesarestrategic
complementsorsubstitutes.53
Thus,thetaxrateofonejurisdictioniistypicallyregressedonthetaxrateofthe
otherjurisdictionsatthesameleveland,ifverticalinterdependenceisalsoconsidered,on
thefederaltaxrate.Jurisdictionalfixedeffectsaretypicallycontrolledforaswellasaset
ofsocio-economiccharacteristicsofjurisdictioniandasetoffactorscommonacrossall
jurisdictions.54Eachoftheotherjurisdictions’taxratesisfurtherassignedaspecificweight,
suchthatacentralissueconcernsthespecificationoftheappropriateweightmatrix,for
whichseveralapproacheshavebeenbroughtforward.Themostwidelyusedisaweighting
schemeaccordingtosharedborders.Suchaweightwijmaybedichotomous,55withwij=1
whenregioniandregionjshareaborderandwij=0otherwise,ornon-dichotomousinsuch
thatthetotallengthofsharedbordersrelativetoacountry’ssizeisconsideredandthen
normalisedtoone.Recentcontributionshaveformulatedmoreelaborateweightingschemes.
Forinstance,amulti-regionalapproachisappliedtoU.S.statesbyRorkandWagner(2008),
51SeealsoCaseetal.(1993),HeyndelsandVuchelen(1998)orBordignonetal.(2003).Thisliteratureis
surveyedinMadiesetal.(2004).
52Brueckner(2003)hadpointedtowardstheseidentificationproblems.Buildingonthat,Revelli(2005)
dev53otesanentiresurveytothisissue.
54See,See,forforeexample,xample,MKeenadiesandetaKl.(2004)otsogiannisfor(a2002)moreordetailedGrazzinioveandrview.Petretto(2007).
55SeeBrueckner(2003).

25

TillmannLisa

FederalTaxCompetition

whoassignweightswithrespecttothepopulationsizeofthecompetingstateandalso
groupcertainstatesintoregions.Weightsarethenfurtherassignedundertheassumption
thatcross-regionalinteractionswillbedifferenttothoseoccuringwithinaregion.Similarly,
Gerardetal.(2010)applyaweightingschemeinwhichBelgianmunicipalitiesaresorted
intoregionsanddifferingweightsareassignedwithrespecttoinner-regionalorcross-regional
sharing.rderobFurtherissuesstemfromendogeneityproblems.Forone,giventhemutualinterdepen-
denceofthestrategyvariables,thevectoroftaxrateschosenbytheotherjurisdictionsmay
itselfbeendogenousandthuscorrelatedwiththeerrorterm.Theerrortermmayfurther
includespatiallydependentomittedvariables.56Typically,theseissuesaresolvedbythe
applicationofmaximum-likelihood(ML)methods57orbytheimplementationofaninstru-
mentalvariableapproachassuggestedbyKelejianandRobinson(1993)orKelejianand
58(1998).haPrucTheincludedvectorofjurisdiction-specificcharacteristicsmayalsobecorrelatedwiththe
errorterm.Thearisingendogeneityproblemscanbesolvedbyappropriatespecificationof
instrumentsfortherespectivecontrolvariables,ortheymaybecircumventedbytheuseof
paneldata,whichmayalsohelptodealwithspatialerrordependence,aspointedoutby
Brueckner(2003).Thisapproachwastaken,forexample,byRevelli(2001)or,morerecently,
2010).(al.etGerard

2.2.2InterregionalTaxCompetition
Avastarrayofliteratureoninterregionaltaxmimickingbehaviourhasbeenproduced,among
whichthemajorityofstudiesusesEuropeandatasets,withonlysomestudiesontheUnited
StatesorCanada.Sinceexcellentsurveysofthepreviousliteraturehavebeengivenby
Brueckner(2003)orMadiesetal.(2004),thefocusofthisworkliesonthemorerecent
tributions.con

CompetitionOverTaxRatesWithinaCountry
ForAmericandata,moststudiesfocusontaxinteractionbetweenstatesorbetweenlocal-
levelcountiesandmunicipalities.Inearlierstudies,Ladd(1992)orBruecknerandSaavedra
(2001)havefoundevidenceofpositivetaxinteractionformunicipalpropertytaxratesin
56Inthatcontext,Revelli(2008)pointstowardsthepresenceofcorrelatedshocks,suchasanincreasein
federaltransferstoallregions,whichincursregionalgovernmentstoreducetheirtaxrates,nottobeconfused
withevidenceofstrategicinteractionbetweenregionaldecision-makers.
57Theseareapplied,amongothers,bySol´e-Olle´(2003),AllersandElhorst(2005)orRorkandWagner
(2008).58See,forexample,FeldandReulier(2009),Devereuxetal.(2008)orEdmarkandAgren(2008).

26

TillmannLisa

FederalTaxCompetition

theUnitedStates.ForpaneldataonthepropertyandsalestaxesinU.S.statesbetween
1993and2003,Hill(2008)examineswhetheragglomerationhasanimpactontaxsetting.
Hefindsacomplementaryrelationshipbetweentaxratesacrossstatesaswellasevidence
thatagglomerationinfactleadstoanoverallhighertaxburdenandalowerlikelihoodof
taxmimickingbehaviour.RorkandWagner(2008)examineinterstateexcise(cigaretteand
generalsales)taxcompetitionintheUSAforthetimefrom1967to2002.Theyfindsignifi-
cantcoefficientsforinterstatetaxinteractionandfurtherconjecturethatthelatternotonly
hingesonthemobilityoftaxbase,butalsoonthestructureofregionalinterrelationswithin
thecountry.Theyspecificallyfindthatthedirectionandthedegreeoftaxcompetitionvaries
significantlyoverspecificregions.DeskinsandHill(2010)surveypaneldataonseveralU.S.
statetaxesinanattempttopinpointwhethertheinterdependenceofstatetaxrateshasvar-
iedovertime.Theyconfirmtheexistenceofstrategiccomplementarityoftaxratesandfind
significantevidencethattheinterregionalsensitivityofstatepersonalincometaxrateshas
infactsystematicallydiminishedovertime.Theyare,however,unabletogenerateevidence
ofsuchtimetrendsforthecorporateincometax.CrowleyandSobel(2010)analysepanel
dataforpropertytaxationinmunicipalities,countiesandschooldistrictsinPennsylvania,
inwhichthenumberofregionsattherespectivelevelservesasaproxyforitsdegreeof
decentralisation.Theyexaminetheinterrelationsoftaxcompetition,revenue-maximising
governmentbehaviourandspatialinterdependenceoftaxratesinmunicipalitiesandcoun-
tiestofindthatdecentralisationandthusahigherdegreeofinterregionaltaxcompetition
reducestaxrates.Theirresultsforschooldistricts,whicharehighlydecentralised,however,
provideevidenceofstronginteraction,yetsignificantlyhighertaxrates.Theauthorsde-
ducethatspatialcorrelationbetweentaxratesasanindicatoroftaxcompetitionbetween
regionsneedstobeseenwithcaution,giventhatitmightaswellimplyevidenceofcollusion
activities,whichmightinturndrivetaxationupwards.59
ForasampleofBrazilianstatesbetween1985and2001,deMello(2008)findssignificant
evidenceofstrategiccomplementarityinthechoiceofthevalue-addedtax.Hefurtherfinds
thattheseinteractionsaremorepronouncedamongstatesbelongingtothesamegeographic
regionandthatonestate,namely,Bahia,infactactsasaStackelbergleader.
VariousstudiesexistforEurope,giventherichnessofdataandthevarietyofinstitu-
tionalreformsandindividualcountries’regulation.Onestrandofresearchisdrivenby
thedecentralisedstructureofEuropeancountries.Amongthemorerecentcontributions,
Vandenbusscheeetal.(2005)findevidencethatBelgianregionscompeteovertheeffective
corporatetaxrateforatimeperiodfrom1993-2002.Geys(2006)examinesFlemishtaxset-
tingfortheyear2000.Hearguesthatwithrespecttotaxbasemobility,theratioofthetax
leveltotheamountofgovernmentspendinginonemunicipalityratherthanonlythetaxrate
59Seealsosection2.2.4.

27

TillmannLisa

FederalTaxCompetition

mayinfluencelocationchoice.Withathuscreatedindex,hedrawstheconclusionthatsig-
nificantcompetitionfortaxbasebetweenmunicipalitiesdoesexistinsuchthatafavourable
tax-expenditureratioofonemunicipalityislikelytopositivelyinfluencethatratioinother
municipalities.60ForapanelofBelgianmunicipalities,Gerardetal.(2010)explorethesig-
nificanceofinterregionaldifferencesforinteractionsinsettinglocaltaxratesurcharges.They
findthatdecision-makersreactspecificallytowardspropertytaxratesurchargessetbytheir
closestneighboursratherthanallotherBelgianmunicipalities.Further,theresultsindicate
thattaxcompetitionismoreintenseamongmunicipalitieswithintheBrusselsregion.Their
findingsvalidatetheresultofBelgianincometaxratesasstrategiccomplementsestablished
byHeyndelsandVuchelen(1998)andRichard,TulkensandVerdonck(2002),butcontradict
theseauthors’positiveresultsforthelocalpropertytaxsurchargebydetectinginsteadan
inverserelationship.
FordataonSwedishlocalincometaxpoliciesbetween1993and2006,EdmarkandAgren
(2008)testfortaxmimickingbehaviourandexplicitlytrytoverifywhethertheirresultsare
drivenbyyardstickcompetitionortaxcompetition.Theyfindstrongevidenceoftaxrates
beingpositivelyspatiallycorrelated,drivenbymobiletaxbaseratherthanpossibleelection
outcomes.FeldandReulier(2009)examinecompetitionoverincometaxratesamongSwisscantons
forpaneldatafrom1984and1999.Theyfindapositiverelationshipforincometaxrate
reactionsbetweenthecantons.Hereby,theysupportpreviousfindingsforSwissdata.61
ReulierandRocaboy(2009)testforstrategiccomplementarityinthesettingoftaxrates
betweenFrenchregions,specificallyaimingtodiscriminatebetweenyardstick-ortaxbase
competition-drivenbehaviour.TheirresultssuggestthattaxmimickingbetweenFrench
regionalgovernmentsexistsandcanbeattributedtotheformerifthetaxburdeniscarried
bythevoters,whereasitcanbeattributedtothemobilityoftaxbaseifthetaxisimposed
onfirms.CharlotandPaty(2010)studytheimpactofagglomerationontaxcompetition
inFrancefor2002andfindevidenceofcomplementarityintaxrates,butdonotdetecta
significantrelationshipbetweentaxratesandthepopulationdensitywithinaregion.For
urbanjurisdictions,theyfurtherfindapositiverelationbetweenthecapitaltaxrateandthe
prevailingcapitalstock,whichtheyattributetoataxableagglomerationrent,asbrought
forwardbyneweconomicgeographymodels.62
RelativelylittleevidenceexistsontaxcompetitioninGermany.Buettner(1999,2001)
60Thatresultmayalsobeseenasindicativeoftherelevanceoflaboratoryfederalismapproachesoutlined
.2.1sectionin6162SeeTheirFeldafindingsndcKirconhtributegaessnertoaf(2001,airlye2002)xtensivoreFeldliteratureandRoneulierFrenc(ht2005).axsetting,whoseearliercontributions,
forexample,byJayetetal.(2002)orFeldetal.(2002),wereaffirmativeofthetaxcompetitionhypothesis.
SeeMadiesetal.(2004)formoredetails.

28

TillmannLisa

FederalTaxCompetition

examinesbusinesstaxsettinginteractionsbetweenGermanmunicipalitiesforthetimebe-
tween1980and1996,wherehefindsevidenceoftaxmimickingbehaviourwithlargerregions
settinghighertaxratesonaverage.Subsequently,Buettner(2003)specificallytestsfortax
basemobilityasasourceoftaxmimickingbehaviourbetweenGermanmunicipalitiesfrom
1980till2000andconfirmshishypothesis.

CompetitionOverTaxRatesAcrossCountries

AnotherstrandofliteratureonEuropeandataisdrivenbytheEUintegrationandenlarge-
mentprocessandaimstocharacterisestrategictaxrateinteractionsamongmemberstates.
Foradatapanelon11WesternEuropeanstatesbetween1970and1999,Redoano(2007)con-
cludesthatthereispositivestrategicinterdependencebetweencorporatetaxlevelsandsees
thatEuropeancountriesspecificallyinteractwiththeirlargecompetitors,whichtheyuseasa
pointofreference.Further,taxmimickingappearstobeparticularlypronouncedbeforejoin-
ingtheEuropeanUnion,whileitbecomeslesssignificantaftermembershipisaccomplished.
TakingtheEUenlargementasapointofdeparture,CassetteandPaty(2008)examinethe
relationshipbetweentheformerEU-15andCentralandEasternEuropeancountries(CEEC)
fortheperiodbetween1995and2005.Theyfindevidenceofpronouncedcomplementary
taxinteractionwithinWesternEuropeancountries,butlesspronouncedstrategicinteraction
amongtheCEEC.TheirresultsalsopointtowardsinterregionalcompetitionbetweenEast
est.WandFinally,someauthorshaveusedtheobservedincreaseininternationalcapitalmarket
integrationtostudyinternationalstrategictaxinteractions.AltshulerandGoodspeed(2003)
examineEuropeanandU.S.taxsettingtofindevidenceoftaxcompetition,inwhichthe
UnitedStatesactasaStackelbergleader.Theyfurtherconcludethattaxcompetitionamong
Europeancountrieshasbecomelessintense,whileithasgainedmomentumbetweentheU.S.
andEurope.Devereuxetal.(2008)aimtofindproofofcompetitionovercorporatetax
ratesamongOECDcountries.Theyestimateupward-slopingreactionfunctionsforOECD
countriesgeneratedfromatheoreticalmodelshowingthatthemobilityoftaxbasedepends
onanindustry’srelocationcost.Theyconcludethatstrategicinteractiononlytakesplace
amongopencountries,inwhichrelaxedcapitalcontrols,leadingtoincreasedcompetition
formobiletaxbase,triggerareductionintaxrates.Kammas(2009)examinesa1982-2000
paneldatasetforOECDcountriestofindapositiveresponseofcapitaltaxrateswithinthe
countriestowardsthosesetinothercountriesandanegativereponseofcapitaltaxrates
towardspublicinvestmentspending.

29

LisaTillmann

FederalTaxCompetition

TaxBaseMobility
Inadditiontothecharacterisationoftaxreactionfunctions,someauthorstrytoestablish
arelationshipbetweenthemobilityoftaxbaseandtheleveloftaxratesprevailingina
jurisdiction.Interstatetaxcompetitionwithrespecttothedegreeofmobilityofthetax
baseisexaminedbyRork(2003)forfivemajorstatetaxinstrumentsintheUnitedStates
overatimeperiodfrom1967through1996.Hesupportsthetaxcompetitionhypothesis
withhisresultsstatingthatthemoremobilethetaxbaseofastate(as,forexample,motor
fuelortobaccosalesandcorporateincometaxesasopposedtopersonalincomeandgeneral
salestaxes),themoresensitiveitsresponsetowardsataxhikeintheneighbouringregions
willbe.Hefurtherconjecturesthatthereisevidenceofcomplementarityfortaxrateson
mobilebasesandindicationsofsubstitutabilityforimmobileones.MintzandSmart(2004)
usepaneldataoncorporatetaxationinCanadianprovincesbetween1986and1999.They
findevidenceofhigherstatetaxratesreducingtherespectivetaxbase,andoffirmswith
anopportunityforincomeshiftingreactingsignificantlymoreelasticallytowardschanges
intaxrates.Carlsenetal.(2005)examineNorwegianmunicipalitiesandtheirstrategic
behaviourwithrespecttothemobilityoffirms.Forthat,thedegreeoffirmmobilityat
thelocallevelisexplicitlyincorporatedintoanapproachdevelopedandlaterpublishedby
Devereuxetal.(2008).63Theyestablishaninverserelationshipbetweenthemobilityoftax
baseandthelevelofinfrastructurefeesimposedonfirmsattheregionallevel.UsingOECD
paneldata,Winner(2005)concludesthatforthetimeperiodof1965-2000,anincreasein
capitalmobilitypusheslabourtaxationupwards,whilehavinganinverseeffectoncapital
taxation.Hefurtherfindsthatthelargeracountry,themorelikelyitistohaveanoverall
highertaxlevel.TheintegrationofcapitalmarketsanditsimpactonEUwealthandtax
levelsforthetimeperiodbetween1970and1996isscrutinisedbyMendozaandTesar(2005)
withinamacroeconomicframework.Theirevidenceimpliesthatwhilecapitaltaxrates
wereonlyslightlydrivendownwards(exceptintheUK,wherethereductionwassomewhat
morepronounced),labourtaxratesweresubjecttosharpincreases(exceptintheUK,
whereitchangedonlymarginally).TheimpactofmonetaryandeconomicEUintegration
oncorporatetaxcompetitionisalsoexaminedbyKarkalakosandMakris(2008)foratime
spanbetween1975and2005.Apositivecorrelationisobservedbetweencorporatetaxrates
andeconomicintegration,whiletheformerisnegativeformonetaryintegration.Daviesand
Voget(2009)askwhethertherecentEUenlargementhasinfactintensifiedtaxcompetition
forunbalancedpaneldataoveratimeperiodfrom1980through2005.Theyconcludethat
taxcompetitionamongEUmembersismoreintensethanbetweenmembersandotherstates.
63Seeabove.

30

TillmannLisa

FederalTaxCompetition

2.2.3FederalTaxCompetition
Giventhetheoreticallydrivenexplicitrecognitionoffederalstructure,itwasonlyamatterof
timeuntilitsresultsweresoughttobeempiricallyverifiedbyaimingtodetectaninteraction
ofupper-andlower-leveltaxrates.

VerticalTaxRateInteractionWithinaCountry
TheearliestempiricalstudycanbefoundinBesleyandRosen(1998).Theyexaminethe
interactionoffederalandstatelevelgasolinetaxratesintheUnitedStatesoveratimeperiod
from1975till1989.Theirapproachonlyaccountsforverticalinteractionsanddoesnotallow
forpossiblehorizontaleffectsamongthelower-levelgovernments.Thegeneratedresultspoint
towardsapositiverelationshipbetweenfederalandstatetaxes.Varioussubsequentpapers
haveanalysedgasolineandcigarettetaxationattheU.S.federalandstatelevel.Devereux
etal.(2007)considerpaneldatafrom1977-1997controllingfortheelasticityoftaxbase
andcross-bordershopping.Forthetaxationofcigarettes,theyfindasignificantpositive
andlargehorizontaleffect,whiletheverticaleffectsareinsignificant.Gasolinetaxrates
aredrivenupwardsbyverticaltaxinteraction,whilehorizontalinteractionisnotsignificant.
Theseresultscorrespondwiththepredictionsfromtheirtheoreticalmodelarguingthatthe
verticalexternalityislikelytobemoresignificant,themoreelasticthetaxbase.Theyrefer
toempiricalevidencesuggestingthatthisisthecaseforgasolineratherthancigarettes.Fora
panelfrom1975-2001,FredrikssonandMamun(2008)examineU.S.cigarettetaxation.While
theregressionresultscoveringtheentireperiodareinsignificant,theirresultsonthetime
between1982-2001suggestthatstatesarelikelytoreactinverselytowardsataxratehikeat
thefederallevel.Miyamoto(2009)estimatesreactionfunctionsforgasolineandcigarettetax
rates.Hefindsverylittleevidenceofhorizontaltaxcompetitionandonaverageamoderate
positiverelationshipbetweenfederalandstatelevelforbothtaxes.Theresultsonvertical
interactionaresomewhatmorepronouncedforcigarettetaxationandthesignandmagnitude
oftheslopesofthereactionfunctionsdifferacrossstates.
Esteller-More´andSol´e-Olle´(2001)examinepersonalincomeandgeneralsalestaxation
atthestateandfederallevelintheUnitedStatesoveratimeperiodbetween1987and
1996andfindasignificantpositiveinterdependenceofbothstateleveltaxrateswithfederal
taxes.WuandHendrick(2009)examinepropertytaxratessetinmunicipalities(lowerlevel)
andincountiesorschooldistricts(upperlevel)inFloridabetween2000and2004.Their
resultsconfirmtheexistenceofhorizontalandverticaltaxinteractions.Namely,theyfind
positivehorizontalinteractionsatthemunicipallevelandaninversereactionofmunicipal-
leveltaxratestowardscounty-leveltaxrates,whilethelatterispositivewithrespectto
schooldistrict-leveltaxrates.
Rizzo(2010)developsanapproachtoestimateverticalandhorizontaltaxinteraction

31

TillmannLisa

FederalTaxCompetition

effectsforgeneralsalesandtobaccoexcisetaxratesintheUnitedStatesandCanadabetween
1984and1994.Hefindsevidenceofcomplementaritybetweenupperandlowerleveltax
rates.Hisresultsfurtherindicatethatanincreaseinthefederaltaxratemaybeefficiency-
enhancingandmightoffsetthetaxcompetitioneffectsatthelowerlevelbyreducingthe
sensitivityoflower-leveltaxrateresponsestohorizontaltaxcompetition.Usingaproxyfor
excisetaxationofproductiongoodsderivedfromadatasetbetween1970and1997,Crisan
(2007)examinesverticalandhorizontaltaxcompetitioneffectsinCanadianprovinces.His
resultsstronglysupporttheexistenceofverticalexternalitiesandcomplementaritybetween
federalandprovincialtaxrates.Theevidenceonhorizontaltaxcompetitionismixed,with
theslopesoftheprovincial-levelreactionfunctionsbeingeitherpositiveornegative,andnot
allofthembeingsignificant.Hedoes,howeverfindhighlysignificantindicationsthatall
provincesreactpositivelytowardsthetaxratechosenintheprovinceofAlberta,pointing
towardsasomewhatleadingfunctionwithintheCanadianprovinces.
TheregionalpatternofmunicipalcorporatepropertytaxrateswithinBritishColumbiais
examinedbyBrettandPinkse(2000).Theyfindevidenceofhorizontaltaxcompetitionand
notefurtherthattheirresultspointtowardstheexistenceofverticalinterdependencieswith
thestatetaxratesbeingdrivendownwardsbyafederaltaxhike.ForCanadianfederaland
provincialcorporateincometaxratesfrom1963through1996,HayashiandBoadway(2001)
estimatereactionfunctionsandfindevidenceofverticalandhorizontaleffects.Precisely,they
findaninverserelationshipbetweenfederalandprovincialtaxratesandcomplementarity
forprovincial-leveltaxrates.Theyfurtherconcludethattheresultsfortheprovinceof
Ontarioarepeculiarinsuchthattaxsettingisneitheraffectedbytheotherprovincesnor
bythefederallevel.Instead,federaltaxationexhibitsapositiverelationshiptowardsthe
taxlevelinOntario.Finally,Esteller-Mor´eandSol´e-Olle´(2002)estimateanupward-sloping
reactionfunctionofCanadianprovincialpersonalincometaxratestowardsotherprovinces’
andfederaltaxratesfordatafrom1982through1996.
AmongthestudiesconsideringdatafromtheEU,Revelli(2001)examinespropertytax
settinginnon-metropolitandistrictsinEnglandforpaneldatainthe1980s.Heisableto
confirmspatialinteractionbetweendistricts,butfindsnoevidenceofverticalinterdependence
oftaxrates.MunicipalanddepartmentlevelsettingofthelocalbusinesstaxinFrancein
1995isexploredbyGoodspeedandLePrince(2003).Theirestimatespointtowardsan
upward-slopingreactionfunctionofthemunicipalitiestowardsthedepartments.LePrinceet
al.(2007)characterisedepartment(intermediatelevel)andregional(upperlevel)government
taxreactionsforthecaseofbusinesstaxationbyuseofaFrenchcross-sectionfrom1995.
Theirresultssuggestthatverticaltaxinteractionsarenotsignificant,whilehorizontaltax
interactionsatthedepartmentlevelare.Theyfurtherpointoutthatthereisnoevidenceof
theupper-leveltaxchoicesoffsettingtheinefficienciescreatedatthelowerlevel.Fordataon

32

TillmannLisa

FederalTaxCompetition

Swedishlocalandregionalpersonalincometaxationderivedfroma1981-1990paneldata
set,Anderssonetal.(2004)findaninverserelationshipbetweentaxrates.Theyfurther
testforwhethertheupper-levelgovernmentisabletointernalisetheinefficienciestriggered
atthelowerlevelandconjecturethatthisisnotthecaseforlocalandregionaltaxrate
choices.BruelhartandJametti(2006)useaSwisspaneldatasetforselectedyearsbetween
1985and2001.Startingfromthetheoreticalresultthattaxratesatthestatelevelwill
declinewithanincreaseoffragmentationifthehorizontalexternalitiesdominateatthestate
level,andviceversa,theyanalysetheinterdependenceofcanton-levelandmunicipal-level
personalandcorporatetaxratesforallofthe26Swisscantonswithrespecttothedegreeof
fragmentation.Theirevidencesuggeststhatonaverageverticalexternalitiesdominatethe
horizontalexternalitiesamongthemunicipalitiesineachoftheSwisscantons.

VerticalTaxRateInteractionAcrossCountries
Goodspeed(2000)aimstoidentifyhorizontalandverticaltaxcompetitioneffectsfordata
on13OECDcountriesbetween1975and1984.Hefindsthatstate-levelgovernmentsreact
towardsanincometaxhikeatthefederallevelbyreducingtheirtaxrates.Asaproxy
forthemobilityoftaxbaseand,thus,internationaltaxcompetition,heusesstatepoverty
ratesandfindsapositiverelationshipbetweenthoseandthelevelofincometaxrates,which
confirmstheclassicaltaxcompetitionresults.Goodspeed(2002b)supportsandenricheshis
previousresultsbyfurtherspecifyingtheexistenceofinteractionofverticalandhorizontal
effects.Hepointsoutthattheverticalexternalityincreaseswithagreaterequalityoftax
basesacrossregions,thatis,withalowerdegreeoftaxbasemobilityand,thus,horizontal
tition.eompctax

2.2.4EfficiencyEnhancingTaxCompetition
Theevaluationofdecentralisationortaxcompetitiondifferssignificantlydependingon
whethergovernmentsareassumedtoberevenueorwelfaremaximisers.64Thissectionaims
togiveanoverviewoftheexistingstudiesseekingtoempiricallyvalidatetheassumptionof
iour.vehabLeviathanMostoftherespectivecontributionsaimtoestablishalinkbetweenthesizeofgovernment
andthedegreeofdecentralisation.65Decentralisationwillleadtoareductioninthesizeof
governmentundertheassumptionthatitreducestheabilityofgovernmentstomaximisetheir
ownbudgetsthroughbetterpossibilitiestocontrolthepolicymakers.Typically,thedegreeof
decentralisationisproxiedbymeansofonejurisdictionallevel’sshareofexpenditure(revenue)
642.1.sectionSee65onpublicOtherasectorpproaches,efficiencywhicorhwillwhetheralsobeelemendiscussedtsofdirectfurtherdemoon,ecracyxaminewillwhetherreducedpublicecensectortralisationsize.hasaneffect

33

TillmannLisa

FederalTaxCompetition

intotalgovernmentexpenditures(revenues)orthetotalnumberofjurisdictionsatonelevel.66
Competingviewsregardingtheeffectoffiscaldecentralisationcallforacarefulspecification
oftherespectiveregressions.Forone,whilefiscaldecentralisationwillmostlikelyreduce
centralandraiselocalgovernmentbudgets,theaggregateeffectremainsopen.Brennan
andBuchanan(1980)suggestedforLeviathangovernmentstoberestrainedbyhorizontal
competitionformobiletaxbase,suchthatdecentralisationwillraisewelfarebyreducing
theoverallsizeofgovernment.Ontheotherhand,their‘collusionhypothesis’impliesthat
thisneednotbethecase.Infact,upperandlowerlevelLeviathangovernmentshavean
incentivetocolludeandtransfertaxationpowertothecentrallevelinordertocircumvent
therestrainingtaxcompetitioneffects,whileonlyexpendituresremaindecentralised.The
collusionhypothesisthuscallsforthecarefulconsiderationofexpenditureasopposedto
revenuedecentralisation.Severalstudies67furthercontrolforcollusionactivitiesbytaking
intoaccounttheprevalenceoffederaltransferschemes.Incontrast,withreferencetoanidea
byJohnWallis,Oates(1985)assertedthatdecentralisationmayleadtoinefficientlyhigh
localandalsoaggregateexpendituresgiventhattaxpayersarebetterabletoexertinfluence
ontheirlocalincontrasttofederal-levelincumbents,suchthattheformermayfeelmore
inclinedtopursuelocalprojectsleavinglower-levelexpendituresinefficientlyhigh.Hence,
anincreaseofaggregateexpendituresfollowingfiscaldecentralisationneitherconfirmsnor
deniestheLeviathanhypothesis.
Empiricalstudiesthusneedtoanalysetheexpenditureandtherevenuesideofdifferent
levelsofgovernmentaswellastheaggregatelevelandcarefullyaccountforthecontrasting
viewsandimplicationsoftheLeviathanhypothesis,thecollusionhypothesisandtheOates-
Wallishypothesis.

DecentralisationWithinaCountry
Oates(1985)firstattemptedtoempiricallyverifytheexistenceofaLeviathangovernment
forcross-sectionaldataonU.S.stateandlocallevelgovernmentsin1977aswellasinter-
nationaldatafrom43countries.Includingbothstaterevenueandexpendituresasashare
oftotalgovernmentrevenue(expenditures)andthenumberoflocaljurisdictionstomeasure
thedegreeofdecentralisation,heestablishesnoevidenceofLeviathanbehaviour.Forthe
samedataset,aftervaryingthemeasuresofthedegreeofdecentralisationandaccountingfor
heterogeneouspreferencesacrossthestates,Nelson(1987)findsamoderatelynegativerela-
tionshipbetweenthesizeofgovernmentanddecentralisation.EbertsandGronberg(1988)
observeareductioninpublicsectorsizeasaconsequenceoffiscaldecentralisationfor1977
will66beOtherpoinmtedeasures,outwithsuchasregardthetdotegreeheofrespfectivragmenestationtudies.ortaxcompetition,havealsobeendeveloped,which
67See,forexample,Lalvani(2002)orChen(2004).

34

TillmannLisa

FederalTaxCompetition

dataontheU.S.countyandmetropolitanlevel.ForU.S.countycross-sectionaldata,Forbes
andZampelli(1989)concludethatpublicsectorsizeexpandswithahigherdegreeofde-
centralisation,whichtheyperceiveascontradictingtheLeviathanhypothesis.Zax(1989)
findsevidenceinfavouroftheLeviathanhypothesisforacross-sectionaldatasetofU.S.
municipalitiesfrom1982.Hestatesthatlocallevelgovernmentbudgetswillbereducedif
decentralisationraisestaxcompetition,whiletheoppositewillbethecaseifitincursaloss
ofscaleeconomies.WithdataontheU.S.statelevel,Marlow(1988)performsatime-series
regressionfor1946-1985withresultspointingtowardsdecentralisationreducingthesizeof
government.JoulfaianandMarlow(1991)measuretheimpactofthelevelofdecentral-
isationongovernmentbudgetsforU.S.state,localandfederalspendingin1983-1985to
confirmtheLeviathanhypothesisforfederalrevenues.68UsingDataEnvelopmentAnalysis
todeterminetheefficiencyofMinnesotancounties,NoldHughesandEdwards(2000)find
evidenceinfavouroftheLeviathanhypothesis.Theyuseaggregatecountypropertyvaluesas
aproxyforpublicsectorefficiency.ThatapproachwassuggestedbyBrueckner(1982),who
showedthattheSamuelsonconditionforefficientpublicgoodprovisionisconsistentwith
themaximisationofaggregatecommunalpropertyvalues.Thatsameproxyisappliedby
BatesandSanterre(2006),whosupporttheLeviathanhypothesisforcross-sectionaldataon
municipalitiesinConnecticutfor1998.Theygeneratetheresultthatpublicsectorefficiency
isreducedwithhighermarketpower.Stansel(2006)examinesmunicipal-levelU.S.data
from1962through1992.Formunicipalgovernmentspendinggrowthoverthreetimeperiods
(1962-1992,1962-1982and1982-1992),heconcludesthatitwillbelimitedbymoreintense
competition.CrowleyandSobel(2010)69analysepaneldataforpropertytaxratesinmu-
nicipalities,countiesandschooldistrictsinPennsylvaniabetween1995and2005tofindthat
decentralisationintensifiesinterjurisdictionalcompetitionandabatestaxrates.Theyfur-
therfindevidenceofcollusionactivitiesbetweenschooldistrictsraisingtheleveloftaxrates.
TheirevidenceoverallverifiestheLeviathanhypothesisandpointstowardstherecognition
ofcollusionactivitiesnotonlyacrosslevelsofgovernment,butalsoatthesamelevel.
ForapaneldatasetonIndianstateandlocalgovernmentsforselectedtimeperiodsbe-
tween1990and1998,Lalvani(2002)concludesthatdecentralisationyieldsasmallertotal
andupper-levelgovernmentsize,whileitraisesthatofsubnationalgovernments.Shefurther
findsevidenceofcollusionactivitiestriggeringanincreaseinexpendituresatalllevelsof
government.Chen(2004)examinesfiscaldecentralisationattheprovinciallevelinChinafor
68Marlow(1988)hadsuggestedthattheinclusionofthenationallevelinordertoexaminetherevenue
effectsofdecentralisationwasofhighrelevanceandhadbeenneglectedsofar.JoulfaianandMarlow(1991)
arguethatwhilecentralgovernmentmaynotbesubjecttoTieboutcompetitionandsomeofitspublicgoods
maynotbepossibletoprovideatthelocallevel,itisjustthatmonopolisticpowerthatislikelytobequite
relev69antintestingtheLeviathanhypothesis.
Seealsosection2.2.2.

35

TillmannLisa

FederalTaxCompetition

adatasetfrom1987through2000.Thegeneratedresultsimplythatprovincialexpenditures
willriseasaconsequenceofstrongerfiscaldecentralisation,whichisperceivedasaconfir-
mationoftheOates-Wallishypothesis.70Chen(2004)alsofindsevidencethatprovincialand
centralgovernmentsaremorelikelytoengageincollusionactivitieswithahigherdegreeof
decentralisation,whichinturnisseenassupportiveofthecollusionhypothesis.
SeveralstudiesexistthatexaminepaneldataonSwisscantons.Forthetimebetween1980
and1998,FeldandMatsusaka(2003)findthatdirectdemocracyintheformofreferendums
significantlyreducestheamountofgovernmentexpenditures,aresultwhichmaybeseen
asaconfirmationoftheLeviathanhypothesis.Forselectedyearsbetween1985and2001,
BruelhartandJametti(2007)findthatahigherdegreeofdecentralisationwithinacanton
entailsareductionofmunicipaltaxrates,concludingthatthismaybeindicativeoftax
competitionhavingarestrainingeffectonLeviathangovernments.FeldandSchaltegger
(2009)examinetheimpactofgovernmentfragmentationonrevenueandexpendituresize
between1980and1998.Theyfindthatcabinetsizeispositivelyrelatedtogovernment
revenuesandspending,whichmaybeseenassupportiveoftheLeviathanhypothesis.They
furtherfindthatdirectdemocracyisameanstoreducethesizeofgovernment.Forthesame
dataset,Feldetal.(2010)find,amongotherthings,thatgovernmentbudgetwillshrinkwith
moreintenseinterregionaltaxcompetition.Thedecentralisationofrevenuepowerleadsto
areductionoftaxrevenues,whileuserchargeswillincrease.

triesCounAcrosstralisationDecenAmongthestudiesfocusingoninternationaldatasets,AndersonandvandenBerg(1998)
examineaworldwidesampleof45countries,butareunabletoestablishresultsinfavour
oftheLeviathanhypothesis.Forpaneldataon32developingandindustrialcountriesfrom
1980-1994,JinandZou(2002)examinehowdecentralisationaffectssubnational,national
andconsolidatedgovernmentsectorsizeandfindevidenceinfavouroftheLeviathanhy-
pothesis,whilealsoaccountingfortheOates-Wallishypothesis.Theirresultspointtowards
expendituredecentralisationculminatinginareductionofcentralgovernmentsizeandan
increaseofsubnationalandaggregatebudgets.Greatersubnationalbutreducednationaland
compoundbudgetsarefoundtoresultfromrevenuedecentralisation.Finally,theexistence
ofintergovernmentalgrantschemesraisesgovernmentbudgetsatnationalandsubnational
level.Rodden(2003)alsostressestheneedtospecificallyaccountforwhetherexpenditure
decentralisationisaccompaniedbytaxrevenueautonomy.Foraninternationalpaneldataset
from1978through1999on29countries,thecombineddecentralisationofexpendituresand
revenuespointstowardsareductionoftotalgovernmentsize,whiletheisolateddecentrali-
sationofexpenditureswillleadtoitsexpansion.For16OECDcountries,Santolini(2009)
70(1985).atesOSee

36

TillmannLisa

FederalTaxCompetition

examinesbalancedcross-sectionaltimeseriesdatabetween1978and1997testingforthe
effectofdecentralisationandofthelegislature’smonopolypowerongovernmentsize.The
resultsimplythatfiscaldecentralisationwillconstrainLeviathans,whilethesameholds
forhighlegislativemonopolypower,whichrunscountertotheBrennanandBuchananhy-
pothesis.Shefurthertestsforthejointeffectsofdecentralisationandlegislativecontrolto
concludethatLeviathancanbetamedbyausterelegislativecontrol,orweakcontrolsofthe
legislatureincombinationwithahighdegreeofsub-nationaltaxationautonomy.Martinez-
VazquezandYao(2009)scrutinisetherelationshipbetweenpublicsectoremploymentand
decentralisationbetween1985and2005forover100countries.Theyfindthatahigherdegree
ofdecentralisationentailshigherlocallevelpublicsectoremployment,whichmorethanoff-
setsthedownsizingofcentrallevelpublicsectoremployment.Todetecthowdecentralisation
affectsgovernmentbudgetsatthenational,thesubnationalandtheaggregatelevel,Cassette
andPaty(2010)examinepaneldataontheEU-15.Theyfindthattheresultingincrease
insubnationalexpendituresmorethanoffsetsthereductionincentrallevelexpenditures,
suchthataggregatebudgetswillbehigherwithastrongerdegreeofdecentralisation.The
existenceofintergovernmentaltransfersisfoundtoraisethesizeofgovernmentatalllevels.

Summary2.3

Federaltaxationischaracterisedbytheco-existenceofverticalandhorizontalexternalities
triggeredbyonejurisdictionneglectingtheimpactitsownchoiceoftaxratehasonother
jurisdictionsatthesameoranotherlevel.Earlytheoreticalapproachessawfederations
asanumberofindependentregionswithapassivefederalgovernment.Onlyhorizontal
externalitiesthatdistortregionaltaxpolicieswillthenoccur.71Insum,theypointtowards
atendencyofinefficientlylowtaxationandpublicgoodprovision.Thatdownwardpressure
maybeattenuatedifregionsarenotassumedtobesymmetric,forexample,iftheydifferin
size,ifoneregioncantakeadvantageofagglomerationrentsorisabletodifferentiateitself
fromcompetitorsviatheprovidedpublicgood.
Ifacentralgovernmentactsasanactiveplayer,additionalverticaltaxexternalitiesoccur.
Whileforeachregionthetaxbaseiselastic,theelasticityofthefederaltaxbasedepends,
forinstance,onthedegreeofthecountry’sopennesstointernationalcapitalmarketsorthe
elasticityofcapitalsupplywithrespecttothenetinterestrateprevailinginthefederation.
Bothlevelsofgovernmentmayneglecttheimpacttheirdecisionshaveonbudgetsatthe
otherlevel,whichmaybeaconsequenceofmyopicbehaviourordifferingobjectivefunctions.
71Thesecanbedividedintoindirectpositiveexternalitiesresultingfromtheinterregionalreallocationof
mobilecapitaltaxbaseanddirectnegativeexternalitiesthatmayresultfromachangeinthefederalnet
capital.onreturn

37

TillmannLisa

FederalTaxCompetition

These(bottom-uportop-down)verticalexternalitiespointtowardstaxratesbeingsethigher
thaninthecentrallyplannedsecond-bestoptimum.
Theverticalandhorizontalexternalitiesresultingfromfederalstructurethuspointin
oppositedirections.Therelativestrengthofverticalandhorizontalexternalitiesmayhingeon
variousfactors:ahighinterestrateelasticityofcapitalsupply,alowinterestrateelasticityof
capitaldemandandastrongdegreeofinternationalcapitalmarketintegrationpointtowards
adominanceofverticaloverhorizontalexternalities.
Toevaluatetheefficiencyofequilibria,thedecision-makers’intentionsareofconcern:For
benevolentgovernments,inefficientlevelsoftaxationorpublicgoodprovisionimplyalossof
welfare.ForLeviathangovernments,dominatingverticalexternalities,suchthatconsolidated
taxratesaretoohigh,pointtowardstherelevanceoftheLafferparadoxon.Inefficientlylow
taxratesmaybeperceivedaswelfareenhancingbyrestrainingtheLeviathan’srevenue-
olicies.pmaximisationInternationalcapitaltaxcompetitionaddsanotherlayerofexternalities:Positivecross-
nationalhorizontal(andvertical)externalitiesbetweenthefederationsmayattenuatethe
verticalexternalitieswithinafederationbypointingtowardsinefficientlylowtaxratesfrom
aglobalperspective.Byexertingadditionaldownwardpressureontaxrates,international
taxcompetitionimpliesfortheverticalexternalitytobelessrelevantfortheoveralllevelof
taxrates.Or,putdifferently,federalstructuredrivingconsolidatedtaxratesupwardswithin
thefederationmayfunctionbyoffsettingthedownwardpressureontaxratesarisingfrom
internationaltaxcompetition.
Themajorityofempiricalstudiesseemstobeaffirmativeofthehypothesisofinterre-
gionalacross-andwithin-countrytaxcompetitionimplyingacomplementaryrelationship
betweentaxrates.Particularlythemorerecentstudiesconveyamoredifferentiatedpicture
accountingforthesignificanceofregionalismandagglomerationeffectswithinacountryas
wellasindividualregionalcharacteristicsthatshapethedetectedstrategicinteraction.The
evidenceisfurthersupportiveoftheEUintegrationandenlargementprocesshaving(had)
animpactontaxratechoices,whiletherespectiveresultsastohowexactlythatstrategicin-
teractioncanbecharacteriseddifferacrossstudies.Inanycase,theresultsareaffirmativeof
thehypothesisthatincreasedinternationalcapitalmarketintegrationintensifiesinterregional
tition.eompctaxByincludingverticaltaxinteractions,theevidencebecomesrathermixed,withaslight
tendencyofstudiessupportingacomplementaryrelationshipbetweenupper-andlower-level
taxrates.Whatismore,givensomeresultsonCanadiandata,72itseemslikelythataregion
maytaketheroleofaStackelbergleaderandthatfederalaswellassame-leveljurisdictions’
taxratesreacttosuchleadingregions.Withrespecttotheabilityofthefederalpolicymakers
72SeeHayashiandBoadway(2001)andCrisan(2007).

38

LisaTillmann

FederalTaxCompetition

tointernalisetheinefficienciescreatedattheregionallevel,theevidenceismixedand,ifat
all,notintheaffirmative.
TheevidenceonLeviathangovernmentbehaviourisfarfromclearanditsexistencecan
neitherbedeniednorconfirmed.WhileitisstraightforwardtotestfortheLeviathanhy-
pothesisinthecontextofdecentralisation,thetestablehypothesesaswellastheresulting
implicationsaredifficulttointerpret.Theexistingstudiesmaybeseenasindicativeof
efficiency-raisingor-deterioratingeffectsofdecentralisation.Itismorelikelytoraiseeffi-
ciencyifreformsincluderevenueinsteadofsolelyexpendituredecentralisation.Whetherthis
maybeattributedtoLeviathanbehaviourisnotnecessarilyclear.Moreworkthusneedsto
bedoneinordertoclarifythespecificchannelsinwhichthepresenceofLeviathanbecomes
manifest.Inanycase,itisfairtosaythattheassumptionofrevenue-maximisingbehaviour
withinamodelmaybejustifiedtoacertainextentandtheevaluationofpoliciesneedsto
becautiousastowhatbehaviouroneismoreinclinedtoassume.
Thefearofthenegativeimpactofglobalisationonwelfareandtaxrevenuessubstantially
shapesfiscalpolicies,asoutlinedinsection1.1.Indeed,someoftheearlytheoreticalcontri-
butionsoninternationaltaxcompetitionsupportthisviewofglobalisedcapitalmarkets.The
standardtheorydevelopedbyZodrowandMieszkowski(1986)andWilson(1986)seesthat
policymakershaveincentivestoreducetaxratesoncapitalasthelatterbecomesincreasingly
mobile.AzeroequilibriumtaxrateonmobilecapitalisevenfoundtoresultbyBondand
Samuelson(1989)orRazinandSadka(1991).Notwithstandingthattheyhavebeenreshaped
andspecifiedinthesucceedingliterature,73theseresultsappeartoremaininfluentialinthe
collectiveconscience.Theempiricalstudiesontaxrevenuelevelscitedinsection1.1further
pointtowardsamoredifferentiatedviewoftheconsequencesofglobalisation.Theexplicit
considerationoffederaldesign(withfederaltaxsettingpossiblybeingdriveninadirection
oppositetothatpredictedbythestandardinterregionaltaxcompetitionliterature)aswell
astherecognitionofthecontrastingevaluationofresultswhenpolicymakersareassumed
tobeeitherrevenueorwelfaremaximisersinthecontextofinternationaltaxcompetition
furtherenhancethediscussion.Whatismore,theoreticalresultspointtowardsadifferen-
tiatedviewofthechallengesofglobalisationinsuchthatcountriesdoinfacthavemeans
todistinguishthemselvesfromtheircompetitorsinvariousways,therebycircumventingthe
downwardpressureontaxrates.Theheterogeneityofjurisdictionsaswellasofcitizens’and
firms’preferencesentailssuchpossibilities,aswasalreadypointedoutbySalmon(2008).
Regardingtheworldwidetrendtowardsdecentralisation,policyrecommendationsneed
tobearinmindthattheempiricalevidencesuggestsfortheefficiency-enhancingeffectsof
decentralisationtobecomeobsoleteifthelatterisrestrictedtoexpenditureratherthanrev-
732.1.1.sectionSee

39

TillmannLisa

FederalTaxCompetition

enueautonomy.74Aglobaltrendtowardsfiscaldecentralisationmaynonethelessculminatein
growingnumbersofcountriesactuallyadaptingfederaltaxsystems,suchthattherecognition
oftheconsequencesoftaxbaseoverlapbecomesevenmorerelevant.
Withtheargumentsbroughtforwardsofar,fortheevaluationofpossibletaxpolicies,
somekeyquestionsneedtobeanswered.Forone,amoredifferentiatedviewofhorizontal
taxcompetitionseemstobeinplace.Beyondthat,thefederalframeworkoftherespective
countriesishighlyrelevant.Itisfurthernecessarytobeawareofwhetheracountry’spolicies
aremorelikelytohavebeenshapedbyrevenue-maximisingratherthanwelfare-maximising
intentionsinthesettingoftaxrates.Theexplicitrecognitionofthecompetingcountries,
particularlytheirfederaldesign,isalsorelevant.WithrespecttotheEuropeanUnion,
policymakersneedtobeawareofthedifferencebetweentheperspectiveoftheEuropean
Unionasaneconomicentityincompetitionwithotherstates,incontrasttotheperspectiveof
asinglemembercountry.WithanintegratedEuropeactuallybecomingmoredeeplyrooted
inthedecision-makingprocessofeachmemberstate,thislastaspectwillbeincreasingly
important.Finally,worldwidedecentralisationtrendsneedtobeevaluatedcautiouslywith
respecttoexpenditureasopposedtorevenuedecentralisation.
Asdescribedinsection1.1,theworldwidetrendtowardsfiscaldecentralisationcombined
withinternationalcapitalmarketintegrationmakestherecognitionofacountry’sfederal
frameworkessentialforeconomicpolicydesignandevaluation.Inthatrespect,researchcan
stillbeextended.Sofar,thetheoreticalmodelsexaminingthedegreeofinternationalcapital
taxcompetitionanditsimpactontaxsettingwithinafederation,havedonesoeitherby
assumingthatthecountryfindsitselfinoneoftheboundarycasesofperfectintegration
oraclosedeconomyorbyvaryingthatdegreeindirectly,forexampleoverthenumberof
competingstates.Noapproachexistssofarthatexplicitlyconsidersthesupplyofcapital
withregardtoacountry’sdegreeofinternationalcapitalmarketintegration.Furthermore,
notmuchworkhasbeendonewithrespecttojusthowthefederalstructureofacountry
influencesitspositionintheinternationalcompetitionforcapitaltaxbase.Afewstudies
inthatrespectexist,75yetthereisnosystematicdiscussionofthechangeinoutcomeswith
twocountriescompetingformobilecapitalwithadifferingfederalsetupwithinonedistinct
setting.Thefollowingtwochaptersaimtofillsomeofthesegapsconcerningacountry’sfederal
frameworkanditsdegreeofopennesswithregardtocapitalmobilityandtherebyfurther
validatethekeyquestionsoutlinedinthepreviousparagraph.

7475SeeSeesectionsection2.1.3.2.2.4.

40

Chapter3

FederalTaxCompetitionandCapital
tegrationInetMark

Thischapteraimstoexplicitlyincorporateafederation’sdegreeofintegrationintointerna-
tionalcapitalmarketstotheconsiderationofefficiencyoftaxrates.
Thejointeffectsofverticalandhorizontaltaxexternalitiesinthecontextofcapitalincome
taxationhavebeenstudiedbyKeenandKotsogiannis(2002).1Intheirmodel,thesupply
ofcapitalisdeterminedbylife-cyclesavings.Itisdemonstratedthatwhetherthehorizontal
taxexternalitydominatestheverticaloneorviceversaessentiallyhingesontheelasticityof
savingswithrespecttotheinterestrate.Ifsavingsareinelastic,thehorizontalexternality
dominatestheverticaloneandtaxratesoflowerleveljurisdictionsaretoolow.Forelastic
savings,theoppositeholdstrue.
TheapproachtakeninthefollowingchapteriscomplementarytothatofKeenandKotso-
giannis.Ratherthanconsideringlife-cyclesavings,itrelatesthesupplyofcapitaltothelevel
ofinternationalcapitalmarketintegration.Thekeyelementoftheframeworkestablishedin
thispaperisaconceptthatallowstoparameterisethedegreeofinternationalcapitalmarket
integration,whichisthenincorporatedintoanefficiency-orientedanalysisofcapitaltaxation
inafederation.Itisshownthatifthedegreeislow,thatis,iftheeconomyhasonlylimited
accesstotheinternationalcapitalmarket,thehorizontalexternalitywilldominatetheverti-
caloneatthestatelevel,whereastheoppositewillholdtrueifthedegreeislarge.Thisresult
impliesthatinafederalsystemwithlimitedaccesstotheinternationalcapitalmarketthere
willbeundertaxationofcapitalincome,whereasinafederalsystemwithextensiveaccess
therewillbeovertaxationbylowerleveljurisdictions.Itisfurthershownthattheefficiency
offederaltaxationessentiallyhingesontheassumptionsaboutthestrategicbehaviourofthe
federalgovernmentandontherelativebehaviourofstateandfederaltaxes.Thegenerated
resultsestablishalinkwiththeresultsbyKeenandKotsogiannis(2002)insuchthatthe
12.1.2.sectionSee

41

TillmannLisa

FederalTaxCompetition

degreeofcapitalmarketintegrationworksthesamewayastheinterestelasticityofsavings.
Thatis,ahighdegreeofinternationalcapitalmarketintegrationfunctionsbyreducingthe
relativeimportanceofthehorizontalexternalityasopposedtotheverticalexternality,given
thatcapitaltaxbasecanbeshiftedmoreflexiblyabroad.Itisfurthershownthatforaspe-
cificdegreeofinternationalcapitalmarketintegration,verticalandhorizontalexternalities
mayoffseteachother.
Thechapterissetupasfollows:Thenextsectionestablishestheanalyticalframework
andsketchestheinternationalinterdependenciesofcapitalmovementsandcapitaltaxation.
Section3analysesthepossibleverticalandhorizontalexternalities.Attentionisfirstpaid
totheefficiencyofstateleveltaxation,followedbyanefficiency-orientedanalysisoffederal
leveltaxation.Withthehelpofanumericalexample,section4willillustratethetheoretical
resultsgeneratedintheprecedingsection.Section5concludes.

3.1TheModel
Theanalyticalframeworkisoutlinedintheproceedingsection.Theexaminedfederationis
asmallcountry,whosepublicsectorconsistsofafederalgovernmentandnidenticalstates.

Households3.1.1Inordertofactoroutequityconsiderations,leteachstatei=1,...,nbepopulatedbya
singlerepresentativeconsumerendowedwithoneunitofcapital.Utilityoftheconsumerin
stateiisdeterminedby

u(ci,gi,G)=ci+Γ(gi,G),

(3.1)

whereciisprivateconsumptionofconsumeri,giistheamountofalocalpublicgoodprovided
instateiandGisfederalspendingperstate.Γ(gi,G)isafelicityfunctionmeasuring
theutilityfromlocalpublicgoodprovisionandfederalspending,whichsatisfiestheusual
monotonicityandconcavityassumptions.Anindividual’sprivateconsumptionisconstrained
to

ci=ρ+(1−θ)πi,

(3.2)

inwhichρdenotesthenetreturnoncapital(whichwillbeseentobeequalisedacross
states).Eachstateraisesan(exogenous)taxrateθ∈[0,1]oneachunitofrentalincomeπi
anindividuallivinginigeneratesfromanimmobilefactor.

42

LisaTillmannFederalTaxCompetition
ductionPro3.1.2Aconstantreturnstoscaleproductionfunctionisconsidered,whichisidenticalinallstates
andatleasttwicecontinuouslydifferentiable.Itdisplaysdiminishingreturnstoscaleinits
mobileinputfactor,capital.Productioninstatei,denotedasyi,isthendeterminedby
yi=f(ki)=f(ki)ki+πi,(3.3)
wherekiistheamountofcapitalemployedinstatei.Assumingthatfirmsareprofitmax-
imisers,f(ki)kiisthentheincomeofcapitalinvestedinstatei.
yobilitMCapital3.1.3Consumerscanchoosetoinvesttheircapitaleitherinthestateinwhichtheylive,insome
otherstateofthefederation,ortheycaninvestintheinternationalcapitalmarket.While
capitalmobilitywithinthefederationcomesatnocost,itisassumedthatinternational
capitalmobilityiscostly.Moreprecisely,thecorrespondingcostfunctionisdefinedas
2
Cσ,n−ki=21σn−ki,withσ∈[0,∞),(3.4)
iiwhereikidenotesthetotalcapitalemployedinthefederation.Thecapitalavailablein
thefederationisgivenbyn.Ifnfallsshortofthetotalamountofcapitalemployedinthe
federation,thecountryisanetimporterofcapital.
σmeasuresthedegreeofinternationalcapitalmobility.Itmaybeinterpretedasarisk-
premiumdemandedforinternationalinvestmentprojects.2Itisassumedthatσisthesame
acrosscountries.3Forσ=0,thecountryisasmalleconomyperfectlyintegratedintothe
internationalcapitalmarket.Forσ→∞,theeconomyisclosedandhasnoaccesstothe
internationalcapitalmarket.
ThecostofinternationalcapitalmobilityisdepictedbyFigure3.1.Thehigherthedegree
ofcapitalmarketintegration,thelowerisacountry’scostofinvestingabroad.Thecostof
internationalcapitalmobilityincreasesprogressivelywiththedifferencebetweenthetotal
capitalavailableandthetotalcapitalemployedinthefederation,whichcanbegrasped
intuitivelybytheassumptionthattheinherentriskofinvestingabroadwillbethehigher,
themorecapitalisalreadyemployedinacountry.Assumethatthemostattractive,orsafest,
investmentopportunitieswillbeseizedfirst.Thus,withamountingintensityofinvestment
activity,suppliersofcapitalmaybeobligedtoresorttoother,possiblymorerisky,projects.
23FoNoterthethatusetohisfassimilarymmetrytaerminssumptiontheconmatyextboefdfropporeignedwithoutindebtedness,furtherseecvhandangingerPlotheega(1996).nalyticalresults,
providedthatthecharacteristiccostfunctionisdifferentiableatallpoints.Theanalysiswould,however,be
complicated.unnecessarilyrendered43

LisaTillmannFederalTaxCompetition
Figure3.1:CostofInternationalCapitalMobility
Thepossibilityofinterstateandinternationalinvestmentgivesrisetotwonon-arbitrage
conditions.Theinterstatenon-arbitrageconditionisgivenby
ρ=f(ki)−τi∀i,(3.5)
inwhichf(ki)isthemarginalreturnoncapitalandτi=ti+Tistheaggregatetaxburden
imposedontheresidentsofonestatebythelocal(ti)andthefederal(T)authorities.It
canbeinferredfrom(3.5)thatthenetreturnoncapitalwillbethesameforinterstate
investments,irrespectiveofwhatstateanindividualdecidestoinvestin.Givendecreasing
returnstoscale,rent-seekinginvestorswillarbitrageawaydifferencesinnetreturnoncapital
betweenstates,thatis,capitalwillmovebetweenjurisdictionsuntilρisequalacrossall
states.Thecostperadditionalunitofcapitalinvestedoutsidethefederationamountstoσ(n−
iki).Theinternationalnon-arbitrageconditioncanthenbedenotedas
nρ=r+σ1−iki∀i,(3.6)
whereristheworldmarketinterestrateandinkiistheratiobetweendomesticcapital
endowmentandcapitalemployment.Iftheamountofcapitalemployedinthefederationis
exceededbyitscapitalendowment(inki>1),thatis,ifthecountryisacapitalexporter,the
netreturnoncapitalprevailinginthefederationwillfallshortoftheworldmarketinterest
rate.Itwillonlybebeneficialtoinvestabroadaslongastheworldmarketinterestnetthe
costofinternationalinvestmentexceedsthenetreturnoncapitalgeneratedfromdomestic
investment.4Inequilibrium,
nf(ki)−τi=r+σ(1−iki)∀i.
4Itisavalidobjectiontowonderwhyinasmallcountryinwhichtheworldmarketinterestisexogenous,
44

LisaTillmann

FederalTaxCompetition

Hence,allpossibilitiesofarbitragewillbeextinctandaninvestorwillbeindifferentbetween
investingabroadorathome.
Twobenchmarkcasesofinternationalcapitalmobilityshallnowbeestablished,which
willbeusefulthroughoutthefurtheranalysis.
Lemma1Letτibefiniteforalli=1,...,n.Providedthatf(ki)isabijectivefunctionof
hen:t,kii.ki=(f)−1(r+τi),ifσ=0
ii.ki=n,ifσ→∞
i:fooPri.Obviousfrom(3.5)and(3.6).
ii.Theproofisbycontradiction.Tostartwith,assumethatiki≥constant>nif
σ→∞.Then,asτiisfiniteandasf(ki)−τi=r+σ(1−inki)foralli,itfollows
thatf(ki)→∞foralli.Asf<0,thisnecessarilyrequireski→0foralli,which
contradictsiki≥constant>nforalli.Nowassumethatiki≤constant<nif
σ→∞.Then,asτiisfinite,itfollowsthatf(ki)→−∞,whichcontradictsf>0.
Q.E.D.

Figure3.2summarisestheabovediscussion.Theinternationalnetreturnoncapital
isdepictedasalinearfunctionincreasinginiki.Interstatenetreturnoncapitalisa
decreasingfunctionofki.Arbitrage-seekingindividualswillnowhaveanincentivetoinvest
insidethefederation,aslongastheinterstatenetreturnoncapitalexceedstheinternational
one,andviceversa.Hence,foragivenworldmarketinterestrateandaspecifictaxburden
ineverystateofthefederation,acertaintotalamountofcapital(forexample(iki)0<n,
implyingthatthecountryisanetexporterofcapital)willbeemployedinsidethefederation.
Consequently,n−(iki)0ofdomesticcapitalwillbeemployedabroad,whichwillredound
toaspecificequilibriumnetreturnoncapitalρ0.
Now,ataxhikewilltriggeradropintheequilibriumnetreturnoncapital,yetthat
decreaseinρwillbedisproportionatelysmall.Assumefirst,forsimplicity,thatthefederal
theindividualinvestor’schoicewouldinfluencethenetreturnoncapital.Yetthelattercanbesolvedby
arguingthateachindividual,whenfacingthedecisionwheretolocatecapital,willchoosethemostattractive
opportunityfirst.Foreachadditionalunitofcapitalprovided,aless‘attractive’investmentopportunitywill
betakenup,suchthattheimplicitcostforeachinvestorwillinfactvarywitheachindividual’sinvestment
decision.

45

TillmannLisa

3.2:FigureEquilibriumNon-Arbitrage

FederalTaxCompetition

taxisraised,which[via(3.5)]willleadtoasimultaneousdecreaseinthenetreturnon
capitalineverystateofthatfederation.Asaconsequence,investorsinthefederationwill
haveanincentivetoremovetheircapitalandinvestabroadinstead.Althoughthatrise
ininternationalinvestmentactivityhasnoimpactonr,foreachadditionalunitofcapital
locatedoutsidethefederation,theincurredcostwillincrease(suchthatρfalls).Inthesame
time,marginalproductivityofcapitalinsidethefederationwillagainapproachtheinitial
levelanddomesticinvestmentwillagainpickup.Inconclusion,thearbitrageopportunities
arisingduetothefederaltaxhikewillleadtoanewequilibriuminwhichalargershare
oftotaldomesticcapitalisemployedinternationallyandthenetreturnoncapitalsettles
atalowerlevelthanbefore.Notethatifasinglestatedecidestoraiseitstax,notonly
willinternationalinvestmentbecomemoreattractive,investorswillalsohaveanincentiveto
relocatetheircapitalintotheotherstatesofthefederation.Inbothcases,thedropinnet
returnoncapitalwillbedisproportionatelysmallincomparisontothechangeintaxburden.
Figure3.3displaysthebenchmarkcasesofσ=0andσ→∞graphically.Inthefirstcase,
anindividualwillincurnocostofrelocatinginvestmentsbetweencountries.Theamount
employedinsidethefederationwillthenbesuchthattheequilibriumnetreturnoncapital
equalstheworldmarketinterest.Inthelattercase,alldomesticcapitalisemployedinside
thefederationandnocapitalwillenterthecountry,suchthatiki=n,irrespectiveof

46

TillmannLisa

FederalTaxCompetition

Figure3.3:Non-ArbitrageEquilibrium-BenchmarkCases

nationaltaxpoliciesorcapitalproductivity.

mplicationsIolicyP3.1.4Asoutlinedintheprevioussection,thenetreturnoncapitalisinfluencedbythetaxregime
inthefederation.Furthermore,therentalincomeanindividualgeneratesalsodependson
theaggregatetaxburdenineachstate.
Equation(3.5)implicitlydefineskiasafunctionki=ki(ρ+τi),withki(ρ+τi)=fi1,5in
whichfi:=f(ki).Then,equation(3.6)implicitlydefinesρasafunctionρ=ρ(τ1,...,τn)
and(applyingtheimplicitfunctiontheorem)
1nσ∂ρ=(iki)2fi,
1n∂τi1−σ(iki)2ifi
andforτ1=τ2=...=τn=τ
σ∂ρ∂τ=nk2f−nσ≤0,(3.7)
5Providedagainthatf(ki)beabijectivefunctionofki.

47

(3.7)

TillmannLisa

FederalTaxCompetition

with∂τ∂ρ=0,asσ=0,and∂τ∂ρ→−n1,asσ→∞.
Leteachindividualreceivearentalincomeπifromthepossessionofafixedfactordefined
byπi=f(ki)−f(ki)ki,
whichthendenotestheamountbywhichthecapitalproductexceedsitsremuneration.In
it,ki=ki(ρ+τi),whichimpliesthatπi=πi(ρ+τi),withπi=−ki.Thus,
∂ρ∂πi∂τi=−ki(∂τi+1).
Hence,thechangeinrentalincomeduetoashiftinτiisgivenbytheeffectofthatshift
onkiandbytheimpactofthetax-inducedchangeinρonki.

3.1.5GovernmentObjectives
Atwo-levelledfederationisconsidered,inwhichthefederalgovernmentiscomplementedby
state-levelgovernments.Bothareassumedtobebenevolent.
Thenstatescompeteformobiletaxbase,thereinengaginginNashcompetition.The
benevolentstatetriestomaximisethewelfareofitsconstituents,irrespectiveofthepossible
impactitsactionsmayhaveoncitizensofotherstates.Consequently,eachstateimaximises
ui=ρ+(1−θ)πi+Γ(gi,G),
tojectsubgi=tiki+θπi
andG=1Tki
nibychoiceoftiandtheresultinglevelofgi.Eachstategovernmentderivesitsrevenuefromthe
taxationofcapital(ki)andrents(πi).Itredistributesthatrevenueamongitsconstituents
bymeansofthesupplyofthestatepublicgood.
Thebenevolentandomniscientfederalgovernmentmaximiseswelfareofallconsumersin
allstatesi=1,...,nacrossitsjurisdictionsubjecttotherespectivebudgetconstraintsby
choiceofGandthecorrespondinglevelofT.Itderivesrevenuefromthetaxationofthetotal
amountofcapitalemployedinthefederationandredistributestaxrevenuesevenlyacrossall
states.

48

TillmannLisa

FederalTaxCompetition

3.2HorizontalandVerticalExternalities

Themodelgivesrisetoverticalaswellashorizontalexternalities.Witheachlocalgovernment
neglectingtheimpactitstaxdecisionshaveonthebudgetoftheotherlocalentities,ittriggers
ahorizontalexternality.Averticalexternalityarisesfromperfecttaxbaseoverlap.Withan
omniscientfederalgovernment,onlyabottom-upverticalexternalitywilloccur,giventhat
eachstateunderestimatesthelossoftaxrevenuethefederalgovernmentincursduetoarise
intaxationatthestatelevel.
Theanalysisiscarriedoutasfollows:First,thestatelevelisconsidered.Itisexamined
whetherundertheoutlinedstructure,stateswillexcessivelytaxcapital.Proceedingtothe
federallevel,thesamequestionisasked,allowingforthefederalgovernmenttoeitherengage
inNashcompetitionoractasaStackelbergleader.Throughouttheanalysis,referenceis
madetothebenchmarkcasesoutlinedinLemma1.

3.2.1AreStateTaxesTooHighorTooLow?
Eachlowerlevelgovernmentintendstomaximiseitsinhabitant’sutilitybychoiceofits
taxrate.AssumingthatstatesengageinNashcompetition,theresultingequilibriumis
characterisedbyeachstateisettingtheoptimaltaxgiventhedecisionofallotherstates.
Equilibriumisthendefinedbysymmetricstrategies,sothatfromthemaximisationproblem
asoutlinedinsection2.5.,eachstate’sfirst-orderconditionisgivenby:

(3.8)

∂u=∂ρ+1−θπ∂ρ+1
∂t∂τ∂τ
+Γgk+tk+θπ∂ρ+1(3.8)
∂τ+ΓG1Tkn∂ρ+1=0.
∂τnFrom(3.8),itcanbeinferredthatthechangeinutilityofarepresentativeconsumerdue
toariseincapitaltaxationinonestatedependsontheresponsivenessofnetreturnoncapital
toashiftincapitaltaxation(∂τ∂ρ).Asshownin(3.7),thelatterisnegative(orzero)andwill
betheclosertozero,themoreintegratedintotheinternationalcapitalmarketacountryis.
Astategovernmentaccountingforthiseffectforitsconstituentstendstoneglectthatthe
citizensofeveryotherjurisdictionwillbesubjecttothesamedropinρ,whichpointstostate
taxesbeingtoohigh.Ontheotherhand,capitalwillberelocatedforagiventaxhike,such
thattheotherjurisdictions’taxbasewillrise,whicheachregionaldecision-makerdoesnot
accountfor.Thatpositivehorizontalexternalitypointstowardsstatetaxesbeingtoolow.
Equation(3.8)isfurtherinfluencedbythedropinafter-taxrentalincomeofarepresentative

49

TillmannLisa

FederalTaxCompetition

consumerduetoacapitaltaxincrease[(1−θ)π(∂τ∂ρ+1)],whichisinturnafunctionofthe
amountofcapitalemployedinthestate.
ΓGandΓgdenotethemarginalvaluationoffederalandstatespendingrespectively.Di-
minishingmarginalutilityleadstothembeingthesmaller,thehighertheprevailinglevel
oftherespectivepublicgoodsupply.Inthecaseofstatepublicgoodprovision(depicted
bythefirstterminsquaredbrackets),thestategovernmentbearsinmindtheincreasein
utilityasaconsequenceofahigherlevelofpublicgoodprovisionduetotheadditionalrev-
enuesfromthehighertaxrateasopposedtothelossoftaxbaseasaconsequenceofthe
arbitrage-seekingbehaviourofindividuals.Thatis,itaccountsfortheresultingtaxincome
asopposedtothetaxbaseeffectuponevaluatingthechangeinutilityderivedfromregional
publicgoodprovision.Thetaxbasereductionwillbethegreater,thehighertheriseinstate
tax,whileitisinthesametime(lessthanfully)offsetbytheresultingoveralldropinnet
returnoncapital.Theutilityderivedfromfederalpublicgoodsupplyisaffectedinsuch
thatthefederaltaxbase,whichisequivalenttothesumofallstates’capitaltaxbase,will
diminishduetothestatetaxhikebycapitalnotonlyshiftingwithinthefederation,butalso
relocatingabroad.Thefederaltaxbasewillthusdecrease,which,inturn,reducesfederal
expenditures.Eachstateonlyaccountsforthenegativeimpactoftheresultingreductionin
federalexpenditureswithregardtothewelfareofitsowncitizens,whichdepictsthevertical
externalitytriggeredbystatepoliciespointingtowardsinefficientlyhighregionaltaxrates.
Sincethefederalgovernmentequallydividesitsexpendituresbetweenthestates,onlyn1th
ofthefederaltaxbasereductionisincorporatedbyeachstate.
Thatis,anincreaseincapitaltaxationinoneofthestateswilltriggerseveraleffects.
Thestategovernmentanticipatestheimpactthehighertaxburdenhasonitscitizens’rental
incomeandutilityderivedfromstatepublicgoodsupply.Meanwhile,itneglectsthepotential
effectsinotherstates,therebytriggeringahorizontalexternality,whichpointstoinefficiently
lowtaxation.Inthesametime,theoptimisingstategovernmentdoesnotfullyaccountfor
thedropinfederalpublicgoodsupplycausedbythedecreaseinfederaltaxbase,suchthat
taxationtendstobeinefficientlyhigh,asdepictedbythesecondtermin(3.8).Tounderstand
whetherNashcompetitionamongthestatesleadstoinefficientequilibriumtaxation,itis
consideredhowacoordinatedtaxhikebyallstatesaffectswelfarewithinthefederation.
StartingfromthedescribedsymmetricNashequilibrium,taxesaresimultaneouslyincreased
inallstatestohave

50

TillmannLisa

FederalTaxCompetition

(3.9)

∂u∂ρ∂ρ
n∂t=n∂τ+1−θπn∂τ+1
+Γgk+tk+θπn∂ρ+1(3.9)
∂τ+ΓGTkn∂ρ+1.
∂τThesignof(3.9)isnowofinterest.Ifitiszero,itindicatesthatstatetaxationisinfact
efficient.Ifitispositive,thehorizontalexternalityislikelytobedominant,whereas(3.9)
beinglessthanzeropointstowardsadominantverticalexternality.Subtract(3.8)from(3.9)
evhato

horizontalexternality
du=∂u−∂u=1+1−θπ+Γgtk+θπn−1∂ρ
dtn∂t∂t∂τ(3.10)
+ΓGTkn−1∂ρ+1.
verticalexternalit∂τyn
Withthehorizontalexternalitypointingtowardsstatetaxesbeingtoolowandthevertical
externalitypointingtowardsstatetaxesbeingtoohigh,thesignof(3.10)isambiguous.In
ordertogainsomefurtherinsight,thetwobenchmarkcasesestablishedinLemma1will
againbereferredto.Assumefirstforσtobezero.Asaconsequence,(3.7)willbezero,so
ecomesb(3.10)thatdu=ΓGTkn−1<0,
ndtimplyingthatforagiventaxhike,theutilityoftherepresentativeconsumerwillfall.Hence,
inthecaseofperfectintegrationintotheinternationalcapitalmarket,theverticalexternality
dominatesthehorizontalexternality.
Thisresultcanbeeasilygraspedintuitively:Ifcapitalisperfectlymobile,non-arbitrage
leadstoanincreaseinstatetaxeshavingnoimpactonrentalincomeoftheconstituents
intheentirefederation.Giventhatthecostofinternationalcapitalmobilityisequalto
zero,thereductioninnetreturnoncapitalduetoanincreasedtaxburdencanbesimply
offsetbyrelocatingcapitalabroad,therebygeneratingreturnsattheexogenousrater.The
withdrawalofcapitalwillinthemeantimeincreasecapitalproductivityinthestatesuptoa
pointatwhichthenetreturnoncapitalsettlesagainwhereitisequaltotheworldmarket
interest.Thatis,ifcapitalisperfectlymobile,ρwillequaltheworldmarketinterestrin
allstatesofthefederation.Thus,thetaxpolicyinstateihasnoimpactonthenetreturn

51

TillmannLisa

FederalTaxCompetition

oncapitaland,subsequently,noimpactontaxrevenueofstatej.Asaconsequence,the
horizontalexternalitywillvanishandtheverticalexternalitytriggeredbycapitalrelocating
abroadremains.Thisdominantbottom-upverticalexternalityiswhythestatetaxlevelis
inefficientlyhighinthecaseofperfectcapitalmarketintegration.Itwillbethestronger,the
greaterk,thatis,thegreaterthetaxbasereductionduetotheincreaseincapitaltaxation.
Thesecondcaseisthatofaclosedeconomy,inwhichcapitalsupplyequalscapitalde-
mand,whichimpliesthatalltheeffectsofataxhikewillbefullyincurredwithinthefed-
eration.Hence,σ→∞,whichimpliesthat∂τ∂ρ→−n1andk=1.Itthenfollowsthat(3.9)
willbeequalton∂∂ut=−1+Γg,whereΓgistheelasticityofconsumerutilitywithrespectto
regionalspending.ThehorizontalexternalitywillthendominateifΓg>1.Sowhenisthis
ase?ctheItfollowsfrom(3.8)that
Γg=1+(1−θ)(n−1).
n+(ft−θ)(n−1)
Therefore,providedthatn+(ft−θ)(n−1)>0,itfollowsthatΓg>1ifandonlyif
1+(1−θ)(n−1)>n+(ft−θ)(n−1).FromthisinequalityitcanbeinferredthatΓg>1
ift>0,suchthat,consequently,n∂∂ut=−1+Γg>0.Hence,thehorizontalexternalitywill
dominateifthecountryisisolatedfromtheinternationalcapitalmarketandift>0.The
intuitionbehindthisisstraightforward:Giventhatσ→∞impliesthattheeconomyis
closed,thefederaltaxbasebecomesindependentofthetaxesraisedbythestates,suchthat
theverticalexternalitywillapproachzero.Withrespecttothetaxburden,itcanbeinferred
thattwillbepositiveifthetaxrateonrentalincomeissufficientlysmall.Thisaspectcan
beeasilygraspedbythefactthatifrentalincomeisnottaxedatall,stategovernmentsmust
raiseataxoncapitalinorderforsufficientincometobegeneratedfortheprovisionofpublic
goods.Undertheassumptionofdecreasingmarginalutilityofpublicgoodprovision,the
utilityderivedfromthefirstsuppliedunitwillapproachinfinity,suchthatstateswillalways
haveanincentivetoprovideatleastacertainamountofpublicgoods.
Now,astheprecedingdiscussionshowed,forσ=0,theverticalexternalitydominatesthe
horizontalexternality,andforσ→∞,thehorizontalexternalitydominatestheverticalone.
Hence,assumingthatσiscontinuous,theremustexistsomeoptimaldegreeofintegration
intotheinternationalcapitalmarket,suchthatbothexternalitiesoffseteachother.Itcan
concluded:ebProposition1ConsideraNashequilibriumwithsymmetricstrategiesofthestates.Then:
i.Theverticalexternalitywilldominatethehorizontalexternality(therewillbeovertax-
ationatthestatelevel)ifthefederationisstronglyintegratedintotheinternational
capitalmarket(σsmall).

52

LisaTillmannFederalTaxCompetition
ii.Thehorizontalexternalitywilldominatetheverticalexternality(therewillbeundertax-
ationatthestatelevel)ifσislargeandifstatetaxessatisfyt>0.
iii.Thereissomeσˆ∈[0,∞),suchthatbothexternalitieswilloffseteachother,ifσ=ˆσ.
Remark:t>0ifθissmall.
WithreferencetoKeenandKotsogiannis(2002),whofindthatthehorizontalexternality
willdominatetheverticalexternalityatthestatelevelifrentalincomeisfullytaxedand
t>0,itcanbefurtherexaminedhowthediscussionwillchangeifthestatesfullytaxrental
incomeinthissetting.Substractingntimes(3.8)from(3.9)andsettingθ=1yields
∂u∂un∂t−n∂t=(1−n)Γgtk>0,(3.11)
whichwillobviouslybethecaseonlyift>0.Hence,thesameresultisgeneratedasinKeen
andKotsogiannis(2002),thatforfullytaxedrentsthehorizontalexternalitywilldominate
theverticalexternalityatthestatelevelifrevenuegeneratedfromrentalincometaxationis
notsufficienttofinancetheprovisionofapublicgood.
Proposition2Letstatesfullytaxrentalincome(θ=1).Then:
i.Thehorizontalexternalitydominatestheverticalexternalityift>0.
3.2.2AreFederalTaxesTooHighorTooLow?
Similartothestategovernments,thefederalgovernmentintendstomaximisethewelfareof
allindividualswithinitsjurisdiction.Ifitdecidestoincreasetaxes,thiswillhaveadirect
impactonaggregatetaxationineverystate,whichwillinturneachaffecttheutilityof
anyindividualwithinthefederation.Thus,givenstatetaxationassetinaNashgame,the
federalgovernment’sfirstorderconditionisdepictedby
∂u=n∂ρ+1−θπn∂ρ+1
∂T∂τ∂τ
+Γgtk+θπn∂τ∂ρ+1(3.12)
+ΓGk+Tkn∂ρ+1=0.
∂τJustasbefore,horizontalandverticaleffectspointinoppositedirections.Withstate-level
taxationasestablishedintheprecedingsection,twocasescannowbedistinguishedwith
respecttotheassumedstrategicbehaviourofthefederalgovernment.
53

LisaTillmannFederalTaxCompetition
Inthefirstcase,thefederalgovernmentplaysNash.Itwillthenmaximisetheutilityof
eachindividualbychoosingitsbest-responseT,takingtheequilibriumstatetaxesasgiven.
Thatis,giventheshortcomingsofstatetax-setting,thefederalgovernmentwillimposean
optimalcapitaltaxrate.
Thesecondcaseischaracterisedbytheassumptionthatthefederalgovernmenthasa
first-moveradvantage.Stateandfederalgovernmentswillthenengageinasequentialgame,
withthelatterchoosinganoptimalT,allowingforthestatestosubsequentlysettheirtaxes
inaNashgamewhiletakingTasgiven.Thefederaldecision-makersmustthenchooseT
withrespecttot=t(T),whichisimplicitlydefinedbythefirst-order-conditionsofthestates.
Theywillthusmaximisetheutilityu=u[T,t(T)]ofarepresentativeindividualbysetting
du=∂ut(T)+∂u=0.(3.13)
∂T∂tdTIn(3.13),thefirsttermontheright-handsidedenotesthechangeinutilityanindividualin-
cursduetoashiftinoptimal-responsestatetaxestriggeredbyavariationinfederaltaxation.
Thesecondtermdenotesthedirectchangeinutilityduetoafederaltaxhikeasdepictedby
(3.12).Aimingfor(3.13)tobefulfilled,thedecision-makersmightdeemitoptimaltoseta
taxbydeviatingfromtheoptimalityconditiongivenby(3.12).Thequestionisthen,whether
thefederalgovernmenthasanincentivetodeviatefromitsoptimaltaxrateasdefinedin
(3.12)bychoosingahigherorlowerleveloftaxation.Infact,federaltaxrateswillbechosen
lowerthanintheNashequilibrium,suchthat∂T∂u>0inorderfor(3.13)tobefulfilled,if
∂t∂ut(T)<0,andviceversa.Inotherwords:IfstatesengageinNashcompetitionandthe
federalgovernmentactsasaStackelbergleader,itcanbeinferredthatthereisovertaxation
(undertaxation)atthefederallevelif∂t∂uandt(T)areofopposite(equal)signs.Morepre-
cisely,iffederalandstatetaxesarestrategicsubstitutes[t(T)<0],thenfederaltaxeswill
besetlower(higher)thanintheNashequilibriumifstatetaxesaretoolow(toohigh).The
reverseisthecaseifstateandfederaltaxesarestrategiccomplements.6Theseresultscan
beeasilygraspedintuitively:Assumingthatstatetaxesaretoolow,theoptimisingfederal
governmentwillaimtooffsetthestate-inducedinefficienciesbychoosingahigher(lower)
federaltaxthanintheNashequilibriumifstateandfederaltaxesarestrategiccomplements
(substitutes)withthegoaltoprovideanincentiveforstatestosubsequentlysettheirtax
ratessuchthattheconsolidatedtaxburdenwillbeefficient.
Propositions3and4summarisethediscussion:
Proposition3Letfederalandstatetaxesbestrategicsubstitutes[ie.t(T)<0].Then,in
aStackelbergequilibriumwithsymmetricstrategiesofthestates:
6Noteforthesakeofc∂uompletenessthatfederaltaxationwillbechosenefficientlyasintheNashequilibrium
ifstatetaxesareefficient(∂t=0).
54

TillmannLisa

FederalTaxCompetition

i.FederaltaxesarehigherthaninaNashequilibrium,iftheverticalexternalitydominates
thehorizontalexternalityatthestatelevel(statetaxesbeingtoohigh).
ii.FederaltaxesarelowerthaninaNashequilibrium,ifthehorizontalexternalitydomi-
natestheverticalexternalityatthestatelevel(statetaxesbeingtoolow).

Proposition4Letfederalandstatetaxesbestrategiccomplements[ie.t(T)>0].Then,
inaStackelbergequilibriumwithsymmetricstrategiesofthestates(ie.ti=tforalli):
i.FederaltaxesarehigherthaninaNashequilibrium,ifstatetaxesaretoolow(the
horizontalexternalitydominatestheverticalexternality).
ii.FederaltaxesarelowerthaninaNashequilibrium,ifstatetaxestiaretoohigh(the
verticalexternalitydominatesthehorizontalexternality).

ExampleNumerical3.2.3Theresultsgeneratedintheprevioussectionsarenowillustratedwiththehelpofanumerical
etLexample.i.f(k)=kα,
suchthatf=αkα−1andf=α(α−1)kα−2
ii.π=(1−α)kα
σ∂ρiii.∂τ=nα(α−1)kα−1−nσ
iv.Γ(g,G)=β(lng+lnG)
suchthatΓg=gβandΓG=Gβ.
Letfurtherα=0.25,β=0.1,n=10,r=0.125.

SimulationofasymmetricNashequilibriumyieldsthenumericalresultsasdepictedin
Table1.7Thesimulationdistinguishesthreecaseswithrespecttocapitalmarketintegration,
thatis,thetwobenchmarkcasesofperfectintegration(σ=0)andanisolatedcountry(with
σ=109),aswellasthecaseof‘optimal’capitalmarketintegration,inwhichtheproposed
valueforσisrealisedforwhichtheexternalitiesoffseteachother.
7‘opt’dNoteenotesthattheinoTableptimal1vaanluesasteriskthepdenotesarametersthesbhouldesttrespakefonseoraninNashefficienctsompeolutiontition,towbehilereacthehed.superscript

55

TillmannLisa

FederalTaxCompetition

σk∗t∗T∗kopttoptTopt
01.1760.0540.0421.3520.0370.037
0.02=σˆ0.9800.0660.0660.9800.0660.066
1091.0000.0680.1001.0000.1000.100

Table3.1:NashSimulation

Table1showsthatk=1forahighvalueofσ.Thatis,iftheeconomyisclosed,allcapital
willbeemployedinsidethefederation.Incomparisontotheoptimalvalues,statetaxeswill
betoolow.Bycontrast,ifthecountryisperfectlyintegratedintotheinternationalcapital
market(σ=0),thecountryisanetimporterofcapitalandstatetaxeswillbeexcessively
high.Inbothcases,thefederalgovernmentisassumedtoplayNashandsetitstaxoptimal
givenstatetaxation.Ifσtakesitsoptimalvalue,thefederationisanetexporterofcapital
andtherealisedtaxescoincidewiththeiroptimalvalues.
IfthefederalgovernmentactsasaStackelbergleader,thesimulationwillyieldtheresults
asdisplayedinTable2,distinguishingthesamethreecasesforσasabove.8

σt∗T∗Ts(sub)Ts(com)toptTopt
00.0540.0420.0520.0370.0370.037
0.02=σˆ0.0660.0660.0660.0660.0660.066
300.0670.0940.0680.0950.0940.094
Table3.2:StackelbergSimulation

StatesareassumedtoplayNashbysettingt∗.Further,T∗depictsthechoiceoftax
rateforthefederalgovernmentinaNashgame.Ifthefederalgovernmenthasafirst-mover
advantageandstateandfederaltaxesactasstrategicsubstitutes,therealisedfederaltax
willfallshortofitsoptimalvalueT∗,ifstatesareexcessivelylow,andviceversa.Onthe
contrary,inthecaseofstrategiccomplementarity,federaltaxationwillbe(slightly)toohigh,
ifstatetaxesareexcessivelylow,andviceversa.
Theresultscorrespondwithwhatthetheoreticalmodelsuggested.Thatis,thelikely
effectsoffiscalexternalitiesarelargelydependentontheassumptionsaboutthestrategic
behaviourofstateandfederalgovernmentsaswellasontheassumptionsabouttherelative
behaviouroffederalandstatetaxes.Thedegreeofinternationalcapitalmarketintegration
8NotethatinTable2thesuperscript‘S’referstothesolutionforthefederalgovernmentactingasa
Stackelbergleader.Theextension‘(sub)’referstothecaseofstrategicsubstitutabilityofstateandfederal
taxes,whiletheextension‘(com)’denotesthatfederalandstatetaxesarestrategiccomplements.For
reference,theoptimalandtheNashsolutionsaregivenaswell.
56

TillmannLisa

FederalTaxCompetition

functionsbycurbingtheextenttowhichachangeinτcausesaverticalorahorizontal
externalityrespectively.

Summary3.3

Thepreviousdiscussionshowedthattheinefficienciesoftaxationinafederalsystemare
potentiallyrelatedtotheextenttowhichacountryisintegratedintotheinternational
financialmarket.Itwasshownthatastrongdegreeofintegrationwillleadtothevertical
externalitydominatingthehorizontaloneatthestatelevel.Statetaxeswillthusbetoo
high.Thehorizontalexternalitywill,ontheotherhand,dominateatthestatelevel,ifthe
economyhasonlylimitedaccesstotheinternationalcapitalmarket.Thatis,ahighdegreeof
integrationintointernationalcapitalmarketsfunctionsbyreducingtherelativeimpactofthe
horizontalexternalityinsuchthatthenetreturnoncapitalisdeterminedmoredirectlyby
theexogenousworldmarketinterest.Atthesametime,theverticalexternalitywillbecome
increasinglyimportantbytaxbasebeingshiftedmoreflexiblyabroad,therebyreducingthe
federaltaxbase.Byendogenisingthedegreeofcapitalmarketintegrationitwasshownthat
thereexistsaspecificvalueforwhichbothexternalitiesoffseteachother.Hence,apolicy
balancewithrespecttotheopennessofaneconomymayinfactincuranoverallefficientlevel
oftaxation.GiventhatthemodelisavariationofKeenandKotsogiannis(2002),theresults
demonstratethattherecognitionofthedegreeofinternationalcapitalmarketintegration
functionsthesameastheinterestrateelasticityofsavings.
Theresultscanbefurtherspecifiedwithrespecttotheassumptionsaboutthestrategic
behaviourofthefederalgovernment.Thatis,giventheaboveresults,aNashfederalgov-
ernmentwillchoosetaxratesefficientlygivenstatetaxation.Ifthefederalgovernmenthas
afirst-moveradvantage,theresulthingesontheefficiencyofstatetaxesaswellasonthe
characterisationoffederalandstatetaxesasstrategicsubstitutesorcomplements.Inthe
formercase,aninefficientlylowstatecapitaltaximpliesthatfederaltaxationwillbelower
thanintheNashequilibrium,andviceversa.Inthelattercase,federaltaxeswillbehigher
thanintheNashequilibriumiftherateatwhichthestatestaxcapitalisinefficientlylow.
Theappliedmodelyieldsaconcisedepictionoftheimpactofinternationalcapitalmarket
integrationonoptimaltaxationstrategiesofstateandfederalgovernments.Obviously,the
analysedstructureissubjecttosomelimitations,especiallygiventhesimplifiedmodelworld
considered.yeconomForone,themodeldoesnotallowforanyheterogeneityamongthecitizens.Theconsider-
ationislimitedtoasinglerepresentativeconsumer,whichallowsforefficiencyconsiderations
only.Thefederalgovernmentdoesnotdifferentiatewithrespecttoindigenceorstructural

57

TillmannLisa

FederalTaxCompetition

differences.Yet,heterogeneityinpreferencesandneedswithinastateorafederation9may
distorttheresultsinsuchthatthefederalgovernmentdistributingitswealthevenlyacross
thestatesmayitselfcreateinefficiencies.Notallowingforheterogeneityinfactenablesfor
somesimplifyingassumptionsmadeinthemodel.Thesameholdsfortherecognitionof
heterogeneityacrossstates.Whatismore,bydroppingtheassumptionofsymmetriccost
functions,ananalysisoftheinteractionofhighandlowriskregionscouldbeatopicfor
furthermodelspecification.
Anassumptioninthediscussionoffederaltaxationistheprevalenceofcompleteand
perfectinformationofallplayers.Theextensionofthemodeltodecisionsunderuncertainty
mightyieldmeaningfulresults.Moreover,howwouldthesituationchangeifthestatesdid
notsimplytakethefederaldecisionasgiven,buthadachancetoreacttothetaxsetby
thefederalgovernment?Theoverallstructureofthetaxsystemiskeptfairlyconcise.More
complextaxinstruments,suchasthepossibilityoftaxdeductibilityorearnedincometax
credits,whichcouldbeofsomerelevancetotherealworld,arenotconsidered.
Whilepossibleinteractionsofthedifferentfederallevelsareincorporatedintothemodel,
thelatterisrestrictedtoasingle-economyanalysis.Noattentionispaidtopossiblepolicy
changesabroadtriggeredbyachangeincapitaltaxationathome.Anextensiontothemodel
couldstemfromtheanalysisofatwo-ormultiple-economyworld,intowhichthestrategic
interactionbetweencountriesisincorporated.Thefederalgovernmentitselfcouldthusreact
totaxcompetitionwithothercountries,therebytriggeringanotherhorizontalexternality.It
isthislastsetofobjectiveswhichthesubsequentchapterisgoingtoaddress.

9AsmodeledbyRizzo(2005),forexample.

58

4Chapter

TaxCompetitionandFederalDesign

Amultitudeofpapersexistsassessingtheeffectsofhorizontalcompetitionintermsofseveral
unitarystatesengagingincompetitionwitheachother.Papersassessingfederaltaxation
havemainlyfocusedontheinterrelationofverticalandhorizontalexternalitieswithinone
country,whileonlyafewapproachescharacterisetheinteractionoftwofederalstatesora
federalandaunitarystate.1Whatappearstobemissingisanintegratedapproachthat
systematicallycompareswithinonedistinctsetting,howachangeintheinstitutionaldesign
oftwocountriescompetingformobiletaxbaseinfluencestaxratechoices.Theintention
istotherebygiveasystematicoverviewoftheoccuringchangesthatarenotaffectedby
differencesinthemodelspecifications.
Inthatrespect,theapproachpursuedinthischapteristhefollowing:Insteadofadding
anotherlayertotheexistingmodelsoffederaltaxation,itgivesanoverviewofhowtax
ratessetbytwocountriescompetingformobilecapitalwillchangewithdifferentfederal
structures.Theaimistostriptheappliedmodelofallunnecessarycomplexity,inorderto
keepitassimpleaspossibleandtoaddclaritytotheresults.Asequential-movegamewith
revenue-maximisingLeviathangovernments2isexamined,which,tothebestofknowledge,
hasnotbeentreatedintheexistingliterature.Abaselinescenariooftwounitarystates
competingovermobilecapitalisfirstconsidered,where,obviously,onlyhorizontaleffects
occur.Themodelisthenextendedtoasecondscenarioinwhichonecountryisunitary
andanotherisafederation.Inathirdstep,theanalysisiscarriedoutforthecaseoftwo
federationsinteractingintheverysamesettingasbefore.Theaimistoverifytheclassic
1extentThesetowhichapproaceachheseachjurisdictionaldifferinleveltheirseesmodelthroughsptheecifications,intentsionsuchofasitsthecompeassumptionstitors.Seewithrsectionegard2t.1.3otforhe
detail.more2ofcourse,Thecsubjectonsiderationtoofdiscussion.aLSeeeviathan-gosectionv2ernmen.2.4tforasoppfurtherosedtodetail.awFromatelfare-maximisingheoreticalpdointoecision-makfview,erithes,
attractivLeviathane,papproacarticularlyhyieldswitharrelativegardetlyotheeasymtouhandleltiple-laymerodinelwithteractionsfairlydepictedclear-cutbyrtwesults,ocompwhicehtingmakesitfederations.quite

59

LisaTillmann

FederalTaxCompetition

resultofLeviathangovernmentsbeingrestrainedbytaxcompetitionwithothercountriesand
todescribetherespectiveexternalitiesatworkineachsettingandtheefficiencyofchosentax
rates.3Theimpactofataxhikeinonejurisdictiononrevenuesinotherjurisdictionsistaken
intoconsiderationaswellasthatofcoordinatedtaxhikesonindividualandconsolidated
es.unerevTherelevanceofthisstudybecomesapparentbyconsideringtheEuropeanUnion,as
discussedintheintroduction.Withfreecapitalmobilitybetweenmemberstates,itisa
relevantquestionwhetheritwillmakeadifferenceforacountrysuchasGermanytobe
competingintaxrateswithanotherfederalcountryoraunitarycountrysuchasFrance.
Fromadifferentperspective,withrespecttoapossiblefurtherenlargement,itmightbea
differentthingforGermanyorFrancewhennewmemberstatesjointheEuropeanUnion
withaninstitutionalstructurethatisunitaryorfederal.Whatismore,thepoliticalreality
frequentlyseesmemberstatesactingasindependententitiesaimingtomaximisetheirown
revenue,whileneglectingtoalargedegreetheeffecttheirownchoiceshaveonothermember
states.Thatis,anotherrelevantquestiontoaskwithrespecttotheEuropeanUnionis
whetherEU-widebudgetswouldprofitfromfurtherintegrationandcoordinationbetween
memberstates,andwhatroleintensifiedcapitaltaxcompetitionplaysinthatcontext.
Thischapterissetupasfollows:Thenextsectiongivesageneraloverviewofthemodel,
whilesections4.2,4.3,and4.4dealwiththeanalysisofthethreecasesoutlinedabove,that
is,twounitarycountries(4.2),aunitaryandafederalcountry(4.3)andtwofederalcountries
(4.4)competingovercapitaltaxbase.Section6concludesbysummarisinganddiscussing
theresultsandpossiblepolicyimplications.

4.1TheModel

Averybriefoverviewofthebasicmodelwillnowbegiven,whichwillbefurtherspecified
ineachofthefollowingsections.TheanalyticalframeworkissetupbytwocountriesA
andBcompetingformobilecapitalwitheachother.CountryA(B)isinhabitedbyn(m)
individuals,whoareimmobileandeachendowedwithoneunitofperfectlymobilecapital.4
IndividualschoosetoinvesteitherincountryAorcountryB,whilenocostisfacedfor
3Theefficiencyoftaxrates,thatis,insuchthatfromtheperspectiveofaLeviathangovernment,atax
willbeoptimalifitmaximisesrevenues.Obviously,itwillbeinefficientlyhigh,ifrevenuesrisebyadropin
the4respTheseectivtweotaaxrate,ssumptionsortoaolopplywitnottheherevcaseerseofcthease.SeeEuropeansectionU2nion,.2.4wforherefurthercapitaldetail.movesfreelybetween
memberstates.Themobilityofcitizensis-bylaw-alsogranted(exceptfortherecentlyjoinedmember
states,wherelabourmarketprotectionismledtothemobilityofthelabourforcebeingsomewhatlimited,
atleastuntil2014[seeEuropeanUnion(2009)],yetgivenlanguageand(tosomeextent)culturalbarriers,
labourmobilityissignificantlylesspronounced,whichcorrespondswiththeassumptionsofthemodel.

60

TillmannLisa

FederalTaxCompetition

internationalinvestment.
Thecapitalavailableineachjurisdictionisusedforproduction,whereaconstantre-
turnstoscaleproductionfunctionisconsidered,whichisidenticalinalljurisdictionsandat
leasttwicecontinuouslydifferentiable.Itdisplaysdiminishingreturnstoscaleinitsmobile
inputfactor,capital.Per-capitaproductionincountryC=A,B,denotedasyiC,isthen
ybdeterminedyiC=f(kiC)=f(kiC)kiC+πiC,
wherekiCistheper-capitaamountofcapitalemployedincountryC.Assumingthatfirms
areprofitmaximisers,f(kiC)kiCisthenthecapitalincomeperheadincountryC.πiC
denotestherentalincomeaccruingtoeachconsumerilivinginCfromanimmobilefactor.
CtrycouninHence,

πiC=f(kiC)−f(kiC)kiC,
fromwhichfollowsthat

πiC=−kiC.(4.1)
Eachindividualderivesutilityfromconsumptionfinancedbyindividualrentalincome
πitaxedattheexogenousrateθ5andthereturnoncapitalinvestments,whichistaxed
ataspecificratebyeachjurisdiction.Hence,theyaimtomaximisethenetreturnon
capitalinvestmentsρ=f(kiC)−τiC,wheref(kiC)isthemarginalproductofper-capita
capitalinvestedineachcountryCandτiCisthetotaltaxburdenimposedoncapitalineach
jurisdiction,whichwillallbefurtherspecifiedintheproceedingsections.Itfollowsthatthe
amountofinvestmentsineachcountryisafunctionofthenetreturnoncapitalandthe
capitaltaxrateeachjurisdictionchooses.Thatis,kiC=kiC(ρ+τiC),suchthat,applying
theorem,functionimplicitthe

1kiC=f(kiC)<0.

4.2TwoUnitaryCountries
ThecaseoftwounitarycountriesAandB6competingforcapitalwillfirstbeconsidered.The
setupissolvedbybackwardinductionasatwo-stagegameinwhichbothgovernmentsact
65θmaybedividedintoaregional(δ)andafederal(Δ)share,suchthatθ=δ+Δinafederation.
Here,obviously,n=0.

61

TillmannLisa

FederalTaxCompetition

asNashplayerssimultaneouslychoosingtheirtaxratetakingthatoftheotherjurisdiction
asgiven,whiletheindividualssubsequentlymaketheirinvestmentdecisions.Eachofthe
followingsubsectionscharacterisesoneparticularstageofthegame.Productionisdefined
.1.4sectioninas

4.2.1CapitalInvestment
Eachindividualhastheoptionofinvestingcapitalathomeorabroad,facinganetreturn
ybnegiv

ρ=f(KC)−TC,
inwhichf(KC)denotesthemarginalproductofcapitalinvestedandTCisthecorresponding
taxrateleviedwithinthejurisdictionofcountryAorBrespectively.Givenperfectcapital
mobility,non-arbitragewillleadtothenetreturnoninvestmentbeingequalacrossboth
thathsuctries,counρ=f(KA)−TA=f(KB)−TB
implyingthatρ=ρ(TA;TB).Totalcapitalsupplyisgivenby

n+m=KA(ρ+TA)+KB(ρ+TB)
withρ=ρ(TA;TB),fromwhichfollowsthat:

∂KA∂ρ∂TA∂TA=−∂KA+∂KB<0∈[−1;0),
∂ρ∂ρ

∂KB∂ρ∂TB=−∂KA∂T+B∂KB<0∈[−1;0).
∂ρ∂ρThatis,thenetreturnoncapitalinonejurisdictionwillfallforacorrespondingtaxhike.
Itwillfalllessthanone-to-one(exceptifcapitalsupplyisfixed),giventhatthereductionin
capitalinvestmentduetothelowernetreturnwillbeaccompaniedbyahigherproductivity
ofcapitalinvestments,justasdescribedinsection3.1.3.

62

TillmannLisa

4.2.2GovernmentObjectives
ForcountryA,theobjectivecanbedescribedthefollowingway:

FederalTaxCompetition

TmaxARA=TAKA(ρ+TA)+θπA(ρ+TA),
thatis,itaimstomaximiserevenuesRAgeneratedfromthetaxationofcapitalandrental
incomebychoiceofthetaxleviedoncapitalwithinitsjurisdiction.Eachgovernmentplays
Nash,thatis,uponchoosingitstaxrate,ittakestheothercountry’staxasgiven.Bearing
inmindthatρ=ρ(TA;TB)thefirst-orderconditionofcountryA’sgovernmentisdescribed
by

∂RA=KA(ρ+TA)+TAKA(∂ρ+1)+θΠA(∂ρ+1)=0.(4.2)
∂TA∂TA∂TA
Thatis,Awilltakeintoaccountthatitstaxratechoicewillhaveadirecttaxincomeeffect,
asdepictedbythefirsttermontherighthandsideof(4.2).Thesecondandthethird
termimplythatthegovernmentwillalsoincurachangeintaxbase.Forone,capitaltax
basewillbereducedbyinvestorswithdrawingcapitalandinvestingelsewhereinsteadasa
consequenceofthehighertaxburden.Ontheotherhand,theresultingshiftinoverallnet
returnoncapital(duetohighertaxratesandtheresultingreductionincapitalproductivity
intheotherregiontriggeredbytherelocationofinvestment)willagainattenuatethecapital
taxbaseeffect.Thesameappliestothereductioninrentalincome.Yet,eachgovernment
accountsfortheseeffectsonlyinsofarasitsowntaxbaseisconcerned,whilenottaking
intoaccountthatittriggersapositivehorizontalexternalityaffectingthetaxbasebythe
neighbouringcountry.Forsimplicity,using(4.1),equation(4.2)canberewrittenas
∂T∂RA=KA(ρ+TA)+(TAKA−θKA)(∂T∂ρ+1)]=0,(4.3)
AAwherethefirsttermontherighthandsideisthetaxrevenueeffectandthesecondtermis
thetaxbaseeffect.GiventhatcountryA’soptimalchoiceoftaxratesimplicitlydependson
thetaxratechosenbytheothercountry’sgovernment(andthesamereasoningholdsforthe
governmentincountryB),equilibriumischaracterisedbyTA=TA(TB)andTB=TB(TA),
suchthatρ=ρ(TA(TB);TB(TA)).Inordertodeterminewhetherfromtheperspectiveofthe
Leviathangovernmenttaxratesthuschosenwillbetoohighortoolowinequilibrium,a
coordinatedtaxhikeofbothcountriesisconsidered,whoseeffectonA’srevenuesisgivenby

∂(T∂R+AT)=KA(ρ+TA)+(TAKA−θKA)(∂T∂ρ+∂T∂ρ+1)].
BABA63

(4.4)

TillmannLisa

FederalTaxCompetition

Togaininsightonthesignof(4.4),equation(4.3),whichiszero,issubstractedfrom(4.4)
evhato

∂ρ∂R∂R∂(TA+ATB)−∂TAA=(TAKA−θKA)(∂TB)>0.(4.5)
TheanalogousexpressionholdsforcountryB.Thatis,revenuescouldbeincreasedbya
coordinatedtaxhikeinbothcountries,giventhatgovernmentsdonottakeintoaccount
thehorizontaleffectstriggeredbytheirchoiceoftaxrates.Thatcorrespondswiththe
classicresultthathorizontalcompetitionrestrainsLeviathangovernmentsintheirrevenue-
ions.tteninmaximising

4.3UnitaryandFederalCountry
Withtheinteractionofaunitaryandafederalcountrythesituationbecomesslightlymore
complex.CountryAisdividedintonregionseachinhabitedbyasingleconsumerendowed
withoneunitofcapital.CountryBisunitaryandinhabitedbymindividualspossessing
oneunitofcapitaleach.7Inhabitantsareassumedtobeevenlydistributedacrosstheir
homejurisdiction.Thesituationisanalysedbybackwardinductioninathree-stagegame.
First,thegovernmentsatthefederallevelsettheirtaxrates.TheyactasaStackelberg
leadertowardstheregionalgovernmentswhileengaginginNashcompetitionwitheachother.
Subsequently,theregionalgovernmentsincountryAchoosetheirtaxratesasNashplayers,
bearinginmindthetaxratessetbythefederalgovernments.Finally,theindividualsdecide
wheretoinvestthecapitalavailabletothem.

4.3.1CapitalInvestment
Individualsderiveutilityfromconsumption,whichisfinancedbythenetreturnoninvestment
plustheafter-taxrentalincome.TheutilitymaximisingindividuallivingincountryAthen
hasthechoiceofinvestingathome,inanyotherregionofthefederationorabroad(incountry
B,thatis).Likewise,eachinhabitantofcountryBchoosestoinvestinsidehishomecountry
orabroad,inanyofA’sregions.Hence,totalcapitalsupplyinoneregionofcountryAis
ybnegiv

kAi=kAiNAT+k∗iB,
7Tobeprecise:Forsymmetryreasons,Bisassumedtobedividedintomregionseachinhabitedbyone
individual,yetthetaxationofcapitaliscentralised,suchthatoneuniformnationaltaxrateislevied.

64

TillmannLisa

FederalTaxCompetition

wherei=1,...,nstandsforeachregionincountryA,kAiNATdenotestheamountofcapital
availableinAthatisactuallyinvestedinregioni,k∗iBdepictstheamountofcapitalavailable
inBthatisinvestedinregioniofcountryA.
TotalcapitalsupplyinanyofB’sregions8isgivenby

kB=kBNAT+kA∗,
wherekBNATdenotestheamountofcapitalavailableinBthatisactuallyinvestedincountry
Bperregion,kA∗depictstheamountofcapitalavailableinAthatisinvestedineachregion
ofcountryB.TotalinvestmentsinBcanbedescribedbyKB=nkB.
AllcapitalinvestedinsideregioniofcountryAwillbesubjecttotheconsolidatedtax
rateτAi=tAi+TA,wheretheformerdenotesthetaxleviedbyregioniandthelatterdenotes
thetaxleviedbythefederalgovernmentincountryA.9Allcapitalinvestedinsidecountry
BwillbetaxedatrateTB.Hence,theinvestmentdecisionsforoneindividuallivinginAor
Brespectivelycanbedescribedthefollowingway:

ρANAT=f(kAi)−τAi,
ρAINT=f(kB)−TB,
ρBNAT=f(kB)−TB,
ρBINT=f(kAi)−τAi,

inwhichρANATdenotesthenetreturntocapitalanindividuallivingincountryAgenerates
byinvestinginA,ρAINTisthenetreturnanindividuallivinginAgeneratesfrominvesting
abroad,thatis,incountryB.Analogously,ρBNATisthenetreturntoanindividuallivingin
BfrominvestinginsidehishomecountryandρBINTisthenetreturnforanindividualfrom
Binvestingabroad.CapitalinvestedinaregionofAwillthenearnitsmarginalproduct,
whichdependsonthetotalamountofcapitalsuppliedinthatregion,thatis,f(kAi),net
thetaxburdenapplicable,thatis,τAi.CapitalinvestedinBwillearnitsmarginalproduct
dependingonthetotalamountinvestedinBnettherespectivetax,whichisTB.Given
decreasingreturnstoscale,non-arbitragemustleadtothenetreturnoncapitalbeingequal
acrossregions10andcountries,suchthatρ=f(kAi)−τAi=f(kB)−TB,whichimpliesthat
ρ=ρ(ti=1,...,n;TA;TB).
8Giventhattheseareassumedtobesymmetricandonlysubjecttoauniformcentral-leveltax,itis
possibletoabstainfromanindexforeachregion.
9Weassumethatthefederalgovernmentwillsetoneuniformtaxacrossitsjurisdiction(whichdoesnot
havetobethecase,asWrede(2002)pointsout).
10WhichisalsowhyitispossibletowriteρANATwithoutanindexinthefirstplace.
65

TillmannLisa

Totalcapitalsupplyisgivenby

n+m=kAi(ρ+τAi)+KB(ρ+TB),
withρ=ρ(ti=1,...,n;TA;TB),fromwhichfollowsthat:

FederalTaxCompetition

∂kAi∂ρ∂t=−∂kAi∂tAi+∂KB∈[−1;0),
Ai∂ρ∂ρ∂kAi∂ρ∂TA=−∂kAi∂T+A∂KB∈[−1;0),
∂ρ∂ρ∂KB∂ρ∂TB=−∂kAi∂TB+∂KB∈[−1;0),
∂ρ∂ρforwhichthesamereasoningholdsasintheprevioussectionandwherefromthefirsttwo
itcanbeinferredthat
∂ρ∂ρ∂TA=n∂tAi.(4.6)
4.3.2GovernmentObjectives
TheLeviathangovernmentsineachjurisdictionaimtomaximisetaxrevenuesbychoiceof
therespectivetaxrate,whiletakingintoaccountthelikelytaxbasereactionstogiventax
licies.op

elevLStateThestategovernmentsi=1,...,nincountryAholdthefollowingobjectives:

(4.6)

tmaxAiri=tAikAi(ρ+τAi)+δπ(ρ+τAi)
Eachstatewillsetitstaxratetakingthoseofallotherstatesandthenationaljurisdictions
(thatactasStackelbergleaders)asgiven,suchthattAi=tAi(tj=i=1,...,n−1;TA;TB).Using
(4.1),eachstate’sfirst-order-conditionisgivenby

∂t∂rAi=kAi(ρ+τAi)+biA(∂t∂ρ+1)=0.
AiAi66

(4.7)

TillmannLisa

FederalTaxCompetition

wherebiA=(tAikAi−δkAi)<0denotesthechangeinregionaltaxbaseduetoarespective
taxhike,whichfollowsthesameargumentationasintheprevioussection.Itcanbeinferred
from(4.7)thateachregionalgovernmentwillagainaccountforthedirecttaxrevenueeffect
generatedbychoiceofitstaxrate(asdepictedbythefirsttermontherighthandside)
aswellasthetaxbaseeffectdepictedbythesecondtermin(4.7).Itneglectsthatthe
latterwillraiserevenuesintheotherjurisdictions,therebytriggeringapositivehorizontal
externality.Eachstategovernmentfurtherneglectsthatthefederaltaxbasewillpossibly
benegativelyaffectedbycapitalnotonlyrelocatingfromoneregiontoanotherwithinthe
federation,butalsobeingshiftedtocountryB.Thereby,thelocalentitieswilltriggera
verticalexternalitywhichpointstowardsstatetaxesbeinginefficientlyhighwithrespectto
consolidatedrevenues.Thesetupofthestatesimpliesthatequilibriumwillbecharacterised
byasymmetricsolutioninwhichallstatessetthesametax.Theeffectofacoordinatedtax
hikeofallstatesonlocalgovernmentrevenueisdepictedby

∂ρ∂rAn∂tA=kA(ρ+τA)+bA(n∂tA+1).(4.8)
Inordertogaininsightonthesignof(4.8),equation(4.7)issubstractedfrom(4.8),which
yields

(4.9)

dtdrA=n∂∂rtA−∂t∂rA=kA(ρ+τA)+tAkA(n∂t∂ρ+1)
AAAA∂ρ−δkA(n∂tA+1)−kA(ρ+τA)
(4.9)∂ρ∂ρ−tAkA(∂tA+1)+δkA(∂tA+1)
=bA(n−1)∂t∂ρA>0.
Thatis,staterevenueswouldincreaseasaconsequenceofacoordinatedtaxhike.This
pointstowardsstatetaxesbeingsetinefficientlylow(thatis,notrevenue-maximising)from
theprespectiveoftheregionalLeviathangovernments.Itisinlinewiththeresultsfrom
section4.2andimpliesthatthetaxbaseeffectfrominvestorsrelocatingtheircapitalabroad
isoffsetbytherevenueeffect.

FederalLevel
Giventhesymmetricstateequilibrium,thefirst-moverfederalgovernment’sobjectiveinA
is

67

TillmannLisa

FederalTaxCompetition

TmaxARA=n[TAkA(ρ+τA)+ΔπA(ρ+τA)].
Itbearsinmindthatregionalgovernmentswillmaketheirtaxchoicedependentonwhat
ratewaspreviouslysetatthefederallevelandtakesthetaxrateofthefederalgovernment
inBasgiven.Itfollowsthat
ρ=ρ(ti=1,...,n(tj=i=1,...,n−1;TA;TB);TA;TB)
anditsfirst-orderconditionwillbegivenby

∂RA=n[kA(ρ+τA)+BA(∂ρ+n∂ρ∂tA+1+∂tA)=0,(4.10)
∂TA∂TA∂tA∂TA∂TA
inwhichBA=(TAkA−ΔkA)<0isthechangeinA’sfederaltaxbaseduetoarespectivetax
hike,analogoustotheprevioussection.Itfollowsfrom(4.10)thatthefederalgovernmentin
Awillperfectlyanticipatehowitstaxwillaffecttheoptimalchoiceofitsregionalgovernments
andthecorrespondingimpactonitstaxbase.Itneglectsthepositiveeffectontaxbasein
B,therebytriggeringahorizontalexternality,whichwillbecomeapparentfurtherdown.
CountryB’sgovernmentaimstosatisfy

TmaxRB=n[TBkB(ρ+TB)+θπB(ρ+TB)],
Bforwhichalsoρ=ρ(ti=1,...,n(tj=i=1,...,n−1;TA;TB);TA;TB).Thatis,itwillperfectly
anticipatethatitsowntaxhasanimpactonthetaxratesetbytheregionsintheneighbouring
country.11Thefirst-order-conditionisgivenby
11Withintheexistingliteratureonfiscalfederalism,theassumptionsmadewithrespecttotheextentto
whichonejurisdictiontakesintoaccountitscompetitors’reactionfunctionsdifferbetweenthemodelsand
aresubjecttodiscussion.Ofcourse,theyessentiallydrivethegeneratedresults.Inanearlycontribution,
Keen(1998)examinestaxationinasinglefederationwherebothlevelsofgovernmentaremyopic.Ina
laterpaper,KeenandKotsogiannis(2002)modelafederationinwhichthedecision-makersateachlevelare
perfectlyawareofthebudgetconstraintsattheotherlevel.Inasettingwheretwofederationscompetewith
eachother,Wrede(2002)assumesforbothlevelsineachfederationtoignorethebudgetconstraintofthe
otherjurisdictions.GrazziniandPetretto(2007)analysetheinteractionofafederalandaunitarycountry,
wherethefederal-levelgovernmentinonecountryisassumedtobeperfectlyawareoftheimpactitsown
taxratehasonthetaxessetbyitsregions.Also,thegovernmentintheunitarycountrytakesintoaccount
thetaxratesetattheneighbouringcountry’sfederallevel,butisignorantoftheinterdependencewith
regionaltaxratesinthatcountry.(Seealsosection2.1forfurtherdetailontherespectivemodels.)From
thisstudy’sperspective,itis,however,justaslikelythatgovernmentsofonecountrydotakethetaxrates

68

TillmannLisa

FederalTaxCompetition

∂RB=n[kB(ρ+TB)
∂TB(4.11)+BB(∂ρ+n∂ρ∂tA+1+∂tA)]=0,
∂TB∂tA∂TB∂TB
whereBB=(TBkB−θkB)<0isthechangeinB’sfederaltaxbaseandworksthesameway
(4.10).asEquilibriumischaracterisedbyTA=TA(ti=1,...,n(tj=i=1,...,n−1;TA;TB);TB)andTB=
TB(ti=1,...,n(tj=i=1,...,n−1;TA;TB);TA),andfurther

ρ=ρ[ti=1,...,n(tj=i=1,...,n−1;TA;TB);TA(ti=1,...,n(◦);TB);TB(ti=1,...,n(◦);TA)](4.12)
Inordertofindoutwhetherbothgovernmentswillsettheirtaxratestoohighortoolow
inequilibrium,acoordinatedtaxratehikebetweenthetwofederalgovernmentsisagain
considered,whichyieldsforcountryA:

∂RA∂ρ∂ρ∂ρ∂tA∂tA
∂(TA+TB)=n[kA(ρ+τA)+BA(∂TA+∂TB+n∂tA∂TA+1+∂TA)].(4.13)
Inordertodeterminethesignof(4.13),equation(4.10),whichiszero,issubstracted
gainto(4.13)from

∂RA∂RA∂ρ
∂(TA+TB)−∂TA=n[BA∂TB]>0.(4.14)
Thatis,giventhesetup,federalrevenueswouldincreasefollowingacoordinatedtax
hikeatthefederallevel.Sincethefederalgovernmentsperfectlyanticipatethereactions
attheregionallevel,theresultgeneratedhereisequivalenttothatwhenonlytwounitary
countriesarecompeting.Unlikethecasewhereonlyoneisolatedfederalstateisanalysed,
ofneighbouring‘foreign’regionsintoaccount,irrespectiveofthejurisdictionallevelatwhichthisrateisset.
TakethecaseofcorporatetaxationinFranceandGermany:itisstraightforwardtoarguethattheFrench
government,ifittakesGermancorporatetaxratesintoaccount,willbeawareoftheeffectivetaxburden
andnotlimititsviewtothefederallevel.Whatismore,theFrenchgovernmentmaybeevenmoreawareof
thetaxratessetinthedifferentregions,particularlyinthosesharingborderswithFrance.Thismodelmay
berestrictedtosymmetricequilibriaacrossregions,yetthepreviouspointstillcontributestothevalidityof
thediscussedassumption.Hence,whileingeneralitmaybeasomewhatvalidassumptionforgovernments
tobeentirelyignorantoftaxreactionsinotherjurisdictions,thisapproachisbuiltontheperceptionthatif
governmentsareassumedtobeawareofeachother’staxrates,thenitmustbemutuallycorrespondingand
shouldnotberestrictedtodecisionstakenatthesamelevel.Thisreasoningwillbeappliedinwhatfollows.
Note,however,thatpossiblethird-orhigher-orderinteractionsintaxratesareassumedtobenegligible.

69

TillmannLisa

FederalTaxCompetition

theoptimisingStackelbergfederalgovernmentneglectsthehorizontaleffecttriggeredbyits
owntaxrateontheneighbouringcountry’staxrevenues.Hence,thefederalgovernmentsin
theirturncauseahorizontalexternalityleadingtotaxratesbeingchosensuchthattheyare
notrevenue-maximising.Thesameholds,obviously,fromtheperspectiveofgovernmentB.

EffectsCross-Jurisdictional4.3.3Whatisinterestingwithrespecttotheanalysisofgovernmentbehaviourinthissettingare
theeffectstriggeredbyataxhikeinonejurisdiction,asdiscussedearlier,onrevenuesatother
levelsorinothercountries.Inordertocharacterisethese,theapproachtakenbyGrazzini
andPetretto(2007)andalsosuggestedbyWrede(2002)isfollowed.Thatis,theimpact
ofataxhikeinonejurisdictiononrevenuesofanotherjurisdictioncanbedeterminedby
assessingthechangeinequilibriumresponsesofthatotherjurisdiction.Forthestatelevel,
tAi=tAi(tj=i=1,...,n−1;TA;TB)foralliinthefederation.Using(4.12),theeffectofachange
inthetaxrateatthefederallevelonregionalbudgetsisthengivenby

drA=n∂tAkA(ρ+τA)+bA(∂ρ+n∂ρ∂tA+1+n∂tA).(4.15)
dTA∂TA∂TA∂tA∂TA∂TA
Here,itcanbeseenthatataxhikeatthefederallevelinAwillinturnaffecttheoptimal
choiceoftaxratesattheregionallevel(∂T∂tAA).Thatwillhaveanimpactdirectlyontax
revenuesasdenotedbythefirsttermontherighthandsideof(4.15).Itssigndependson
theassumptionsmadewithrespecttothestrategicrelationshipbetweenregionalandfederal
taxrates.Byassumingcomplementarity(ie.∂T∂tAA>0),itwillbepositive,andnegative
otherwise.12Thesecondtermdenotesthetaxbaseeffecttriggeredbythetaxhikeatthe
federallevel.GiventhatbAisclearlynegative,itssignhingesonthelastterminbrackets.
Ontheonehand,thetaxbasewillbereducedasaconsequenceofthehigherconsolidatedtax
ratetriggeredbythefederalhikeandtheresultingshiftinoptimaltaxratesattheregional
12Thestrategicrelationshipbetweentaxratesofdifferentjurisdictionsisofcoursesubjecttodiscussion.
Thereactioncurveestimatedforhorizontalinteractiontendstobeupward-sloping(seesection2.2.2),such
thatforcross-nationalandhorizontaleffectsitisafairapproachtoassumestrategiccomplementarity.Ina
federalsetting,itis,however,notunlikelyforthefederalgovernmenttosetitstaxrateoptimalgivenstate
taxation,whichmightincludeloweringitstaxasaresponsetoastate-leveltaxhikeinorderforaconsolidated
taxburdennottobeexceeded-especiallyifthefederalgovernmentbearsthefederation’srelativepositionto
othercountriesintheinternationalcontextinmind.Thatwouldpointtowardstheassumptionoffederaland
regionalgovernmentsbeingstrategicsubstitutes.Theempiricalevidenceforfederalinteractionismixed,as
discussedinsection2.2.3.Inthiscase,withtheconsiderationofrevenue-maximisingLeviathangovernments
focusingonlyontheirownbudgets,theassumptionofstrategiccomplementaritymightbesensible.Inwhat
follows,thefocuswilllieonthecaseofstrategiccomplementarity,buttheimplicationofdifferentassumptions
willbehighlightedwhereapplicable.

70

TillmannLisa

FederalTaxCompetition

level(1+n∂T∂tAA).Ontheotherhand,theseshiftswilleachreducetheoverallnetreturnon
capital,whichwillattenuatethedirectnegativetaxbaseeffectduetothetaxhikes.That
is,theimpactofafederaltaxincreaseonregionalrevenuesisnotnecessarilynegative.It
dependsonthestrategicinteractionofregionalandfederaltaxrates.Theobviouseffect
isthenegativetop-downexternalityresultingfromthelossofregionaltaxbaseduetothe
higherconsolidatedtaxburden.Thenagain,if,asaresponsetothefederalhike,theregions
choosetoraisetheirtaxrates(thatis,if∂T∂tAA),theywillgenerateapositivetaxrevenueeffect
andatthesametimefurtheraugmentthetaxbaseeffect.Therelativesizeofthesetwowill
thendrivethesignof(4.15).Ifitisgreaterthanzero,itpointstowardsfederaltaxrates
beinginefficientlylow.Ifitislessthanzero,itpointstowardstheverticalexternalityleading
tofederaltaxratesbeinginefficientlyhighfromtheperspectiveoftherevenue-maximising
regionalLeviathangovernment.Inordertodeterminethesignof(4.15),applying(4.6),it
canberewrittenas

∂t∂t∂ρ∂tdrdTAA=∂TAAkA(ρ+τA)+bA(∂TA(1+∂TAA)+1+n∂TAA).(4.16)
Here,itbecomesclearthatthelastterminbracketswillbeunambiguouslygreaterthanzero,
iftaxratesarestrategiccomplements,giventhat0<∂T∂ρA<−1and1+n∂T∂tAA>1+∂T∂tAA.
Hence,thetwotermsontheRHSof(4.16)haveoppositesigns,suchthatthetotaleffect
dependsontherelativemagnitudeofthetaxincomeeffectgeneratedfromtheshiftinregional
leveltaxrates,asopposedtothetaxbaseeffecttriggeredfromthereactionofinvestorstothe
changedinvestmentconditionsintherespectivejurisdiction.Thatis,dependingonwhether
thereactionoftaxratesisassumedtobemoreelasticthanthereactionofthetaxbase
towardsachangeinnetreturnoncapital,oneeffectwilloutweightheother.
AcoordinatedtaxhikebytheregionsinA,asconsideredinsection4.3.2,willaffectequi-
libriumfederalrevenueasfollows,bearinginmindthat
TA=TA(ti=1,...,n(tj=i=1,...,n−1;TA;TB);TB)andTB=TB(ti=1,...,n(tj=i=1,...,n−1;
TA;TB);TA):

dRA=n[kA(ρ+τA)∂TA
∂tdtAA∂ρ∂ρ∂TA∂ρ∂TB∂TA(4.17)
+BA(n∂tA+n∂TA∂tA+n∂TB∂tA+1+∂tA)],
fromwhichitcanbeinferredthattheequilibriumresponsesofnationaltaxratesinboth
countrieswillchangeasaconsequenceofacoordinatedtaxratehikeattheregionallevel
(∂t∂TAAand∂t∂TAB).A’schangeinfederaltaxratewillhaveadirectincomeeffect,depending
onthestrategicrelationshipbetweennational-levelandregional-leveltaxrates,asdepicted

71

LisaTillmann

FederalTaxCompetition

bythefirstterminsquaredbrackets.Thetaxbaseeffectwillagainbedrivenbythetax
increaseattheregionallevelandtheresultingshiftintheoptimalfederaltaxrate.Further,
investorswillrelocatetheirinvestmentsduetoachangeinnetreturntocapital,whichis
triggeredbythetaxhikeattheregionallevelaswellasbythechangeinoptimaltaxrates
setatthefederallevelinAandinB.Thesignofthelatterisinturndrivenbythestrategic
relationshipofTAandTBwithtAandwilleitheraugmenttheimpactoftheregionaltaxon
ρ(if∂t∂TAA>0and∂t∂TAB>0)orotherwiseattenuateit.Assumingagainfortaxratestobe
strategiccomplements,itbecomesapparentthatthesignofthetaxbaseeffectand,thus,
thesignof(4.17),willagainhingeonthemagnitudeofthereductioninnetreturnoncapital
incontrasttothetaxhikesatthefederalandregionallevel.Rewriting(4.17)yields

dtdRAA=n[kA(ρ+τA)∂t∂TAA
∂ρ∂TA∂TA∂ρ∂TB(4.18)
+BA(∂TA(1+n∂tA)+1+n∂tA+n∂TB∂tA)].
Itcanbeseenthatgiven−1<∂T∂ρA<0,itfollowsthat∂T∂ρA(1+n∂t∂TAA)+1+n∂t∂TAAwillbegreater
thanzero,suchthatthesignofthetaxbaseeffectisambiguous.Itwillbenegative,such
that(4.18)becomespositive,iftheelasticityofthefederaltaxrateinBasaresponsetothe
taxregimechangeinAissufficientlylargeforthetotaleffectonthetaxbaseinA(triggered
bytaxrisesonbothlevelsandattenuatedbythecorrespondingdropinthenetreturnon
capital)tobeoffset.Otherwise,thetaxbaseeffectwillbenegative,suchthatthesignof
(4.18)dependsontherelativemagnitudesofthetaxbaseeffectandthedirecttaxincome
effectinA.Now,whatthistellsusisthatifregionsinAagreeuponarevenue-increasing
taxhike,ontheonehand,forassumedstrategiccomplementarityoftaxrates,thefederal
governmentwillrealiseanincreaseintaxincome,whichmaybenamedapositivebottom-up
verticalexternality.Ontheotherhand,thecoordinatedhikewillcauseareductionintax
baseduetothehighertaxburdenandthelowernetreturntocapital,whichmightbecalled
afirst-ordernegativeverticalexternality.Whatismore,thateffectwillbeenhancedbythe
complementaryreactionoffederaltaxratesinA,butontheotherhandoffsettoacertain
degreebythecomplementaryreactionofnationaltaxratesinB(duetowhichcountryA
becomesinturnmoreattractiveforinvestors).13Therelativemagnitudeoftheselasttwo
effectsdrivesthesignofthetotaltaxbaseeffect.Considerthe(possiblymorelikely)case
wherethegovernment’sreactioninBisnotsufficientlystrongforthetaxbaseeffectofthe
coordinatedhikeandtheresultingriseatthefederalleveltobeoffset:Thentheimpactof
13Thatis,theacross-countryverticalexternalityresultingfromaregional-leveltaxincreasemayactby
couldlimitingatttenheuatetreductionheofhorizontfederalalrevexternalitiesenuesintA.riggeredPutbyadifferencross-btly,obyrderrtaxaisingcomptaxes,etitionA’satregionalthefgoederalvelevrnmenel.ts

72

TillmannLisa

FederalTaxCompetition

acoordinatedregionaltaxincreaseonfederalrevenuesinAinturndependsonwhetherthe
taxincomeeffectrealisedbythefederalgovernmentissufficientlylargetooutweighthetax
basereduction,whichagainhingesonhowelasticallycapitalownersreacttowardsagiven
changeintaxregime.14
Withrespecttothecross-nationaleffectsoftaxhikes,theimpactofahigherfederaltax
inBonregionalbudgetsinAisgivenby

dr∂t∂ρ∂ρ∂t∂t
dTBA=n∂TABkA(ρ+τA)+bA(∂TB+n∂tA∂TAB+∂TAB).(4.19)
Here,obviously,thetaxbaseeffecttendstobepositive,albeititisatleastalleviatedby
thetriggeredshiftinoptimalregionaltaxrates,ifstrategiccomplementarityisassumed.
Theimpactisfurtheramplifiedbythecorrespondingtaxincomeeffect.Dependingonhow
pronouncedthatreactioninAwillbe,thepositivetaxbaseeffectmayevenbeinverted,
suchthattheoverallsignwouldthenagainhingeontherelativesizeoftaxrevenueandtax
baseeffect.Notwithstandingtheinterestingimplications,thiswould,ofcourse,callforquite
extremereactionsbytheLeviathangovernmentsinA.Yetsuchaneffectistheoretically
possible.Withrespecttothemagnitudeoftaxratereactions,itisfurtherinterestingtonote
that∂T∂tBAislikelytoberathersmall.Inanycase,comparedto(4.16),itseemsreasonable
toassumethatthecross-nationalelasticitiesdifferfromtheinner-nationalelasticities(ie.
∂tA=∂tA).Forexample,German‘L¨ander’mightreactdifferentlytoahighertaxrateat
∂TtheBfederal∂TAlevelasopposedtoataxratehikeofnationaltaxratesinFrance.Notonlymay
thisbeduetothefactthattheawarenessoftaxbasereactionsmaynotbethesameacross
nations.Whatismore,dependingontheregionalproximitytoFrance,someregionsmay
bequitedirectlyaffectedbytaxratessetinFrance(takeSaarlandorBaden-W¨urttemberg,
forexample),whileothers(suchasSaxoniaorSchleswig-Holstein)arenot.Theseformer
regions’taxbaseswillthenagainbedirectlyaffectedbydifferenttaxregimesinFrance,to
whichtheymighthaveanincentivetoreact.Ontheotherhand,theywillstillbecompeting
withotherGermanregions,which,fortheirpart,mighthavenointentionofdirectlyreacting
towardstaxessetinFrance,giventhattheimpactontheirtaxbaseisratherlimited.Hence,
cross-nationalelasticitiesmaydifferacrossregionswithinonecountryand,onaverage,be
lowerthaninner-nationalelasticities.Thiswould,ofcourse,makeacasefortheconsideration
14Obviously,onemightarguethattheassumptionofstrategiccomplementarityoftaxratesacrossall
cojurisdictionsordinatedhdoikeesatnotthemakeregionalmuchlevselense.byloThewferingederalitsgotaxveraternmenintordermightforraathermorechofaosevotourablereactintowaternationalrdsa
positiontobeachieved.Themodeldoesgivethenecessarytoolstoanalysesuchbehaviouratthefederal
tolevethel,ibutnttheeresteddiscussionreader.ofallpossibletypesofreactionswouldbebeyondthescopeofthisstudyandisleft

73

TillmannLisa

FederalTaxCompetition

ofnon-symmetricequilibria,yetinthesametimecorroboratestheassumptionofadifference
in‘average’sensitivitywithrespecttoforeignnationaltaxrates.
Likewise,theimpactofanisolatedtaxhikeinBonthefederalbudgetinAisdescribed
by

dRA=n[kA(ρ+τA)∂TA
∂TdTBB∂ρ∂ρ∂tA∂ρ∂TA∂TA∂tA(4.20)
+BA(∂TB+n∂tA∂TB+∂TA∂TB+∂TB+∂TB)],
forwhichsimilarargumentationsregardingthedirectrevenueandthetaxbaseeffectholdas
for(4.19),exceptthatthetriggeredchangeinfederaltaxratesandthechangeintaxratesset
bythefollowingregionsfurtherattenuatethelatter.Hence,cross-nationalstrategicreactions
willinbothcasesleadtothepositiveimpactofataxhikeinBonrevenuesintherespective
jurisdictionallevelinAtobedepletedbythestrategicreactionofA’sjurisdictions.15
TheimpactofafederaltaxhikeinAonrevenuesinBisgivenby

dRB=n[kB(ρ+TB)∂TB
∂TdTAA∂ρ∂ρ∂tA∂ρ∂TB∂TB∂TB∂tA(4.21)
+BB(∂TA+n∂tA∂TA+∂TB∂TA+∂TA+n∂tA∂TA)],
whichbasicallyfollowsthesamelineofargumentationas(4.20),exceptthatitmightagain
bepossiblethat∂T∂tAA=∂T∂tBA,suchthat(4.20)and(4.21)mightdiffer.

4.3.4TotalRevenueEffects
AssumethattheLeviathangovernmentsofacountryarerequiredtoprovideacertainshare
oftheirrevenuestothecitizenship,whichwillbeusedfortheprovisionofpublicgoods.
Then,theinhabitantsofthatcountrywillbepositivelyaffectedbythechosenpoliciesand
thequestionwhetheranytaxreformwillmaximisetotalrevenueswithintheentirefederation
becomesrelevant.ThetotaleffectofacoordinatedtaxhikebytheregionsinAisgivenby

dRA+drA=n[kA(ρ+τA)∂TA
dtAdtA∂tA
+BA(∂ρ(1+n∂TA)+1+n∂TA+n∂ρ∂TB)](4.22)
∂TA∂tA∂tA∂TB∂tA
+bA(n−1)∂ρ,
∂tA15Withassumedstrategicsubstitutabilityoftaxrates,ofcourse,thequestiontobeaskedwouldbewhether
thenegativedirectrevenueeffectthustriggeredcouldbeoffsetbythefurtheramplificationofthetaxbase
effect.

74

TillmannLisa

FederalTaxCompetition

inwhichtheeffectonregionalbudgets(thirdline)isclearlypositiveandthatonfederal
budgets(firsttwolines)isambiguous.Whetherornottheentiretermbecomespositive,
dependsontheelasticityofthetaxregimeinBwithrespecttothetaxratechosenin
theregionsofcountryA.Ifthelatterissufficientlylarge,thetotaleffectispositiveand
consolidatedrevenueswillrisefollowingataxhikeattheregionallevel.Ifitisnot,thesign
of(4.22)hingesontherelativemagnitudesofthetaxbaseeffectasopposedtotheincome
effectgeneratedfromthetaxhikeaswellasthatofthe(positive)effectonrevenuesinthe
regions.(4.22)canberewrittenas

dRA+drA=n[kA(ρ+τA)∂TA
dtAdtA∂tA
+BA(∂ρ(1+n∂TA)+1+n∂TA+n∂ρ∂TB)(4.23)
∂TA∂tA∂tA∂TB∂tA
∂ρ11bA(n−n2)(∂TA)],
whereitcanbeseenthatfornsufficientlylarge,thethirdlinebecomeszero,suchthat
thesignof(4.23)isdefinedinthesamewayas(4.18)andhingesontherelativesizeof
thegeneratedtaxincomeandtaxbaseeffect.Theseareinturndrivenbythestrategic
interactionsoftaxratesandtheelasticityofthetaxbase.
Theeffectofafederaltaxhikeonconsolidatedrevenuesisdepictedby

dRA+drA=n[B∂ρ∂TB]+
AdTAdTA∂TB∂TA
∂tAkA(ρ+τA)+
∂TAbA(∂ρ+n∂ρ∂tA+1+∂tA),
∂TA∂tA∂TA∂TA
whichcanalsobeexpressedas

(4.24)

dRA+drA=n[BA∂ρ∂TB]+
dTAdTA∂TB∂TA
∂T∂tAkA(ρ+τA)+(4.24)
AbA(∂ρ(1+∂tA)+1+∂tA),
∂TA∂TA∂TA
where,ifstrategiccomplementarityisassumedthroughout,thefirsttwolineswillbeclearly
positive,whilethethirdlinewillbenegative.Whetherornotthetotaleffectispositive
dependsonwhetherthereductionofregionaltaxbaseinA(duetothehigherconsolidated
75

TillmannLisa

FederalTaxCompetition

taxrateandthecorrespondingreductioninnetreturntocapital)canbyoffsetbythetax
revenuegainattheregionallevel,whichresultsfromthehigherregionaltaxratesandthe
increaseinfederaltaxbasetriggeredbythetaxratehikeinBasaresponsetothefederal
taxhikeinA.Thatis,aclearpredictionofwhetherataxhikeatthefederallevelwould
raisetotalrevenuesinAcannotbegiven.

4.4InteractionofTwoFederalStates

Thenextcaseconsideredisthatoftwofederationscompetingformobiletaxbase.Thebasic
setupisthesameasintheprevioussection,exceptthatnow,countryBalsohasafederal
structure,suchthatineverycountry,eachregion’sinhabitanthasthechoiceofinvesting
capitalinanyregioni=1,...,nofcountryAorinanyregiong=1,...,mofcountry
Bwithcapitaltaxratessetatbothlevelsinbothcountries.Thethreestagesofdecision-
makingarecharacterisedthefollowingway:Bothfederalgovernmentschoosetheirtaxrates
first.Subsequently,theregionalgovernmentsinAandBchoosetheirtaxratestakingthe
taxessetpreviouslyatthefederallevelasgiven.Then,theinhabitantsofeveryregionchoose
wheretoinvesttheircapital.

4.4.1CapitalInvestment
Theinvestmentdecisionsbyeachindividualarecharacterisedinlinewiththoseinsection
4.3.1,exceptthatthefederalstructureofcountryBisnowalsoexplicitlyrecognised.That
is,capitalsupplyineachofB’sregionsisgivenby

kBg=kBgNAT+k∗gA,
whereg=1,...,mstandsforeachregioninB,kBgNATdenotestheamountofcapital
availableinBthatisactuallyinvestedinregiong,k∗gAdepictstheamountofcapitalavailable
inAthatisinvestedinregiongofcountryB.
AllcapitalinvestedinsideregiongofcountryBwillbesubjecttotheconsolidatedtax
rateτBg=tBg+TB,wherethefirstisthetaxrateleviedbyregiongandthelatterdenotes
thetaxrateleviedbythefederalgovernmentinB.Capitalprovisionandtaxationincountry
Aisdefinedthesamewayasinsection4.3.1.Investmentdecisionswillbesuchthat

76

TillmannLisa

ρANAT=f(kAi)−τAi,
ρAINT=f(kBg)−τBg,
ρBNAT=f(kBg)−τBg,
ρBINT=f(kAi)−τAi,

FederalTaxCompetition

whichfollowsthesamelinesasinsection4.3.1.
Non-arbitragewillagainleadtothenetreturnoncapitalbeingequalisedacrossjurisdic-
tions,suchthattotalcapitalsupplyisnowgivenby

n+m=kAi(ρ+τAi)+kBg(ρ+τBg),
withρ=ρ(tAi=1,...,n;TA;TB;tBg=1,...,m),fromwhichfollowsthat:

∂kAi∂ρ+Ai∂t=−∂kAi∂tAi∂kBg<0∈[−1;0),
∂ρ∂ρ∂kAi∂ρ∂TA+∂T=−∂kAiA∂kBg<0∈[−1;0),
∂ρ∂ρ∂kBg∂ρ∂t+Bg∂t=−∂kAiBg∂kBg<0∈[−1;0),
∂ρ∂ρ∂kB∂ρ∂T+B∂T=−∂kAiB∂kBg<0∈[−1;0),
∂ρ∂ρwhichisanalogoustosection4.3.1.Italsofollowsthat
∂ρ∂ρ∂TA=n∂tAi
and

∂ρ∂ρ∂TB=m∂tBg.

77

(4.25)

(4.26)

TillmannLisa

FederalTaxCompetition

4.4.2GovernmentObjectives
elevLStateThegovernmentdecisionswillbeanalysedforcountryA,giventhatthefederalsetupof
bothAandBissuchthattheeffectswillbethesameregardlessoftheperspectivetaken.
Eachregionaimstomaximiserevenuesbychoiceofitstaxrate,whiletakingtaxesset
intheotherregions(alsothoseofcountryB)andfederaltaxratesasgiven,suchthat
tAi=tAi(tAj=i=1,...,n−1;TA;TB;tBg=1,...,m).Theobjectiveisagain

tmaxAiri=tAikAi(ρ+τAi)+δπAi(ρ+τAi),
whereρ=ρ(tAi=1,...,n;TA;TB;tBg=1,...,m)andeachstate’sfirst-order-conditionisgivenby

∂rAi=kAi(ρ+τAi)+bAi(∂ρ+1)=0,(4.27)
∂t∂tAiAifromwhichthesameconclusionsasinsection4.3canbedrawn.Theanalogousobjective
holdsforregionalbudgetsinB,suchthat
tBg=tBg(tBh=g=1,...,m−1;TA;TB;ti=1,...,n).
Hence,equilibriumwillbecharacterisedbythetaxchoicesofeveryregioninAaswellasin
countryB,whichwilleachaffectthenetreturnoncapital,suchthat

ρ=ρ[tAi=1,...,n(tAj=i=1,...,n−1;TA;TB;tBg=1,...,m(◦));TA;TB;
tBg=1,...,m(tBh=g=1,...,m−1;TA;TB;tAi=1,...,n(◦))].
ItwillfurtherbecharacterisedbysymmetricstrategiesnotonlyofthestatesinAbutalso
ofthoseinB.
Inordertofindoutwhetherornottaxratesestablishedintheone-shotNashgamewillbe
toohighortoolowfromtheperspectiveoftherevenue-maximisingLeviathan,acoordinated
taxhikeattheregionallevelisagainconsidered.Theequilibriumeffectofacoordinated
hikeinAonregionalbudgetsinAisgivenby

∂rA=kA(ρ+τA)+bA(n∂ρ+m∂ρ∂tB+1).(4.28)
n∂tA∂tA∂tB∂tA
Thatis,thebudgetisnotonlyaffectedbythereductioninthenetreturnoncapitaldue
tothetaxhikeinA.ThathikealsotriggersanadaptionofchosentaxratesinallofB’s

78

TillmannLisa

FederalTaxCompetition

regions,whichwillinturnaffectthenetreturnoncapital,therebyreducingthelossof
taxbasefollowingthehikeinA.Theoptimisingregionalgovernmentdoesnottakethese
horizontaleffectsintoaccount.Substracting(4.27)from(4.28)yields

n∂∂rtAA−∂t∂rAA=bA[(n−1)∂t∂ρA+m∂t∂ρB∂t∂tAB].(4.29)
Hence,thesignof(4.29)hingesontheassumedrelationshipbetweenregionaltaxesinthe
twocountries,thatis,∂t∂tAB.Assumingstrategiccomplementarity,thetotaleffectwillbe
clearlypositive,whichisinaccordancewiththestandardresult16statingthattheLeviathan
governmentwouldprofitfromariseintaxrates.17Thatwouldpointtowardstaxrates
beingsetinefficientlylowfromtheperspectiveoftherevenue-maximisingregionalLeviathan
government.Incontrast,iftaxratesarestrategicsubstitutes,therevenueeffectdependson
theelasticityoftaxratesinBtowardsthosesetinA(∂t∂tAB)aswellasontherelativesize
ofbothcountries.ThelargerBiscomparedtoA,themorelikelyitsreactiontowardstax
hikesinAwillbetooffsettherevenue-increasingeffectofthattaxhikeforA.
Iftherewasachanceforbindingacross-countryagreements,theeffectofacoordinated
hikeofallregionsbothinAandinBonthebudgetofarepresentativeregioninAwould
begivenby:

∂ρ∂ρ∂rA∂(ntA+mtB)=kA(ρ+τA)+bA(n∂tA+m∂tB+1),(4.30)
wherethedifferenceto(4.29)isthat∂t∂tAB=1.Substracting(4.27)from(4.30)yields

drA∂rA∂rA∂ρ∂ρ
dtA=∂(ntA+mtB)−∂tA=bA[(n−1)∂tA+m∂tB]>0.(4.31)
Thatis,regionalbudgetsinA(andB)wouldclearlyincreasefollowingacoordinatedtax
hikeinbothcountries.Hence,horizontaltaxcompetitionagainseemstorestrainLeviathan
governmentsintheirrevenue-maximisingambitions.

FederalLevel
ThefederalgovernmentincountryAchoosesitsowntaxratetakingthatoftheother
country’sfederaldecision-makerasgiven.Itisawareofthefactthatthelower-leveltax
16See,forexample,KeenandKotsogiannis(2003).
17From(4.29)itbecomesapparentthat,undertheassumptionofstrategiccomplementarityofregional
taxratesacrosscountries,itwouldalwaysberevenue-enhancingfortheLeviathangovernmenttoraisetax
rates.Thatis,evenifregionalgovernmentsinonecountrydonothaveanyacross-countrytaxagreements,
theycanstillprofitfromaunilateraltaxrisewithintheirownjurisdictionifthestrategicreactionofthe
othercountryiscomplementary.Obviously,suchbehaviourwouldnotleadtoanequilibrium.

79

TillmannLisa

FederalTaxCompetition

ratesaresetasfunctionsofalltheothertaxrates,whichitwillperfectlyanticipate,notonly
forallofitsownregions,butalsoforallregionsofcountryB.ThefederalgovernmentinB
facesthesamesituation,fromwhichitcanbeinferredthat

TA=TA[ti=1,...,n(tj=i=1,...,n−1;TA;TB;tg=1,...,m);TB;
tg=1,...,m(th=g=1,...,m−1;TA;TB;ti=1,...,n)]
and

TB=TB[ti=1,...,n(tj=i=1,...,n−1;TA;TB;tg=1,...,m);TA;
tg=1,...,m(th=g=1,...,m−1;TA;TB;ti=1,...,n)].
TheobjectiveofthefederalgovernmentinAcanthusbedescribedby

TmaxARA=n[TAkA(ρ+τA)+Δπ(ρ+τA)],
anditsfirstorderconditionisgivenby

(4.32)

∂T∂RAA=n[kA(ρ+τA)
(4.33)BA(∂T∂ρ+n∂t∂ρ∂T∂tA+n∂t∂ρ∂T∂tB+1+n∂T∂tA)]=0.
AAABAA
Thatis,thefederalgovernmentinAwilltakeintoaccountthereactionattheregional
levelinresponsetoitschosentaxrate.Itwill,however,neglectthefactthatitstaxregime
affectstheoptimalchoiceoftaxrateatthefederallevelinB.Hence,inequilibrium

ρ=ρ[ti=1,...,n(tj=i=1,...,n−1;TA;TB;tg=1,...,n(◦));
TA(ti=1,...,n(◦);TB;ti=1,...,n(◦));
TB(ti=1,...,n(◦);TA;ti=1,...,n(◦));
tg=1,...,n(th=g=1,...,n−1;TA;TB;ti=1,...,n(◦))]
andtheequilibriumreactioninresponsetoataxhikeinAisdepictedby

∂eRA=n[kA(ρ+τA)
T∂Ae∂ρ∂ρ∂tA∂ρ∂tB∂ρ∂TB∂tA
+BA(∂TA+n∂tA∂TA+m∂tB∂TA+∂TB∂TA+1+n∂TA)].
80

(4.34)

(4.35)

TillmannLisa

FederalTaxCompetition

Substracting(4.33)from(4.35)tofindoutthesignofthelatteryields

dRA∂eRA∂RA∂ρ∂TB
dTA=∂e(TA)−∂TA=n[BA∂TB∂TA].(4.36)
Equation(4.36)capturesthehorizontalexternalitytriggeredbyataxhikeatthefederallevel
inA.Itssignessentiallyhingesontheassumptionregardingthestrategicrelationshipbe-
tweenfederaltaxratesinthetwocountries.Again,forstrategiccomplementsthehorizontal
externalitywillbepositive,pointingtowardstaxratesbeinginefficientlylowfromtheper-
spectiveoftherevenue-maximisingLeviathangovernment.Iftheyarestrategicsubstitutes,
itwillbenegative.
Iffederalgovernmentsinbothcountrieshadanincentivetocooperate,theeffectofa
coordinatedtaxhikeatthefederallevelinbothcountriesonfederalrevenuesinAwouldbe
ybaracterisedhc

∂R∂(TA+ATB)=n[kA(ρ+τA)+
∂t∂ρ∂ρ+BA(∂(TA+TB)+n∂tA∂(TA+ATB)(4.37)
∂t∂t∂ρAB+m∂tB∂(TA+TB)+1+n∂(TA+TB))].
Again,equation(4.33)issubstractedfrom(4.37)tofindoutthesignofthelatter,which
yields

dRA∂RA∂RA
dTA+TB=∂(TA+TB)−∂TA=(4.38)
∂ρ∂ρ∂tA∂ρ∂tB
n[BA(∂TB+n∂tA∂TB+m∂tB∂TB)].
Assumingfortaxratesattheregionallevelinbothcountriestobestrategiccomplements
withrespecttothetaxratesetatthefederallevelinB,(4.38)willbegreaterthanzero.On
theotherhand,iftheywerestrategicsubstitutes,thesignwouldhingeontheelasticityof
regionaltaxratesinbothcountrieswithrespecttofederaltaxratesinBasopposedtothe
effectonnetreturntocapitaltriggeredbythefederaltaxhikeinB.Thatis,bydropping
theassumptionoftaxratesbeingstrategiccomplements,therevenue-increasingeffectofa
coordinatedhikeacrosscountriesisnolongerclear.

EffectsCross-Jurisdictional4.4.3Inordertodeterminetheeffectstriggeredbyataxhikeatthefederalorregionallevelon
revenuesgeneratedattheotherlevels,theimpactoftheresultingchangesintheequilibrium
81

TillmannLisa

FederalTaxCompetition

choicesoftaxratesonrevenuesisagainconsidered.Forataxhikeattheregionallevelin
A,theeffectonA’sfederalrevenuesischaracterisedby

∂TdRA=n[kA(ρ+τA)A
∂tdtAA∂ρ∂ρ∂tB∂ρ∂TA∂ρ∂TB∂TA
BA(n∂tA+m∂tB∂tA+∂TA∂tA+∂TB∂tA+1+∂tA)],
which,using(4.25)and(4.26),canberewrittenas

dRA=n[kA(ρ+τA)∂TA
∂tdtAA∂ρ∂TA∂TA∂ρ∂tB∂TB(4.39)
BA(∂TA(1+∂tA)+1+∂tA+∂TB(∂tA+∂tA))].
Thatis,theimpactofachangeintaxregimeattheregionallevelinAonthebudgetof
thefederalgovernmentinAisdrivenbytheelasticityofthereactionoftaxratesatboth
levelsinBtothechangeintaxregimeinA.Equation(4.39)followsthesamelinesas
(4.18)inthecaseofafederationcompetingwithaunitarystate,exceptthatanadditional
verticalacross-countryeffectresultsfromthestatelevelgovernmentsinBreactingintheir
optimalresponsetaxratestothechoicesmadebyA’sregionalgovernments,therebyhaving
animpactonfederalrevenues.Inthecaseofstrategiccomplementarity,thereactionsatboth
levelsincountryBwillattenuatethetaxbaseeffectsresultingfromthetaxrateincreasein
A.Notonlywillinvestorsbefacedwithareductioninnetreturntocapitalforinvestments
duetothehighertaxinA,theywillalsoencounterlessfavourableinvestmentconditionsin
theothercountryasaresultofthestrategicinteractioninAandB,whichinturnreduces
theirincentivetorelocateinvestments.Thatis,thecross-nationaleffectsatbothlevelswill
makeitmorelikelyforthetaxhikeattheregionalleveltoraisefederalrevenues.
Inordertoassesstheeffectofafederaltaxincreaseonregionalbudgets,theresulting
changeinequilibriumtaxchoicescanbecharacterisedby:

∂tdrAAdTA=kA(ρ+τA)∂TA
∂ρ∂ρ∂tA∂ρ∂tB∂tA
+bA(∂TA+n∂tA∂TA+m∂tB∂TA+1+∂TA),
which,using(4.25)and(4.26),canberewrittenas

drA=kA(ρ+τA)∂tA
dT∂TAA+bA(∂ρ(1+∂tA)+1+∂tA+m∂ρ∂tB).
∂TA∂TA∂TA∂tB∂TA
82

(4.40)

(4.41)

TillmannLisa

FederalTaxCompetition

Thesignof(4.41)essentiallyhingesonthemagnitudeofm∂t∂ρB∂T∂tBA.Assumingagainstrategic
complementarityoftaxrates,thattermwillbenegative.Justasinsection4.3.3,therestof
thesecondlineof(4.41)willbepositive.Hence,ifthatlasteffectissufficientlylarge(thatis,
ifregionaltaxratesinBreactsufficientlyelasticallytowardsataxhikeinAtooutweighthe
totalimpactonthetaxbasetriggeredbythechangesinregionalandfederaltaxratesinA),
thetaxbaseeffectand,hence,(4.41)willbepositive,pointingtowardsfederaltaxationbeing
inefficientlylowfortheregionalLeviathans.18Otherwise,thesameargumentationholdsas
insection4.3.3.Thatis,incontrasttothecaseofafederalcountrycompetingwithaunitary
state,iftwofederationsinteract,thehorizontalcompetitioneffectismademoresignificant
bymeansoftheadditionalreactionofstategovernmentsincountryB.Inconsequence,just
asbefore,ataxhikeatthefederallevelbecomesmorelikelytoincreaseregionalrevenues,
giventhatthecorrespondingreactioninB’sregionsmayalleviatethelossoftaxbase.That
impliesthattheexistenceofverticalexternalitiesthatleadstoinefficientlyhighfederaltax
ratesismadelesssignificantthroughtheacross-countryeffectsarisingatbothlevelsofB.
TheimpactofafederaltaxhikeinBonregionalrevenuesinAcanbediscernedby

drA=kA(ρ+τA)∂tA
∂TdTBB∂ρ∂ρ∂tA∂ρ∂tB∂tA(4.42)
+bA(∂TB+n∂tA∂TB+n∂tB∂TB+∂TB),
forwhichthesamereasoningholdsasfor(4.19),exceptthatthepositiveeffectonA’sregional
taxbaseisfurtherintensifiedbythestrategicreactionattheregionallevelinB.
ThechangeinA’sfederalbudgetsfollowingashiftintaxregimeincountryBcanbe
ybeddescrib

dRA=n[kA(ρ+τA)∂TA
∂TdTBB∂ρ∂ρ∂tA∂ρ∂TA∂ρ∂tB∂TA∂tA(4.43)
+BA(∂TB+n∂tA∂TB+∂TA∂TB+n∂tB∂TB+∂TB+∂TB)].
Here,thesameconjecturesholdasfor(4.20),exceptthatthetaxbasereductionisfurther
attenuatedbythepossiblechangeintheoptimalchoiceoftaxratesattheregionallevelof
countryB,whichinturnleadstothetotaleffectbeingmorelikelytobepositive.Likewise,
theimpactofataxhikeattheregionallevelinBonfederalrevenuesinAisgivenby
18Themagnitudeofm∂t∂ρB∂T∂tABisalsoaffectedbytheelasticityofthenetreturntocapitalwithrespectto
B’sregionaltaxrateaswellasB’srelativesizecomparedtoA(ie.mopposedton).

83

LisaTillmann

FederalTaxCompetition

dRA=n[kA(ρ+τA)∂TA
∂tdtBB∂ρ∂ρ∂tA∂ρ∂TA∂ρ∂TB∂tA∂TA(4.44)
+BA(n∂tB+n∂tA∂tB+∂TA∂tB+∂TB∂tB+∂tB+∂tB)],
whichfollowsthesameargumentationas(4.43).

4.4.4TotalRevenueEffects
TheeffectofafederaltaxhikeinAontotalrevenueswithinthefederationisdescribedby

dRA+drA=n[BA(∂ρ∂TB)]+kA(ρ+τA)∂tA
dTAdTA∂TB∂TA∂TA(4.45)
+bA(∂ρ(1+∂tA)+1+∂tA+n∂ρ∂tB),
∂TA∂TA∂TA∂tB∂TA
whichwillbegreaterthanzero,iftheelasticityoftaxratesintheneighbouringregionsis
sufficientlyhigh,suchthattheoveralltaxbaseeffectwillbepositive.Otherwise,thesignof
(4.45)dependsontherelativemagnitudesofthetaxbaseeffectatthefederallevel(resulting
fromthetaxresponseinB)andthetaxrevenueeffectattheregionallevel(whichareboth
positive)incontrasttothe(negative)taxbaseeffectattheregionallevel.Theseareinturn
essentiallydrivenbytherelativemagnitudesofthestrategicreactionsoftaxratesinAand
B,respectively.ThetotalimpactofaregionaltaxhikeinAisdepictedby

dRA+drA=n[kA(ρ+τA)∂TA
dtAdtA∂tA
+BA(n∂ρ(1+∂TA)+1+∂TA+∂ρ(∂tB+∂TB)](4.46)
∂TA∂tA∂tA∂TB∂tA∂tA
bA[(n−1)∂ρ+n∂ρ∂tB],
∂tA∂tB∂tA
whichwillclearlybegreaterthanzero,iftheelasticityoftaxratesatthefederalandregional
levelinBissufficientlyhigh.Otherwise,itdependsontherelativemagnitudeofthetax
baseeffectinA’sregionsandtherevenueeffectatthefederallevel(whicharebothpositive)
incontrasttothetaxbaseeffectatthefederallevel(whichwillthenbenegative).

Summary4.5Theaimofthischapterwastoenhancethediscussionofefficientcapitaltaxationinthe
lightofinternationalcapitalmarketintegrationbyexaminingtheimpactofachangeina
84

TillmannLisa

FederalTaxCompetition

country’sfederalframeworkontaxsetting.Themodelisrestrictedtoaverybasicsetup.
Notwithstandingitssimplicity,ithelpstospecifyeachrelativeeffectofataxregimechange
onrevenuesinoneparticularjurisdiction.Whilethetotalimpactofthespecifiedeffectsis
ambiguous,theresultscanbesummarisedasfollows:
Inthereferencecaseoftwocompetingunitarystates,governmentswouldprofitfroma
coordinatedtaxhike,whichpointstowardstaxratesbeingsetinefficientlylowinequilibrium.
Thisfollowsfromthebroadlydiscussedhorizontalexternalitiescausedbyeachgovernment
neglectingthepositiveeffecttheirowntaxratechoicemighthaveonanothercountry’stax
base.Withtheintroductionofafederallayerinoneofthecountries,theupper-levelaswell
asthelower-levelgovernmentswillbesubjecttohorizontaltaxcompetitionandtendto
settaxratesinefficientlylow,thatis,notrevenue-maximisingfromtheirperspective,in
equilibrium.The‘standard’negativeverticalexternalitytriggeredbytax-settingatdifferent
levelsofgovernmentwithinafederation(whichpointstowardsregionaltaxratesbeingset
ineffcientlyhighfromthepointofviewofthefederalLeviathanandviceversa)isthen
accompaniedbypositivecross-nationalverticalexternalitiesfromtheregionsofthefederation
totheupper-levelunitarygovernment(whichpointtowardsregionaltaxratesbeingset
inefficientlylowfromtheperspectiveoftheforeigncountry)andviceversa.Whileitis
easilyvalidatedthattheirowntaxrateswillbesetinefficientlylowfromtheperspectiveof
eachrevenue-maximisingLeviathangovernment,theresultsregardingthecross-jurisdictional
revenueeffectsandthusalsotheimpactofataxhikeinonejurisdictiononconsolidated
revenues,areambiguous.Ifonejurisdictionincreasestaxes,ittriggersataxbaseeffectand
ataxrevenueeffectintheotherjurisdictions.Thetaxbaseeffectisdrivenbytherespective
externalitiesjustoutlined.Ittendstobenegativefortwojurisdictionsatdifferentlevels
withinthesamecountry.Ittendstobepositiveforcross-nationalrevenueeffectsoftax
regimechanges.Thesignineachcasehingesontheelasticityofthenetreturnoncapital
andtheelasticityoftheothergovernments’taxrateresponsetoagiventaxratechange.The
latterisincurredbythefactthattheequilibriumtaxrateforonejurisdictionwillchange
givenataxhikeinanotherjurisdiction,whichmayfurtheraugmentorattenuatethetaxbase
effect.Thatstrategicreactionintaxratesalsofeedsintothetaxincomeeffect,whichwillbe
positiveifconsideredtaxratesarestrategiccomplements,andnegativeotherwise.Taxbase
effectandtaxincomeeffecttendtohaveoppositesigns,suchthatthetotalimpactofatax
reforminonejurisdictiononrevenuesinanotherjurisdiction(and,thus,theanswertothe
questionwhethertherespectivetaxrateswillbetoohighortoolowfromtheperspective
oftheotherjurisdiction)isambiguousandessentiallyhingesontherelativeelasticityoftax
ratesincontrasttotheelasticityofthenetreturnoncapitaltoagiventaxrate.
Ifbothcountriesconsideredhaveafederalsetup,theresultschangeinsuchthatanother

85

TillmannLisa

FederalTaxCompetition

layerofcross-nationaleffectswillbeincurredbyataxhike.Forone,horizontalcompetitionis
intensifiedacrossbordersattheregionallevel.Therevenue-increasingeffectofacoordinated
taxhikewithinonecountryistheneitherattenuatedoraugmenteddependingontheassumed
strategicreactionofregionaltaxratesintheothercountry.Thefederal-levelgovernments
arestillsubjecttohorizontalcompetitionandtendtosettaxesinefficientlylow.Theimpact
ofataxreforminonejurisdictiononrevenuesinotherjurisdictionsofotherlevelsisagain
ambiguous.Basically,thesameargumentationholdsasinthepreviouscase,exceptthatthe
cross-nationaltaxcompetitioneffectsarenowaugmentedbythestrategicreactionofregions
federation.secondtheinInsummary,therecognitionofinternationaltaxcompetitionfunctionsbycross-national
reactionsatthecentrallevelattenuatingtheupwardpressureonconsolidatedtaxrates
resultingfromacountry’sfederalstructure.Ifbothcountriesarefederations,thesecross-
nationaleffectsarefurtherintensifiedbytaxinteractionsattheregionallevel.
Themodelmaybeextendedtoyieldfurthermeaningfulresultsbyincludingheterogeneity
amonglower-levelregions.Avariationintheorderoftaxratechoices(suchthatoneofthe
competingcountriesmovesfirstwiththeotheronefollowing)maymakethesetupmore
realistic.Whatismore,notwithstandingsomeempiricalsupport,thediscussioninevery
sectionmadeitclearthatitisfarfromsatisfyingtoassumethatalltaxrateswillbesetas
strategiccomplements.Whilethismaybevalidforthecaseofcross-countryinteractions,it
neednotbeforthestrategicinteractionoffederalandstatelevelgovernmentswithinone
country.Furthermore,themagnitudeofthetaxrateresponses,whichdrivesthesignofthe
totaleffect,mayalsodifferbetweenjurisdictions.Giventhattheempiricalevidenceonthe
strategicinteractionbetweendifferentlevelsofgovernmentismixed,itmaybewisetotry
anddepicttheimpactoftaxregimechangesbearingverycarefullyinmindtherespective
situation.Nonetheless,themodelprovidesanopportunitytohavearoughestimateoftherevenue
effectstriggeredbyapossibletaxregimechange,whileallowingforveryspecificindividual
assumptionswithrespecttothestrategicinteractionoftaxratesbetweendifferentlevelsof
governmentandcountriestobemade.

86

5Chapter

Conclusion

Theaimofthisstudywastosetoutthecurrentstateofresearchoninternationalfederal
taxcompetitionandtoaddfurtherinsighttothediscussionintwoways:Byanalysinghow
thedegreeofintegrationintointernationalcapitalmarketsaffectschosentaxratesandby
examininghowthefederalframeworkofacountryaffectsitspositioninthecompetitionfor
aninternationallymobilecapitaltaxbase.Inwhatfollows,thederivedresultswillagainbe
summarisedandtheirpossibleimplicationsfortaxpolicyevaluationwillbediscussed.

5.1ResultsofSummary

Theexistingtheoreticalliteraturesuggeststhathorizontaltaxcompetitionleadstoineffi-
cientlylowlevelsoftaxation.Thattendencymayforexamplebeoffsetbyaccountingfor
possibleasymmetriesbetweenregionsoragglomerationrents.Theconsiderationoftaxbase
overlapbetweendifferentlevelsofgovernmentpointstowardsinefficientlyhightaxrates.
Whichofthesedominatesissubjecttoavarietyofaspects.Theverticalexternalityislikely
todominateifacountryisstronglyintegratedintointernationalcapitalmarkets,ifcapi-
talsupplyisveryelasticwithregardtothenetinterestprevalentinoneregionorifthe
reverseisthecaseforcapitaldemand.Byincludingtaxcompetitionbetweenfederations,
theresultingcross-nationalexternalitiespointtowardsinefficientlylowtaxrates,suchthat
theinner-nationalverticalexternalityislesslikelytodominate.Theevaluationofresulting
equilibriaalsohingesontheassumptionwithregardtopolicymakersbeingbenevolentor
Leviathans.Theempiricalevidenceisaffirmativeofhorizontaltaxinteraction,whileatrendtowards
amoredifferentiatedviewrecognisingagglomerationeffectsandregionalisationhasenriched
theresults.Theevidenceonverticaltaxinteractiondifferssignificantlyacrosscountriesand
hes.approacChapters3and4extendthesmallexistingliteratureonfederaltaxationwithapar-

87

TillmannLisa

FederalTaxCompetition

ticularfocusontheinternationalisationofcapitaltaxcompetition.Chapter3developsan
approachinwhichtheintensityofcompetitionformobilecapitaltaxbaseisexplicitlymod-
elledandincorporatedintotheanalysis.Itestablishestheresultthatafederation’sdegreeof
internationalcapitalmarketintegrationfunctionsbyreducingtherelativeimportanceofthe
horizontalexternalityasopposedtotheverticalexternality,giventhatcapitaltaxbasecan
beshiftedmoreflexiblyabroad.Itisfurthershownthatforaspecificdegreeofinternational
capitalmarketintegration,verticalandhorizontalexternalitiesmayoffseteachother.
Chapter4takesamoredetailedlookatinternationalcompetitionforamobiletaxbase
withregardtoacountry’sfederalstructure.Itanalysestheimpactoftaxregimechanges
onindividualandconsolidatedrevenuesbygivingadetailedoverviewofthespecificeffects
triggeredineachjurisdiction.Insummary,byconsideringtwocompetingcountries,thetax
choicestakenateachdistinctlevelnotonlyaffectrevenueswithinthecountry,butalsoacross
theborders.Nexttotheimmediateimpactonrevenuesduetotaxbasereactions,taxregimes
intheneighbouringcountrymaychangeasaconsequenceofataxhikeathome.Thatis,
forafederationcompetingwithaunitarycountry,thefirst-movercentrallevelgovernments
becomesubjecttohorizontalcompetitionandsubsequentlytendtosettaxesinefficientlylow.
Theverticalexternalitiestriggeredbyonegovernmentattheregionallevelfeedbacknotonly
toitsowncentral-levelincumbent(negativeverticalexternality),butalsotothatoftheother
country(positiveverticalexternality).Inthereversecase,theregionalgovernmentinone
countryisalsosubjecttoverticalexternalitiesnotonlygeneratedbythefederalgovernment
athome,butalsobythatabroad,eachpointinginoppositedirections.Theintroductionofa
two-levelledgovernmentinbothcountriesfurtherintensifiesthehorizontaltaxcompetition
effectsineachjurisdiction.Thatis,nexttothecross-nationaleffectstriggeredbytheupper-
levelgovernments,theregionsinbothcountriesnoweachcausethesketchedexternalities.
Thesecross-nationalinteractionsworkbypotentiallycounteractingtheverticaltaxexternal-
itiestriggeredwithinonefederation.Theexplicitrecognitionofanothercompetingcountry
hencepointstowardstherelevanceofverticaltaxexternalitiesasasourceofinefficientlyhigh
taxratesbeingreduced.Theextenttowhichthisoccursessentiallyhingesontherelative
strategicrelationshipbetweentaxratesintherespectivejurisdictions,thesensitivityofthat
relationshipandtherelativesizeofthetwocountriesconsidered.
Fromadifferentperspective,chapter4alsoimpliesthattwocompetingunitarycountries
willfindthemselvesinequilibriawithinefficientlylowtaxrates.1Withtheconsideration
ofcompetingfederations,theprobabilityofinefficientlylowtaxationwillbereducedfrom
theperspectiveoftheindividualfederationaswellasfromanaggregateview.2Assuch,
theresultsbychapter4havetwodimensions:Incontrasttothestandardresultoftaxation
14.2.sectionSee2Seesections4.3and4.4.

88

TillmannLisa

FederalTaxCompetition

withinoneisolatedfederation,3therecognitionofanothercompetingcountrypointstowards
theprobabilityoftheverticalexternalitydominatingthehorizontalexternalitybeingreduced.
Incontrasttothestandardtaxcompetitionresult,chapter4indicatesfortheprobability
ofinefficientlylowtaxratesinequilibriumtobereducedbytakingintoaccountthefederal
structureofacountry.
Withregardtothefirstaspectderivedfromchapter4,theresultsderivedinchapters
3and4indicatethattheconsiderationofcapitalmarketintegrationpointsintheopposite
directiontotherecognitionofcross-nationalinteractions.Thatis,inacountrywithahigh
degreeofopenness,theverticalexternalityresultingfromfederalstructurewillbemorelikely
todominate,whiletheinterdependenceofafederationwithanothercountrywillresultin
theverticalexternalitybeinglesslikelytodominateasaconsequenceoftaxrateinteraction
andtaxbaseeffects.Theresultshavebeengeneratedfromtwodistinctmodelsconsidering
revenue-maximisersinonecaseandwelfare-maximisersintheother,andneedtobedirectly
comparedwithcaution.4Theymaynonethelesspointtowardsamoredifferentiatedviewof
globalisationandEUintegration,aswillbediscussedinthefollowingsection.

IolicyP5.2mplications

Theintroductionhasoutlinedthefollowingmainfieldsofpolicyrelevanceforthisstudy:
(i)theworldwideintegrationofcapitalmarketsandaresultinghigherdegreeofcapital
mobility;(ii)EUintegrationandenlargement,withtheconsequenceofcompetitionwithin
itsbordersbetweenseveralcountriesofdifferentfederalframeworks;(iii)theincreasingrole
oftheEuropeanUnionasanactiveplayerinitsownrightcompetingwithothereconomic
regionsandthepossibleintroductionofanEU-leveltax;(iv)aworldwidetrendtowards
decentralisation.Theimplicationsofthederivedresultswithregardtotheseissuesshallbe
outlinedinthissection.
Inasynthesisoftheexistingliterature,fourkeyaspects,asformulatedinchapter2.3,
needtobeborneinmindwithregardtotheevaluationoftaxpolicies.Forone,horizontal
taxinteractionsoughttobeaccountedforwithregardtotheheterogeneityofcompeting
regions.Nexttopossibleasymmetriesacrossregions,acountry’sfederalcompositionaswell
asthatofthecompetingcountriessignificantlyshapeshowitisaffectedbytaxcompetition.
Itisfurtherusefultoanalysewhetherornottheconsideredtaxpoliciesaremorelikely
tobeshapedbyrevenueorwelfaremaximisingintentionsofadecision-maker.Finally,
32.1.2.sectionSee4Inthatrespect,Wrede(1996,2002)haspointedtowardsthefactthattheLeviathanapproachmay
beseenasequivalenttoabenevolentgovernmentinthelimitifthemarginalvaluationofpublicspending
approachesinfinity.

89

TillmannLisa

FederalTaxCompetition

worldwidedecentralisationtrendsneedtobeevaluatedcautiouslywithrespecttoexpenditure
asopposedtorevenuedecentralisation.
Chapters3and4furthervalidatethementionedquestionsrelevantfortheanalysisof
taxpoliciesinlightofaninternationalisationofcapitalinvestments.Therespectiveresults
supporttheargumentthattheperceptionoftheconsequencesofglobalisationneednot
necessarilybecharacterisedbyreductionsinrevenuesorwelfareofthecitizens.Theresults
arerelevantforasingleEUmemberstateaswellasfortheconsiderationoftheEuropean
Unionasoneeconomicinstitutionwhoseoverallwelfareoughttobemaximised.Inthe
contextoftherecentfinancialcrisiswithresultingconcertedactivitiesbyallmemberstates,
suchaperceptionmaybeofhighrelevance.
Theworldwideintegrationofcapitalmarketshasasignificantimpactontheconsequences
ofacountry’sfederalframeworkfortaxratechoices,asshowninchapter3.Thatis,the
betterintegratedintointernationalcapitalmarketsafederationis,themorelikelyitisto
besubjecttoinefficientlyhighconsolidatedtaxrates.Thefearoftaxcompetitiontriggering
welfarelossesasaconsequenceoftheresultingdownwardpressureontaxratesthusneednot
bejustified.Infact,withintheframeworkofchapter3,welfareismorelikelytoberaised
bytaxratereductions.Itfurtherconcludesthatafederationmayfinditselfinastateof
‘optimal’integrationintocapitalmarketswithverticalandhorizontalexternalitiesoffsetting
eachother,suchthatconsolidatedtaxrateswillbeefficient.Thatis,bystrikingabalance
betweenopennessandclosednesswithregardtocapitalmarkets,afederationcompetingfor
mobilecapitalmayendupwithefficientlevelsoftaxation.Putdifferently,thederivedresults
suggestthattheregulationofcapitalmarketsneednotnecessarilyleadtowelfareorefficiency
losses,whichmaybeofparticularrelevanceinthecontextofcurrentdebatesintheaftermath
oftherecentfinancialcrisis.FromtheperspectiveoftheEU,perfectcapitalmobilitywithin
theUnionimpliesthatwithcountriescharacterisedbyfederalrevenuestructures,suchas
Austria,BelgiumorGermany,thewelfareeffectsofobservedtaxratereductionsinthecourse
ofcapitalmarketintegrationneednotnecessarilybenegativefromanaggregateperspective.
ThisviewalsocorrespondswiththeresultsbyBeckerandFuest(2010),5whoobservea
reductioninGerman‘L¨ander’taxrateswithrevenuesthatremainstable.
Withrespecttothefederalsetup,chapter4showsthatcompetitionfromanothercountry
mayfunctionbyoffsettingthetendencytowardsinefficientlyhightaxationpossiblyarising
fromacountry’sfederalstructure,therebyraisingwelfare.Towhatextentthisoccurshinges,
amongotherthings,onhowelasticallythecompetingcountriesreacttowardspossibletax
ratevariationswiththeirownchosentaxratesasopposedtothechangeinthenetreturn
oncapitalprevalentinthecountry.6Inthatrespect,afederalstructuremaythusbeseen
51.1.sectionSee6Thatelasticitymaybefurthervalidatedintheempiricalliterature.

90

LisaTillmann

FederalTaxCompetition

asimprovingacountry’spositionincompetingwithothercountriesthroughtheprevalent
inefficienciesitinitiallycreated.Itisfurtherimpliedthatfortheevaluationoftaxpolicies,it
isessentialforonecountrytoanticipatethelikelyreactionsintaxchoicesbyitscompetitors.
Theresultscanbeappliedtocross-nationaltaxcompetitionwithintheEUorglobaltax
interaction.Itmayfurtherbeconcludedthatpossibletaxbaseoverlapresultingfromthe
introductionofanEU-leveltaxmightimprovethepositionoftheUnionincompetitionwith
egions.rotherFromadifferentperspective,itisdemonstratedthatafederalstructuremaybefoundto
actbyattenuatingtheimpactofhorizontalcompetitioneffectsacrosscountries.Thatis,for
aunitaryEUmemberstate,thetrendtowardsinefficientlylowtaxratesresultingfromtax
competitionbetweenunitarycountriesmaybeoffsetbyfederalstructure.Possibletaxbase
overlaparisingfromanEU-leveltaxthusneednotbenegativeinthelightofinternational
taxcompetition.Fromthatperspective,theintroductionofagenuineEUrevenuesource
maybeseenasbeneficialbydampeningthehorizontalcompetitioneffectswithintheUnion
aswellasbyattenuatingthecompetitionmemberstatesarefacingfromoutsidetheUnion.
Obviously,withthedescribedtrendtowardsworldwidedecentralisation,theoutlined
resultsmayjustaswellapplytoaglobalperspectiveifthattrendimpliesrevenuerather
thanexpendituredecentralisation.Theformerhasalsobeenassociatedwithefficiencygains
inthecontextoftheempiricalLeviathanliterature.Fortheevaluationofresultingequilibria
itisofcourseoffurtherrelevancetoaccountforwhetheracountry’spoliciesarelikelyto
bedrivenbyrevenueorbywelfaremaximisation.Asacurrentexample,takethereform
offiscalstructureinItalyreferredtointheintroduction.Notwithstandingthatthisreform
mayhavebeenpoliticallydrivenbythedemandofthewealthyNorthernItalianregionsin
oppositiontotheSouth,itmayyetraiseefficiencyifoneconsiderstheargumentsmadeinthe
empiricalLeviathanliterature.Itmayfurtherbeaninterestingsubjectofanalysisregarding
theimplicationsoftheresultsderivedinthisstudy.
Theobservedstabilityofrevenuesdespiteareductionoftaxratesoutlinedintheintro-
ductionsuggestsforthewelfare-enhancingeffectsoftaxcompetitiontobeindeedrelevant.
Inviewofthederivedresults,thesemaybedrivenbyLeviathanbehaviourbeingthusre-
strainedor,possibly,bytaxcompetitionreducingoveralltaxrates,whichmayhavebeen
inefficientlyhighduetofederalstructurepreviously.
Giventhattheresultsderivedinchapters3and4pointinoppositedirections,there-
spectiveconsequencesoughttobeseenwithcautiousness.RegardingtheEuropeanUnion,
twothingsareworthmentioning:Forone,theprevalentperfectmobilityofcapitalmay
beindicativeoftheverticalexternalitybeingdominantwithinthefederalmemberstates,
suchthataggregatetaxlevelsareinefficientlyhigh.Yettheconsequencesofcapitalmarket
integrationmaybeoffsetbyconsideringstrategictaxinteractionsacrosscountries.Assuch,

91

TillmannLisa

FederalTaxCompetition

itmaybearguedthatahighlymobiletaxbaseneednotbeharmfuliftheEUasawholeis
considered.Ontheotherhand,regardingtheintroductionofanEU-leveltax,whichmaybe
characterisedbypossibletaxbaseoverlap,thesamemayholdfromtheperspectiveofthe
EUasawholeincompetitionwithothereconomicregions.
Insummary,thisstudyhascontributedtothewiderangeofliteratureontaxcompetition
byincludingthefederalstructureofacountryinasystematicanalysisandbyscrutinisingthe
impactafederation’sdegreeofopennesshasonchosentaxrates.Ithastherebyderivedtools
toconsiderindepththepolicyimplicationsofcapitalmarketintegrationandaworldwide
trendtowardsdecentralisationandregionalisation,particularlywithregardtotheEuropean
Union.Notwithstandingthatthetrade-offinresultsbetweenthesetwostrandsoughttobe
borneinmind,theresultsmaycontributetoamoredifferentiatedviewoftheconsequences
globalisation.of

92

6Chapter

References

Ahmad,EhthishamandGiorgioBrosio(2008),HandbookofFiscalFederalism,1stedition,
Cheltenham,UK,EdwardElgar.

Ahmad,EhthishamandMercedesGarcia-Escribano(2006),FiscalDecentralizationand
SubnationalFinancialManagementinPeru,IMFWorkingPaperNo.WP/06/120.

Allers,MaartenandJ.PaulElhorst(2005),TaxMimickingandYardstickCompetition
AmongLocalGovernmentsintheNetherlands,InternationalTaxandPublicFinance
493-513.12,

Altshuler,RosanneandTimothyJ.Goodspeed(2003),FollowtheLeader?Evidenceon
EuropeanandU.S.TaxCompetition,RutgersUniversity,DepartmentofEconomics,
DepartmentalWorkingPaperNo.200226.

Ambrosanio,MariaF.andMassimoBordignon(2008),NormativeversusPositiveTheories
ofRevenueAssignmentsinFederations,in:HandbookofFiscalFederalism,editedby
EhthishamAhmadandGiorgioBrosio,1stedition,Cheltenham,UK,EdwardElgar,
306-338.

Anderson,JohnE.andHendrikvandenBerg(1998),FiscalDecentralizationandGovern-
mentSize:AnInternationalTestforLeviathanAccountingforUnmeasuredActivity,
InternationalTaxandPublicFinance5,171-186.

Andersson,Linda,ThomasAronssonandMagnusWikstrom(2004),TestingforVertical
FiscalExternalities,InternationalTaxandPublicFinance11,243-263.

Aronsson,Thomas(2010),OptimalIncomeTaxationandDecentralizedFiscalFederalism,
RegionalScienceandUrbanEconomics40,187-195.

93

LisaTillmann

FederalTaxCompetition

Auerbach,AlanJ.(2007),WhyHaveCorporateTaxRevenuesDeclined?AnotherLook,
CESifoEconomicStudies,53(2/2007),153-171.

Bagaka,Obuya(2008),FiscalDecentralizationinKenya:TheConstituencyDevelopment
FundandtheGrowthofGovernment,MPRADiscussionPaperNo.11813.

Bahl,Roy(2001),FiscalDecentralization,RevenueAssignmentandtheCasefortheProp-
ertyTaxinSouthAfrica,InternationalStudiesProgramWorkingPaper01-7.

Banzhaf,SpencerandBenjaminA.Chupp(2010),HeterogeneousHarmvs.SpatialSpillovers:
EnvironmentalFederalismandU.S.AirPollution,NBERWorkingPaperno.15666,
NationalBureauofEconomicResearch,Cambridge,Massachusetts.

Bates,LaurieJ.andRexfordE.Santerre(2006),LeviathanintheCrosshairs,PublicChoice
133-145.127,

Becker,JohannesandClemensFuest(2010),InternationalizationandBusinessTaxRevenue
-EvidencefromGermany,InternationalTaxandPublicFinance17,174-192.

Besfamille,MartinandBenLockwood(2008),BailoutsinFederations:IsaHardBudget
ConstraintAlwaysBest?,InternationalEconomicReview49(2),577-593.

Besley,TimothyJ.andAnneCase(1995),IncumbentBehaviour:Vote-Seeking,Tax-Setting
andYardstickCompetition,AmericanEconomicReview85(1),25-45.

Besley,TimothyJ.andStephenCoate(2003),CentralisedversusDecentralisedProvision
ofLocalPublicGoods:APoliticalEconomyApproach,JournalofPublicEconomics
2611-2637.87,

Besley,TimothyJ.andHarveyS.Rosen(1998),VerticalExternalitiesinTaxSetting:
EvidencefromGasolineandCigarettes,JournalofPublicEconomics70,383-398.

Besley,TimothyJ.andMichaelSmart(2007),FiscalRestraintsandVoterWelfare,Journal
755-773.91,EconomicsPublicof

Bird,RichardM.andRobertD.Ebel(2008),FiscalFederalismandNationalUnity,in:
HandbookofFiscalFederalism,editedbyEhthishamAhmadandGiorgioBrosio,1st
edition,Cheltenham,UK,EdwardElgar,499-200.

94

TillmannLisa

FederalTaxCompetition

Boadway,Robin(2006),IntergovernmentalRedistributiveTransfers:EfficiencyandEquity,
in:HandbookofFiscalFederalism,editedbyEhtishamAhmadandGiorgioBrosio,
Cheltenham,UK,EdwardElgar,355-380.

Boadway,RobinandMichaelKeen(1996),EfficiencyandtheOptimalDirectionofFederal-
StateTransfers,InternationalTaxandPublicFinance3,137-155.

Boadway,RobinandDavidE.Wildasin(1984),PublicSectorEconomics,2ndedition,
Boston:LittleBrown.

Boadway,Robin,MauriceMarchandandMarianneVigneault(1998),TheConsequencesof
OverlappingTaxBasesforRedistributionandPublicSpendinginaFederation,Journal
453-478.68,EconomicsPublicof

Bond,EricW.andLarrySamuelson(1989),StrategicBehaviourandtheRulesforInter-
nationalTaxationofCapital,EconomicJournal99,1099-1111.

Bordignon,Massimo,FlorianaCernigliaandFedericoRevelli(2003),InSearchofYardstick
Competition:ASpatialAnalysisofItalianMunicipalityPropertyTaxSetting,Journal
199-217.54,EconomicsUrbanof

Brennan,GeoffreyandJamesBuchanan(1977),TowardsaTaxConstitutionforLeviathan,
JournalofPublicEconomics8,255-273.

Brennan,GeoffreyandJamesBuchanan(1978),TaxInstrumentsasConstraintsonthe
DispositionofPublicRevenues,JournalofPublicEconomics9,301-318.

Brennan,GeoffreyandJamesBuchanan(1980),AnalyticalFoundationsofaFiscalConsti-
tution,Cambridge:CambridgeUniversityPress.

Brett,CraigandJorisPinkse(2000),TheDeterminantsofMunicipalTaxRatesinBritish
Columbia,CanadianJournalofEconomics33(3),695-714.

Brueckner,JanK.(1982),ATestforAllocativeEfficiencyintheLocalPublicSector,
JournalofPublicEconomics19,311-331.

Brueckner,JanK.(2000),ATiebout/Tax-CompetitionModel,JournalofUrbanEconomics
285-306.77,

Brueckner,JanK.(2003),StrategicInteractionAmongGovernments:AnOverviewof
EmpiricalStudies,InternationalRegionalScienceReview26(2),175-188.

95

TillmannLisa

FederalTaxCompetition

Brueckner,JanK.(2004),FiscalDecentralizationwithDistortionaryTaxation:Tieboutvs.
TaxCompetition,InternationalTaxandPublicFinance11,133-153.

Brueckner,JanK.(2009),PartialFiscalDecentralization,RegionalScienceandUrban
23-32.39,Economics

Brueckner,JanK.andLuzA.Saavedra(2001),DoLocalGovernmentsEngageinStrategic
Property-TaxCompetition?,NationalTaxJournal54(2),203-230.

Bruelhart,MariusandMarioJametti(2006),VerticalVersusHorizontalTaxExternalities:
AnEmpiricalTest,JournalofPublicEconomics90,2027-2062.

Bruelhart,MariusandMarioJametti(2007),DoesTaxCompetitionTametheLeviathan?,
CEPRDiscussionPapers6512.

Bryson,PhillipJ.,GaryC.CorniaandGloriaE.Wheeler(2004),FiscalDecentralizationin
theCzechandSlovakRepublics:AComparativeStudyofMoralHazard,Environment
andPlanning:GovernmentPolicy22(1),103-113.

Buchanan,JamesandRichardA.Musgrave(2001),PublicFinanceandPublicChoice:Two
ContrastingVisionsoftheState,Cambridge,MITPress.

Bucovetsky,Sam(1991),AsymmetricTaxCompetition,JournalofUrbanEconomics30,
167-181.

Bucovetsky,Sam,MauriceMarchandandPatrickPestieau(1998),TaxCompetitionand
RevelationofPreferencesforPublicExpenditure,JournalofUrbanEconomics24,
441-459.

Bucovetsky,SamandJohnD.Wilson(1991),TaxCompetitionWithTwoTaxInstruments,
RegionalScienceandUrbanEconomics21,333-350.

Buettner,Thiess(1999),DeterminantsofTaxRatesinLocalCapitalIncomeTaxation:A
TheoreticalModelandEvidencefromGermany,Finanzarchiv56,362-388.

Buettner,Thiess(2001),LocalBusinessTaxationandCompetitionforCapital:TheChoice
oftheTaxRate,RegionalScienceandUrbanEconomics31,215-245.

Buettner,Thiess(2003),TaxBaseEffectsandFiscalExternalitiesofLocalCapitalTaxation:
EvidencefromaPanelofGermanJurisdictions,JournalofUrbanEconomics54,110-
128.

96

TillmannLisa

FederalTaxCompetition

Carlsen,Fredrik,BjorgLangsetandJornRattso(2005),TheRelationshipBetweenFirm
MobilityandTaxLevel:EmpiricalEvidenceofFiscalCompetitionBetweenLocal
Governments,JournalofUrbanEconomics58,273-288.

Case,AnneC.,HarveyS.RosenandJamesR.Hines(1993),BudgetSpilloversandFiscal
PolicyInterdependence:EvidencefromtheStates,JournalofPublicEconomics52,
285-307.

Cassette,AurelieandSoniaPaty(2008),TaxCompetitionAmongEasternandWestern
EuropeanCountries:WithWhomDoCountriesCompete?,EconomicSystems32,
307-325.

Cassette,AurelieandSoniaPaty(2010),FiscalDecentralizationandtheSizeofGovern-
ment:AEuropeanCountryEmpiricalAnalysis,PublicChoice143,173-189.

Cassing,JamesH.andAryeL.Hillman(1982),State-FederalResourceTaxRivalry:The
QueenslandRailwayandtheFederalExportTax,EconomicRecord58(162),235-241.

Charlot,SylvieandSoniaPaty(2010),DoAgglomerationForcesStrengthenTaxInterac-
tions?,UrbanStudies47(5),1099-1116.

Chen,Chien-Hsun(2004),FiscalDecentralization,CollusionandGovernmentSizeinChina’s
TransitionalEconomy,AppliedEconomicsLetters11,699-705.

Coase,Ronald(1960),TheProblemofSocialCost,JournalofLawandEconomics3,1-44.

Comola,MargheritaandLuizdeMello(2010),FiscalDecentralizationandUrbanizationin
Indonesia,UNU-WIDERWorkingPaperNo.2010/58.

Congleton,RogerD.(2008),AsymmetricFederalismandthePoliticalEconomyofDe-
centralization,in:HandbookofFiscalFederalism,editedbyEhthishamAhmadand
GiorgioBrosio,1stedition,Cheltenham,UK,EdwardElgar,131-153.

Crisan,Daria(2007),AnEmpiricalInvestigationofTaxCompetitionBetweenCanadian
Provinces,IAPRTechnicalPaperNo.TP-07002.

Crowley,GeorgeR.andRussellS.Sobel(2010),DoesFiscalDecentralizationConstrain
Leviathan?NewEvidencefromLocalPropertyTaxCompetition,mimeo.

97

TillmannLisa

FederalTaxCompetition

Dahlby,Bev(1994),TheDistortionaryEffectofRaisingTaxes,in:W.B.P.Robsonand
W.M.Scarth(eds.),DeficitReduction:WhatPain,WhatGain?,PolicyStudy23,
C.D.HoweInstitute,Toronto,44-72.

Dahlby,Bev(1996),FiscalExternalitiesandtheDesignofIntergovernmentalGrants,In-
ternationalTaxandPublicFinance3,397-412.

Dahlby,Bev,JackMintzandSamWilson(2000),TheDeductibilityofProvincialBusiness
TaxesinaFederationwithVerticalFiscalExternalities,CanadianJournalofEconomics
677-694.33,

Davies,RonaldB.andJohannesVoget(2009),TaxCompetitioninanExpandingEuropean
Union,IIISDiscussionPaperNo.276,Dublin.

deMello,Luiz(2008),TheBrazilian‘TaxWar’:TheCaseofValue-AddedTaxCompetition
AmongtheStates,PublicFinanceReview36(2),169-193.

dePater,JamesA.andGordonM.Myers(1994),StrategicCapitalTaxCompetition:A
PecuniaryExternalityandaCorrectiveDevice,JournalofUrbanEconomics36,66-78.

Deskins,JohnandBrianHill(2010),HaveStateTaxInterdependenciesChangedOver
Time?,PublicFinanceReview38,244-270.

Devereux,MichaelP.,RachelGriffithandAlexanderKlemm(2004),WhyHastheUK
CorporationTaxRaisedSoMuchRevenue?,FiscalStudies25(4),367-388.

Devereux,MichaelP.,BenLockwoodandMichaelaRedoano(2007),HorizontalandVertical
IndirectTaxCompetition:TheoryandSomeEvidencefromtheUSA,JournalofPublic
451-479.91,Economics

Devereux,MichaelP.,BenLockwoodandMichaelaRedoano(2008),DoCountriesCompete
OverCorporateTaxRates?,JournalofPublicEconomics92,1210-1235.

Dhillon,Amrita,MyrnaWoodersandBenZissimos(2007),TaxCompetitionReconsidered,
JournalofPublicEconomicTheory9(3),391-424.

Diamond,PeterandJamesMirrlees(1971),OptimalTaxationandPublicProduction:I
andII,AmericanEconomicReview61(8),261-278.

98

LisaTillmann

FederalTaxCompetition

Dinan,TerryM.,MaureenL.CropperandPaulR.Portney(1999),EnvironmentalFeder-
alism:WelfareLossesfromUniformNationalDrinkingStandards,in:Environmental
andPublicEconomics,editedbyArvindPanagariya,PaulR.PortneyandRobertM.
Schwab,Cheltenham,UK,EdwardElgar,13-31.

Eberts,RandallW.andTimothyJ.Gronberg(1988),CanCompetitionAmongLocalGov-
ernmentsConstrainGovernmentSpending?,FederalReserveBankofClevelandEco-
-9.224(1)Reviewnomic

Edmark,KarinandHannaAgren(2008),IdentifyingStrategicInteractionsinSwedishLocal
IncomeTaxPolicies,JournalofUrbanEconomics63,849-857.

Edwards,JeremyandMichaelKeen(1996),TaxCompetitionandLeviathan,European
113-134.40,ReviewEconomic

Eichenberger,ReinerandBrunoS.Frey(2008),Functional,OverlappingandCompeting
Jurisdictions(FOCJ):AComplementandAlternativetoToday’sFederalism,in:Hand-
bookofFiscalFederalism,editedbyEhthishamAhmadandGiorgioBrosio,1stedition,
Cheltenham,UK,EdwardElgar,154-181.

Esteller-Mor´e,´AlejandroandAlbertSol´e-Olle´(2001),VerticalIncomeTaxExternalities
andFiscalInterdependence:EvidencefromtheU.S.,RegionalScienceandUrban
247-272.31,Economics

Esteller-Mor´e,´AlejandroandAlbertSol´e-Olle´(2002),TaxSettinginaFederalSystem:The
CaseofPersonalIncomeTaxationinCanada,InternationalTaxandPublicFinance9,
235-257.

EuropeanCommission(2010),EUBudgetReview,SEC(2010)7000final,accessonlinevia:
http://eur-lex.europa.eu/SECMonth.do?year=2010&month=10,
(19th2011).anJ

EuropeanUnion(2009),PressReleaseIP/09/19,accessonlinevia:
ttp://europa.eu/rapid/pressReleasesAction.do?reference=IP/09/19,h2011).anJ(19th

Feld,Lars,Jean-MichelJosselinandYvesRocaboy(2002),Lemimetismefiscal:uneappli-
cationauxregionsfrancaises,EconomieetPrevision156(5),43-49.

99

TillmannLisa

FederalTaxCompetition

Feld,LarsandGebhardKirchgaessner(2001),IncomeTaxCompetitionattheStateand
LocalLevelinSwitzerland,RegionalScienceandUrbanEconomics31,181-213.

Feld,LarsandGebhardKirchgaessner(2002),TheImpactofCorporateandPersonalIn-
comeTaxesontheLocationofFirmsandonEmployment:SomePanelEvidencefor
theSwissCantons,JournalofPublicEconomics87,129-155.

Feld,Lars,GebhardKirchgaessnerandChristophA.Schaltegger(2010),DecentralizedTax-
ationandtheSizeofGovernment:EvidencefromSwissStateandLocalGovernments,
27-48.77(1),JournalEconomicSouthern

Feld,LarsandJohnG.Matsusaka(2003),BudgetReferendumsandGovernmentSpending:
EvidencefromSwissCantons,JournalofPublicEconomics87,2703-2724.

Feld,LarsandEmanuelleReulier(2005),LeFederalismeFinancierenSuisse,Revuede
351-371.94,ECl’OF

Feld,LarsandEmanuelleReulier(2009),StrategicTaxCompetitioninSwitzerland:Evi-
dencefromaPaneloftheSwissCantons,GermanEconomicReview10(1),91-114.

Feld,LarsandChristophA.Schaltegger(2009),DoLargeCabinetsFavorLargeGovern-
ments?EvidenceontheFiscalCommonsProblemforSwissCantons,JournalofPublic
35-47.93,Economics

Fenge,RobertandMatthiasWrede(2007),EUFinancingandRegionalPolicy:Vertical
FiscalExternalitieswhenCapitalisMobile,PublicFinanceAnalysisvol.63no.4.

Fink,AlexanderandThomasStratmann(2009),InstitutionalizedBailoutsandFiscalPol-
icy:TheConsequencesofSoftBudgetConstraints,CESifoWorkingPaperno.2827.

Fischel,WilliamA.(1976),AnEvaluationofProposalsforMetropolitanSharingofCom-
mercialandIndustrialPropertyTaxBase,JournalofUrbanEconomics3,253-263.

Flochel,LaurentandThierryMadies(2002),InterjurisdictionalTaxCompetitioninaFed-
eralSystemofOverlappingRevenueMaximisingGovernments,InternationalTaxand
121-141.,9FinancePublic

Flowers,MarilynR.(1988),SharedTaxSourcesinaLeviathanModelofFederalism,Public
67-77.16,uarterlyQFinance

100

TillmannLisa

FederalTaxCompetition

Forbes,KevinF.andErnestM.Zampelli(1989),IsLeviathanaMythicalBeast?,American
568-577.79(3),ReviewEconomic

Fredriksson,PerG.andKhawajaA.Mamun(2008),VerticalExternalitiesinCigarette
Taxation:DoTaxRevenuesGoUpInSmoke?,JournalofUrbanEconomics64,35-48.

Galligan,Brian(2008),InstitutionsofFederalismandDecentralizedGovernment,in:Hand-
bookofFiscalFederalism,editedbyEhthishamAhmadandGiorgioBrosio,1stedition,
Cheltenham,UK,EdwardElgar,521-543.

Gerard,Marcel,HubertJayetandSoniaPaty(2010),TaxInteractionsAmongBelgianMu-
nicipalities:DoInterregionalDifferencesMatter?,RegionalScienceandUrbanEco-
336-342.40,nomics

GermanCouncilofEconomicAdvisors(2005),DieChancennutzen-Reformenmutigvo-
ranbringen,Jahresgutachten2005/06,accessonlinevia:
http://www.sachverstaendigenrat-wirtschaft.de/50.html,(19Jan2011).

Geys,Benny(2006),LookingAcrossBorders:ATestofSpatialPolicyInterdependence
UsingLocalGovernmentEfficiencyRatings,JournalofUrbanEconomics60,443-462.

Goodspeed,TimothyJ.(1998),TaxCompetition,BenefitTaxesandFiscalFederalism,
NationalTaxJournal51(3),579-586.

Goodspeed,TimothyJ.(2000),TaxStructureinaFederation,JournalofPublicEconomics
493-506.75,

Goodspeed,TimothyJ.(2002a),BailoutsinaFederation,InternationalTaxandPublic
409-421.,9Finance

Goodspeed,TimothyJ.(2002b),TaxCompetitionandTaxStructureinOpenFederal
Economies:EvidencefromOECDCountriesWithImplicationsfortheEuropeanUnion,
EuropeanEconomicReview46,357-374.

Goodspeed,TimothyJ.andMatthieuLePrince(2003),VerticalCapitalTaxReactionFunc-
tions:EvidencefromSub-NationalGovernmentsinFrance,mimeo.

Gordon,RonaldH.(1992),CanCapitalIncomeTaxesSurviveinOpenEconomies?,Journal
1159-1180.47,Financeof

101

TillmannLisa

FederalTaxCompetition

Grazzini,LisaandAlessandroPetretto(2007),TaxCompetitionBetweenUnitaryandFed-
eralCountries,EconomicsofGovernance8,17-36.

Guriev,Sergej,EvgenyYakovlevandEkaterinaZhuravskaya(2010),InterestGroupPolitics
inaFederation,JournalofPublicEconomics94,730-748.

Hayashi,MasayoshiandRobinBoadway(2001),AnEmpiricalAnalysisofIntergovernmen-
talTaxInteraction:TheCaseofBusinessIncomeTaxesinCanada,CanadianJournal
481-503.36(2),Economicsof

Heyndels,BrunoandJozefVuchelen(1998),TaxMimickingAmongBelgianMunicipalities,
NationalTaxJournal51,89-101.

Hill,BrianC.(2008),AgglomerationandStrategicTaxCompetition,PublicFinanceReview
651-677.36,

Hindriks,JeanandGarethD.Myles(2006),IntermediatePublicEconomics,Cambridge,
Press.MIT

Ho,ChristinaS.(2010),HealthReformandDeFactoFederalisminChina,China:An
InternationalJournal8(1),33-62.

Hoyt,WilliamH.(1991),PropertyTaxation,NashEquilibriumandMarketPower,Journal
123-131.34,EconomicsUrbanof

Jayet,Hubert,SoniaPatyandAlainPentel(2002),Existe-t-ildesinteractionsfiscalesstrate-
giquesentrelescollectivit´eslocales?,EconomieetPrevision154(3),95-105.

Janeba,EckhardandGuttormSchjelderup(2009),TheWelfareEffectsofTaxCompetition
Reconsidered:PoliticiansandPoliticalInstitutions,TheEconomicJournal119,1143-
1161.

Jin,JingandHeng-fuZou(2002),HowDoesFiscalDecentralizationAffectAggregate,
National,andSubnationalGovernmentSize?,JournalofUrbanEconomics52,270-
293.

Johnson,WilliamR.(1988),IncomeRedistributioninaFederalSystem,AmericanEco-
570-573.78,Reviewnomic

Joulfaian,DavidandMichaelL.Marlow(1991),CentralizationandGovernmentCompeti-
tion,AppliedEconomics23,1603-1612.

102

TillmannLisa

FederalTaxCompetition

Kammas,Pantelis(2009),StrategicFiscalInteractionAmongOECDCountries,MPRA
WorkingPaperNo.15841,Munich.

Karkalakos,SotirisandMiltiadisMakris(2008),CapitalTaxCompetitionintheEuropean
Union:TheoryandEvidencefromTwoNaturalExperiments,MPRAWorkingPaper
h.Munic21437,No.

Keen,Michael(1998),VerticalTaxExternalitiesintheTheoryofFiscalFederalism,IMF
StaffPapers45,454-485.

Keen,MichaelandChristosKotsogiannis(2002),DoesFederalismLeadtoExcessivelyHigh
Taxes?,AmericanEconomicReview92,363-370.

Keen,MichaelandChristosKotsogiannis(2003),LeviathanandCapitalTaxCompetition
inaFederation,JournalofPublicEconomicTheory5,177-199.

Keen,MichaelandChristosKotsogiannis(2004),TaxCompetitioninFederationsandthe
WelfareConsequencesofDecentralization,JournalofUrbanEconomics56(3),397-407.

Keen,MichaelandMauriceMarchand(1997),FiscalCompetitionandthePatternofPublic
Spending,JournalofPublicEconomics66,33-53.

Kelejian,HarryH.andIngmarR.Prucha(1998),AGeneralizedSpatialTwo-StageLeast
SquaresProcedureforEstimatingaSpatialAutoregressiveModelwithAutoregressive
Disturbances,JournalofRealEstateFinanceandEconomics17,99-121.

Kelejian,HarryH.andDennisP.Robinson(1993),ASuggestedEstimationforSpatial
InterdependentModelswithAutocorrelatedErrors,andanApplicationtoaCounty
ExpenditureModel,PapersinRegionalScience72,297-312.

Kempf,HubertandGregoireRota-Graziosi(2010),EndogenizingLeadershipinTaxCom-
petition,JournalofPublicEconomics94,768-776.

Kerber,WolfgangandMartinaEckardt(2007),PolicyLearninginEurope:TheOpen
MethodofCo-OrdinationandLaboratoryFederalism,JournalofPublicPolicy14(2),
227-247.

Klor,Esteban(2006),APositiveModelofOverlappingIncomeTaxationinaFederationof
States,JournalofPublicEconomics90,703-723.

103

TillmannLisa

FederalTaxCompetition

K¨othenb¨urger,Marco(2002),TaxCompetitionandFiscalEqualisation,InternationalTax
andPublicFinance9(4),391-408.

K¨othenb¨urger,Marco(2004),TaxCompetitioninaFiscalUnionwithDecentralizedLead-
ership,JournalofUrbanEconomics55(3),498-513.

K¨othenb¨urger,Marco(2005),Leviathans,FederalTransfersandtheCartelizationHypoth-
449-465.122,ChoicePublicesis,

Kornai,Janos(1979),TheSoftBudgetConstraint,Kyklos39(1),3-30.

Kornai,Janos(1986),Resource-ConstrainedversusDemand-ConstrainedSystems,Econo-
47(4),metrica01-819.8

Kotsogiannis,ChristosandRobertSchwager(2006a),OntheIncentivestoExperimentin
Federations,JournalofUrbanEconomics60,484-497.

Kotsogiannis,ChristosandRobertSchwager(2006b),PoliticalUncertaintyandPolicyIn-
novation,JournalofPublicEconomicTheory8(5),779-805.

Kotsogiannis,ChristosandRobertSchwager(2008),AccountabilityandFiscalEqualisation,
JournalofPublicEconomics92,2336-2349.

Ladd,HelenF.(1992),MimickingofLocalTaxBurdensAmongNeighboringCounties,
PublicFinanceReview20,450-467.

Laffer,ArthurB.,StatementPreparedfortheJointEconomicCommittee,in:Laffer,Arthur
B.andJamesP.Seymour(eds.),TheEconomicsoftheTaxRevolt:AReader,New
rk.oY

Lalvani,Mala(2002),CanDecentralizationLimitGovernmentGrowth?ATestofthe
LeviathanHypothesisfortheIndianFederation,PubliusJournalofFederalism32(3),
25-45.

Lee,Kangoh(1997),TaxCompetitionWithImperfectlyMobileCapital,JournalofUrban
222-242.42,Economics

Lee,Kangoh(2003),FactorOwnershipandGovernmentalStrategicInteraction,Journalof
345-361.5(2),TheoryEconomicPublic

104

TillmannLisa

FederalTaxCompetition

LePrince,Matthieu,ThierryMadiesandSoniaPaty(2007),BusinessTaxInteractions
AmontLocalGovernments:AnEmpiricalAnalysisoftheFrenchCase,Journalof
603-621.47(3),ScienceRegional

Levi,Lucio(2009),Italy’sFiscalDevolutionMovesitTowardsFederalism,Federations
5-7.1,Magazine

Llanto,GilbertoM.(2009),FiscalDecentralizationandLocalFinanceReformsinthePhilip-
pines,PIDSDiscussionPaperSeriesNo.2009-10.

Lockwood,Ben(2002),DistributivePoliticsandtheCostsofCentralization,Reviewof
313-337.69(2),StudiesEconomic

Lockwood,Ben(2006),ThePoliticalEconomyofDecentralization,in:HandbookofFiscal
Federalism,editedbyEhtishamAhmadandGiorgioBrosio,Cheltenham,UK,33-60.

Madies,Thierry(2008),DoVerticalTaxExternalitiesLeadtoTaxRatesBeingTooHigh?
ANote,AnnalsofRegionalScience42,225-233.

Madies,Thierry,SoniaPatyandYvonRocaboy(2004),HorizontalandVerticalExter-
nalities:AnOverviewofTheoreticalandEmpiricalStudies,UrbanPublicEconomics
63-93.2,Review

Makris,Miltiadis(2006),CapitalTaxCompetitionUnderaCommonCurrency,Journalof
54-74.59,EconomicsUrban

Marlow,MichaelL.(1988),FiscalDecentralizationandGovernmentSize,PublicChoice56,
259-269.

Martinez-Vazquez,JorgeandMing-HungYao(2009),FiscalDecentralizationandPublic
SectorEmployment-ACross-CountryAnalysis,PublicFinanceReview37(5),539-
571.

Martinez-Vazquez,JorgeandMarkRider(2005),FiscalDecentralizationandEconomic
Growth:AComparativeStudyofChinaandIndia,InternationalStudiesProgram
WorkingPaperNo.05-19.

Mendoza,EnriqueG.andLindaL.Tesar(2005),WhyHasn’tTaxCompetitionTriggered
aRacetotheBottom?SomeQuantitativeLessonsfromtheEU,JournalofMonetary
163-204.52,Economics

105

TillmannLisa

FederalTaxCompetition

Mintz,JackandMichaelSmart(2004),IncomeShifting,Investment,andTaxCompetition:
TheoryandEvidencefromProvincialTaxationinCanada,JournalofPublicEconomics
1149-1168.88,

Mintz,JackandHenryTulkens(1986),CommodityandTaxCompetitionBetweenMember
StatesofaFederation:EquilibriumandEfficiency,JournalofPublicEconomics29,
133-172.

Miyamoto,Kazuko(2009),ANewApproachtoEstimatingTaxInteractionsinFiscalFed-
eralism,GlobalCOEHi-StatDiscussionPaperSeriesNo.053,HitotsubashiUniversity.

Nelson,MichaelA.(1987),SearchingforLeviathan:CommentandExtension,American
ReviewEconomic198-204.77(1),

Niskanen,WilliamA.Jr.(1971),BureaucracyandRepresentativeGovernment,Chicago:
Aldine.

Noiset,Luc(1995),Pigou,Tiebout,PropertyTaxation,andtheUnderprovisionofLocal
PublicGoods:Comment,JournalofUrbanEconomics38,312-316.

Noiset,Luc(2003),IsitTaxCompetitionorTaxExporting?,JournalofUrbanEconomics
639-648.54(3),

NoldHughes,PatriciaA.andMaryE.Edwards(2000),Leviathanvs.Liliputian:AData
EnvelopmentAnalysisofGovernmentEfficiency,JournalofRegionalScience40(4),
649-669.

NZZ-NeueZ¨uricherZeitung(2005),SuchenachM¨unteferings‘Heuschrecken’,3rdofMay
via:onlineaccess2005,http://www.nzz.ch/2005/05/03/wi/articleCSAVD.html,(19thJan2011).

Oates,WallaceE.(1972),FiscalFederalism,NewYork:HarcourtBraceJovanovich.

Oates,WallaceE.(1985),SearchingforLeviathan:AnEmpiricalStudy,AmericanEconomic
48-757.775(4),Review

Oates,WallaceE.(2002),AReconsiderationofEnvironmentalFederalism,in:RecentAd-
vancesinEnvironmentalEconomics,editedbyJohnListandArthurdeZeeuw,Chel-
tenham,UK,EdwardElgar,1-32.

106

TillmannLisa

FederalTaxCompetition

Oates,WallaceE.(2005),TowardASecond-GenerationTheoryofFiscalFederalism,Inter-
nationalTaxandPublicFinance12,349-373.

Oates,WallaceE.(2008),OntheEvolutionofFiscalFederalism:TheoryandInstitutions,
NationalTaxJournal61(2),313-334.

Oates,WallaceE.andRobertM.Schwab(1988),EconomicCompetitionAmongJurisdic-
tions:EfficiencyEnhancingorDistortionInducing?,JournalofPublicEconomics35,
333-354.

OECD(2010),RevenueStatistics2010,OECDPublishing,accessonlinevia:
www.oecd.org/ctp/revenuestats,(19thJan2011).

Parry,IanW.H.(2003),HowLargearetheWelfareCostsofTaxCompetition?,Journalof
39-60.54,EconomicsUrban

Persson,TorstenandGuidoTabellini(1992),ThePoliticsof1992:FiscalPolicyandEuro-
peanIntegration,ReviewofEconomicStudies59,689-701.

Persson,TorstenandGuidoTabellini(2002),PoliticalEconomics:ExplainingEconomic
Policy,MITPress.

Pigou,ArthurC.(1947),AStudyinPublicFinance,Macmillan,London.

Ploeg,Frederickvander(1996),BudgetaryPolicies,ForeignIndebtedness,theStockMar-
ket,andEconomicGrowth,OxfordEconomicPapers48,382-396.

Qian,YingyiandGerardRoland(1998),FederalismandtheSoftBudgetConstraint,Amer-
1143-1162.88(5),ReviewEconomicican

Rao,Giovinda(2008),FiscalFederalisminPlannedEconomies,in:HandbookofFiscal
Federalism,editedbyEhthishamAhmadandGiorgioBrosio,1stedition,Cheltenham,
UK,EdwardElgar,224-239.

Rauscher,Michael(1998),LeviathanandCompetitionamongJurisdictions:TheCaseof
BenefitTaxation,JournalofUrbanEconomics44,59-67.

Rauscher,Michael(2000),InterjurisdictionalCompetitionandPublicSectorProdigality:
TheTriumphoftheMarketOvertheState,Finanzarchiv57(1),89-105.

107

TillmannLisa

FederalTaxCompetition

Razin,AssafandEfraimSadka(1991),InternationalTaxCompetitionandGainsFromTax
Harmonization,EconomicsLetters37,69-76.

Redoano,Michela(2007),FiscalInteractionsAmongEuropeanCountries.DoestheEU
Matter?,CESifoWorkingPaperNo.1952.

Reulier,EmmanuelleandYvonRocaboy(2009),RegionalTaxCompetition:Evidencefrom
FrenchRegions,RegionalStudies43(7),915-922.

Revelli,Federico(2001),SpatialPatternsinLocalTaxation:TaxMimickingorErrorMim-
icking?,AppliedEconomics33,1101-1107.

Revelli,Federico(2003),ReactionorInteraction?SpatialProcessIdentificationinMulti-
TieredGovernmentStructures,JournalofUrbanEconomics53,29-53.

Revelli,Federico(2005),OnSpatialPublicFinanceEmpirics,InternationalTaxandPublic
475-492.12,Finance

Revelli,Federico(2008),SpatialInteractionsAmongGovernments,in:HandbookofFiscal
Federalism,editedbyEhthishamAhmadandGiorgioBrosio,1stedition,Cheltenham,
UK,EdwardElgar,106-130.

Richard,Jean-Francois,HenryTulkensandMagaliVerdonck(2002),DynamiquedesInter-
actionsFiscalesEntrelesCommunesBelges.1984-1997,EconomieetPrevision156,
1-14.

Richter,WolframF.andDietmarWellisch(1996),TheProvisionofLocalPublicGoodsand
FactorsinthePresenceofFirmandHouseholdMobility,JournalofPublicEconomics
73-93.60,

Rincke,Johannes(2006),PolicyInnovationinLocalJurisdictions:TestingforNeighborhood
InfluenceinSchoolChoicePolicies,PublicChoice129,189-200.

Rincke,Johannes(2009),YardstickCompetitionandPublicSectorInnovation,International
TaxandPublicFinance16,337-361.

Rizzo,Leonzio(2005),InteractionBetweenVerticalandHorizontalTaxCompetition:The-
oryandEvidence,Universit´adegliStudidiFerrara,Quadernon.5/2005.

Rizzo,Leonzio(2010),InteractionBetweenFederalTaxationandHorizontalTaxCompeti-
tion:TheoryandEvidencefromCanada,PublicChoice144,369-387.

108

TillmannLisa

FederalTaxCompetition

Rodden,Jonathan(2003),RevivingLeviathan:FiscalFederalismandtheGrowthofGov-
ernment,InternationalOrganization57,695-729.

Rork,JonathanC.(2003),CovetingThyNeighbors’Taxation,NationalTaxJournal56(4),
775-787.

Rork,JonathanC.andGaryA.Wagner(2008),DoRegionsMatterinInterjurisdictional
Competition?EvidencefromStateTaxation,EconomicsLetters101,116-118.

Rose-Ackerman,Susan(1980),RiskTakingandReelection:DoesFederalismPromoteIn-
novation?,JournalofLegalStudies9,593-616.

Salmon,Pierre(1987),DecentralizationasanIncentiveScheme,OxfordReviewofEconomic
Policy3,24-43.

Salmon,Pierre(2008),HorizontalCompetitionAmongGovernments,in:Handbookof
FiscalFederalism,editedbyEhthishamAhmadandGiorgioBrosio,1stedition,Chel-
tenham,UK,EdwardElgar,61-85.

Santolini,Raffaella(2009),LeviathanandLegislativeMonopolyPower:EmpiricalEvidence
FromOECDCountries,SocietaItalianadiEconomiaPubblica,SIEPWorkingPaper
635.No.

Salop,StevenC.(1979),MonopolisticCompetitionWithOutsideGoods,BellJournalof
41-156.110(1),Economics

Schubert,KlausandMartinaKlein(2006),DasPolitiklexikon,4thedition,Bonn,Dietz.

Shah,Ajay(2004),FiscalDecentralizationinDevelopingandTransitionEconomies:Progress,
Problems,andthePromise,WorldBankPolicyResearchWorkingPaper3282.

Shah,AjayandShekhaShah(2006),TheNewVisionofLocalGovernanceandtheEvolv-
ingRolesofLocalGovernments,in:Shah,A.(ed.),LocalGovernanceinDeveloping
Countries.WorldBank,WashingtonD.C..

Sobel,RussellS.(1997),OptimalTaxationinaFederalSystemofGovernments,Southern
468-485.64,JournalEconomic

Sol´e-Oll´e,Albert(2003),ElectoralAccountabilityandTaxMimicking:TheEffectsofElec-
toralMargins,CoalitionGovernmentandIdeology,EuropeanJournalofPoliticalEcon-
685-713.19,yom

109

TillmannLisa

FederalTaxCompetition

Sorin-Dinca,MariusandGheorghitaDinca(2009),FiscalDecentralizationinRomania-
PresentStateandPerspectives,AlexandruIoanCuzaUniversityWorkingPapers56,
103-116.

Stansel,Dean(2006),InterjurisdictionalCompetitionandLocalGovernmentSpendingin
U.S.MetropolitanAreas,PublicFinanceReview34(2),173-194.

Strumpf,Koleman(2002),DoesGovernmentDecentralisationIncreasePolicyInnovation?
JournalofPublicEconomicTheory4,207-224.

Tiebout,CharlesM.(1956),APureTheoryofLocalExpenditures,JournalofPolitical
416-424.64,yEconom

Uchimura,HirokoandSuzukiYurika(2009),MeasuringFiscalDecentralizationinthe
Philippines,IDEDiscussionPaperNo.209.2009.07.

Vandenbussche,Hylke,KarenCrabbe,andBoudewijnJanssen(2005),IsThereRegional
TaxCompetition?FirmLevelEvidenceForBelgium,DeEconomist153,257-276.

Wannisky,Jude(1978),Taxes,RevenuesandtheLafferCurve,PublicInterest50,Winter
-16.3Issue

Wartha,Udo(2007),DieReformdesFinanzierungssystemsderEurop¨aischenUnion:Eine
eigeneSteuerf¨urdieEU-Ebene?,Europ¨aischeHochschulschriften,SeriesV,Vol.3267,
terePang.L

Wellisch,Dietmar(2000),TheoryofPublicFinanceinaFederalState,CambridgeUniver-
sityPress,NewYork.

White,MichelleJ.(1976),FirmSuburbanizationandUrbanSubcenters,JournalofUrban
323-343.,3Economics

Wigger,BertholdU.andUdoWartha(2004),VerticalTaxExternalitiesandtheComposi-
tionofPublicSpendinginaFederation,EconomicsLetters84,357-361.

Wildasin,DavidE.(1988),NashEquilibriainModelsofFiscalCompetition,Journalof
229-240.5,3EconomicsPublic

Wildasin,DavidE.(1989),InterjurisdictionalCapitalMobility:FiscalExternalityanda
CorrectiveSubsidy,JournalofUrbanEconomics25,193-212.

110

TillmannLisa

FederalTaxCompetition

Wildasin,DavidE.(1997),HardandSoftBudgetConstraintsinIntergovernmentalFiscal
mimeo.Relations,

Wildasin,DavidE.andJohnD.Wilson(2004),CapitalTaxCompetition:BaneorBoon,
JournalofPublicEconomics88,1065-1091.

Wilson,JohnD.(1986),ATheoryofInterregionalTaxCompetition,JournalofUrban
296-315.19,Economics

Wilson,JohnD.(1991),TaxCompetitionWithInterregionalDifferencesinFactorEndow-
ments,RegionalScienceandUrbanEconomics21,423-451.

Wilson,JohnD.(1999),TheoriesofTaxCompetition,NationalTaxJournal52(2),269-304.

Wilson,JohnD.(2003),TaxCompetitioninaFederalSetting,PaperPreparedforthe
conference‘FiscalDecentralization:AnalysisandPerspectives’,BocconiUniversity.

Wilson,JohnD.(2005),Welfare-ImprovingCompetitionforMobileCapital,Journalof
-18.157(1),EconomicsUrban

Wilson,JohnD.andEckhardJaneba(2005),DecentralizationandInternationalTaxCom-
petition,JournalofPublicEconomics89,1211-1229.

Winner,Hannes(2005),HasTaxCompetitionEmergedinOECDCountries?Evidence
fromPanelData,InternationalTaxandPublicFinance12,667-687.

Wrede,Matthias(1996),VerticalandHorizontalTaxCompetition:WillUncoordinated
LeviathansEndUpontheWrongSideoftheLafferCurve?,Finanzarchiv53,461-479.

Wrede,Matthias(1997),TaxCompetitionandFederalism:TheUnderprovisionofLocal
PublicGoods,Finanzarchiv54(4),495-515.

Wrede,Matthias(2000),SharedTaxSourcesandPublicExpenditure,InternationalTax
andPublicFinance7,163-175.

Wrede,Matthias(2002),FiskalischeExternalit¨atenimf¨oderativenStaat,Beitr¨agezurFi-
nanzwissenschaft13,Mohr-Siebeck.

Wu,YonghongandRebeccaHendrick(2009),HorizontalandVerticalTaxCompetitionin
FloridaLocalGovernments,PublicFinanceReview37,289-311.

111

TillmannLisa

FederalTaxCompetition

Zax,JeffreyS.(1989),IsThereALeviathaninYourNeighborhood?,AmericanEconomic
60-567.579(3),Review

Zheng,Yongnian(2008),DeFactoFederalisminChina-ReformsandDynamicsofCentral-
LocalRelations,SeriesonContemporaryChina,Vol.7,NewJerseyetal.,World
ific.tScien

Zissimos,BenandMyrnaWooders(2005),RelaxingTaxCompetitionThroughPublicGood
Differentiation,WarwickEconomicResearchPaperNo.737.

Zodrow,GeorgeR.andPeterMieszkowski(1986),Pigou,Tiebout,PropertyTaxation,and
theUnderprovisionofLocalPublicGoods,JournalofUrbanEconomics19,356-370.

112

arungErkl¨heEidesstattlic

Ichversicherewahrheitsgem¨aß,dieDissertationbisaufdieinderAbhandlungangegebe-

nenHilfeselbst¨andigangefertigt,allebenutztenHilfmittelvollst¨andigundgenauangegeben

undgenaukenntlichgemachtzuhaben,wasausArbeitenanderer

unver¨andertodermit

19.01.2011Karlsruhe,

Lisa

Tillmann

anderungenAb¨

ernommenub¨

wurde.

und

aus

igenene

Ver¨offentlichungen