Labor Market Institutions and Wage and Inflation Dynamics

393

Economic AnAlysis & Policy, Vol. 40 no. 3, DEcEmBER 2010

Labor Market Institutions and Wage and Inflation Dynamics

Fatih Macit*

Department of Economics Beykent University

Istanbul, Turkey 34396 (Email: fatihmacit@beykent.edu.tr)

Abstract: ThispaperdevelopsaNewKeynesian(NK)modelthatincorporatesstandardsearchandmatchingstructurewithfiringcosts.Ianalyzehowlabormarketinstitutionsaffectthemacroeconomicdynamics,inparticular,wageandinflationdynamics.Iparticularlylookattwoimportantlabormarketinstitutionsnamelyunemploymentbenefitsandfiringcosts.Ifindthatincountrieswhereunemploymentbenefitsarehigherandtherearemorestrictemploymentprotectionlegislations,inflationandwagesbecomelessvolatileandmorepersistent.Ialsofindthattheleveloftheselabormarketinstitutionsaffecthowwagesandinflationrespondtoexogenousshocks, inparticular, toproductivityandmonetarypolicyshocks.Ifirstpresentsomeempiricalevidencethatshowsacross-countrylinkbetweenlabormarketinstitutionsandwagesandinflation.ThenIbuildadynamicstochasticgeneralequilibriummodelwhichprovidestheoreticalsupportforthisempiricalevidence.

I.INTroducTIoN

TherecentdevelopmentofmodelsthatcombinethetraditionalNewKeynesianmodelsandstandardsearchandmatchingmodelshavebeenquitesuccessfulinreplicatingthemainbusinesscycledynamicsthatstandardmodelsfailtoachieve.Forinstance,thesemodelsareabletoobtainlargeandpersistentresponsesofoutputtoexogenousshocksandrelativelysmoothbehaviorofwagesoverthecycle.Gertler,Sala,andTrigari(2008)findthatthesemodelsaccompaniedbystaggeredwagecontractingfitthedataroughlywell.InarelatedpaperMacit(2010)showsthatincorporatingon-the-jobsearchdoesthejobofstaggeredwagecontractingandachievesthesameresultswithfullyflexiblewages.Trigari(2006),KrauseandLubik(2006),andchristoffel,Kuester,andLinzert(2006)aresomerecentexamplesofthismodellingliteratureandtheyhavebeenquitesuccessfulinmatchingimportantbusinesscyclefacts.Besidesmatchingthebusinesscycledynamicstherehavealsobeenpapersthatattempttolookatoptimalmonetarypolicyinthesemodels.Faia(2006),Thomas(2008),andArseneauandchugh(2007)aresome

1 IwouldliketothankmyadvisorProfessorBehzaddibaforhisguidanceandsupportinthepreparationofthispaper.IwouldalsoliketothankProfessorJamesAlbrechtandProfessorSusanVromanfortheirveryvaluablesuggestionsandcomments.Allerrorsaremyown.

lABoR mARkEt institutions AnD WAgE AnD inflAtion DynAmic

394

recentexamplesthatlookatoptimalmonetarypolicywithsearchandmatchingfrictionsinthelabormarketinanotherwisestandardNKmodel.

Theaimofthispaperisdifferentfromtheoneslistedaboveasitdoesnotaimtomatchbusiness cycle dynamics or look at optimalmonetary policy. I investigatewhether labormarketinstitutionsaffectbusinesscycledynamics,inparticular,wageandinflationdynamics.Iparticularlyfocusontwoimportantlabormarketinstitutionsnamely,thebenefitreplacementrateandfiringcosts.Ichoosethesetwoinstitutionsastheysubstantiallyinfluencetheworker’sincentivetokeepajobandthefirm’sincentivetopreserveanexistingmatchwithaworker,respectively.Figure 1showstherelationshipbetweenthevolatilityofinflationandemploymentprotectionlegislationindexfortheoEcdeconomies.Figure 2showsthesamerelationshipforthevolatilityofrealwagesandFigure 3forthevolatilityofmarginalcost.1Thevolatilitiesofinflation,realwages,andmarginalcostarecalculatedfortheHPfiltereddata.ThedataforemploymentprotectionlegislationindexistakenfromNickell(2006)andthedataforinflation,realwages,andmarginalcostisobtainedfromoEcddatabase.Themarginalcostismeasuredastheunitlaborcost.Asonecanconsidertheemploymentprotectionlegislationasaproxyforthefiringcoststhegraphsshowthatthereisanegativerelationshipbetweenvolatilitiesofthesevariablesandtheleveloffiringcosts.TheseresultsareonlysuggestiveandinordertobeabletounderstandtheeffectoffiringcostsoninflationandwagedynamicsIbuildatheoreticalmodeltoisolatetheeffectoffiringcosts.ForthispurposeIdevelopadynamicstochastocgeneralequilibriummodelinwhichthefirmpaysafixedfiringcostwhenanexistingemploymentrelationshipbreaksup.Ifindthathigherlevelsoffiringcostsgeneratelessvolatileandmorepersistentmovementsininflationandwagesinresponsetomonetarypolicyandproductivityshocks.Theresultsalsoshowthatwhenfiringcostsarelowerinflationandwagesshowalargerresponseonimpactbuttheyadjustmorequickly.

Firingcostshaveespeciallybeenanimportantareaofstudyforthestandardsearchandmatchingmodels.Themain focus of these studies has been to explain thedifferences inunemploymentratesbetweenEuropeanunioncountriesandtheu.S.Therehavealsobeenpapersthattrytoestablishalinkbetweentheleveloffiringcostsandbusinesscycledynamics.Veraciarto(2004)buildsarealbusinesscyclemodelwithestablishmentleveldynamicsandfindsthatwhenfiringcostsarehigheremploymentbecomeslessvariableandmorepersistent.Thomas(2006)findsthatfiringcostsreducethevolatilityofbusinesscyclefluctuations.Healsoshowsthatintroducingfiringcostsintothestandardsearchandmatchingmodelscanbearemedyforthefailureofthestandardmodelingeneratinganegativecorrelationbetweenthecyclicalcomponentsofunemploymentandvacancies.Thispaperdiffersfromthesepapersinthatitincorporatesfiringcostsintoastickypricedynamicstochasticgeneralequilibriummodelandtothebestofmyknowledgeitisthefirstpaperthatinvestigateshowfiringcostsaffectinflationdynamics.Asfiringcostsaffectthesurplusforanemploymentrelationshiponemayexpecttheleveloffiringcosttoinfluencewagedynamics.TheinterestforinflationcomesfromthefactthatwagesaffectthelevelofrealmarginalcostandthedeviationofrealmarginalcostsfromtheirnaturallevelsentersintoaNKPhillipscurve.Sotheeffectoffiringcostson inflation isexpected tooccur through its influenceonmarginalcosts. I find that

1 IpresenttheresultsforthemarginalcostasthelevelofmarginalcostisanimportantdeterminantofinflationinNewKeynesianPhillipscurve.

fAtih mAcit

395

higherfiringcostsmakeinflationlessvolatileandmorepersistent.Theleveloffiringcostalsoaffectsthepatternthatinflationshowsinresponsetoexogenousshocks,inparticular,toproductivityandmonetarypolicyshocks.

BesidestheeffectoffiringcostsonwagesandinflationdynamicsIalsoinvestigatehowthesevariablesareaffectedbythelevelofunemploymentbenefits.InthisregardanimportantpreviousstudyhasbeendonebycampolmiandFaia(2010).Theybuildadynamicstochasticgeneralequilibriumopeneconomymodelforamonetaryunion.TheythenaskthequestionwhethertheinflationdifferentialsbetweenEuropeanunioncountriescanbeexplainedbythedifferencesinlabormarketinstitutions.Theyparticularlylookatthelevelofunemploymentbenefits inEuropeanunioncountriesmeasuredbybenefit replacement rateand find thathigherlevelsofunemploymentbenefitsareassociatedwithlowervolatiliesofinflation,wages,andrealmarginalcost.InthecurrentmodelIlookattheeffectofunemploymentbenefitsonwagesandinflationfromtheperspectiveofaclosedeconomyasopposedtotheopeneconomymodelofcampolmiandFaia(2007)andIalsoinvestigatehowthesevariablesrespondtoproductivityandmonetarypolicyshocks.

Therestofthepaperisorganizedasfollows.Inthenextsection,Ibuildacloseddynamicstochasticgeneralequilibriummodelinordertobeabletoisolatetheeffectoffiringcostoninflationandwagedynamics.SectionIIIdealswiththecalibrationofthemodelandSectionIVpresentstheresults.SectionVconcludes.

II.ModEL

ThemodelthatIadoptisveryclosetothatdevelopedbyTrigari(2006).ThedifferencebetweenthismodelandtheoneproposedbyTrigari(2006)isthatfirmsincurfiringcostswhenanexistingemploymentrelationshipbreaksup.Therearefouragentsintheeconomy:workers,intermediategoodfirms,retailfirms,andamonetaryauthority.Ifirstcharacterizetheproblemoftherepresentativehouseholdandthenthelaborandproductmarkets.

2.1 Households

Eachhouseholdconsistsofacontinuumofmemberswithnamesontheunitinterval.Eachmemberhasthefollowingutilityfunction:

€

u(ct ,ct−1) − g(ht ) = log(ct −ξct−1) −Ψh

ht1+φ

1+φ  (1)

wherectistheconsumptionofthefinalgood,htisthehoursworkedandIallowforhabitpersistence.2Therepresentativehouseholdmaximizeslifetimeutilitybychoosingconsumption,ct,andbondholdings,Bt,subjecttothebudgetconstraint.Thelifetimeutilityofthehouseholdisgivenby:

€

Et

s=0

∞

∑β s[u(c

t+s,ct+s−1) −Gt+s]  (2)

whereβε(0,1)istheintertemporaldiscountfactorandthevariableGt isthefamily’sdisutility

lABoR mARkEt institutions AnD WAgE AnD inflAtion DynAmic

396

fromsupplyinghoursofwork.3Idonotwritethisfunctionexplicitlyashoursworkedisnotachoicevariableforthehousehold.Ineachperiodhouseholdsaresubjecttothefollowingbudgetconstraint:

€

ct +Bt

ptrt= dt +

Bt−1

pt  (3)

wherept is theaggregatepricelevelandrt is thegrossnominal interestrateonthebond.FollowingMerz(1995)andAndolfatto(1996),Iassumethatthereisperfectconsumptionrisksharingbetweenemployedandunemployedfamilymembers.Thevariabledtincludeswageincomeearnedbyemployedmembers,unemploymentbenefitsearnedbyunemployedmembers,theshareofprofitsfromretailers,netofagovernmentlump-sumtaxusedtofinanceunemploymentbenefits.

Therepresentativehouseholdmaximizes(2)subjecttotheperiodbudgetconstraintbychoosingconsumptionandbondholdings.Thefirstorderconditionsforaninteriorsolutionareasfollows:

€

λt=

1

ct−ξc

t−1

−Etβξ

1

ct+1 −ξct

  (4)

€

λt= E

t[βr

t

λt+1

πt+1

]  (5)

whereλtistheLagrangemultiplierassociatedwiththebudgetconstraintandπt+1isthegrossinflationrate.

2.2 Firms and the Labor Market

Therearetwotypesoffirmsinthemodel:intermediategoodsfirmsandretailfirms.Intermediategoodsfirmscarryouttheactualproductionusinglaborastheonlyfactorofproduction.Thesefirmsaresubjecttosearchandmatchingfrictionsinthelabormarketandselltheiroutputinaperfectlycompetitivemarket.retailfirmsfacemonopolisticcompetitionandaresubjecttonominalrigiditiesinthepricesettingdecision.

2.2.1TheLaborMarket

Thematchingprocessbetweentheworkersandthefirmsischaracterizedbyamatchingfunctionwhichgivesthenumberofmatchesinagivenperiodbetweenjobseekersandvacancies.Thetotalnumberofperperiodnewmatchesisgivenbythefollowingfunction:

€

mt=Mu

t

µvt

1−µ  (6)

wherevtisthemeasureofvacanciespostedbyfirmsandutisthemeasureofunemployedworkerssearchingforajob.Iassumeaconstantreturntoscalematchingfunctionwhichischaracterizedby

€

mt = Mutµvt

1−µ .TheconstantMreflectstheefficiencyofthematchingprocess.

3 Assumingthatthereisperfectconsumptionrisksharingbetweenemployedandunemployedfamilymembersallowsonetoaggregatetheutilityfunctionforthefamily.

fAtih mAcit

397

Icanderivetheprobabilitiesofmakingamatchforfirmsandsearchingworkersusingthematchingfunction.Idefine

€

θt = vt /ut asthemeasureoflabormarkettightnessinthematchingmarket.Afirmfillsavacancywithprobabilityqt ≡ mt /vt andaworkersearchingforajobmakesamatchwithprobabilityst ≡ mt/ut.Theseprobabilitiesaregivenby:

€

qt =Mθt−µ  (7)

€

st=Mθ

t

1−µ  (8)

Iassumethatmatchesbreakupexogenouslywithprobabilityρ.Giventhistheevolutionofemploymentisasfollows:

€

nt+1 = (1− ρ)(nt +m

t)   (9)

wherent isthenumberofpeopleemployedinperiodt.Thisequationimpliesthatnewmatchedcannotenter into theproductionfunctionif therelationshipisbrokenuprightafterbeingnegotiated.Asthemeasureoflaborisequaltoone,thenumberofunemployedpeople,ut,isgivenby:

ut =1– nt (10)

2.2.2ValueFunctions

TheproblemoftheworkersandthefirmsischaracterizedbyBellmanequations.Ifirststartwiththevaluefunctionsforthefirms.ThevalueofacontinuingemploymentrelationshipforafirmisdenotedasJt andthethevalueofanewemplomentrelationshipdenotedas

€

Jtn

andaregivenby:

€

Jt = xtat f (ht ) −wtht +Etβt,t+1[(1− ρ)Jt+1 + ρ(Vt −F)]  (11)

€

Jtn = xtat f (ht ) − wt

nht + Etβt,t +1[(1− ρ)Jt +1 + ρ(Vt − F)] (12)

wherewt isthewagepaidforanexistingemploymentrelationshipand

€

wtn istheonefora

newemploymentrelationship,xt isthepriceoftheintermediategoodandatistheproductivityshock.Thevalueof a continuingmatch is equal to the current profitswhich is givenby

€

xtat f (ht ) − wtht plusthecontinuationvalue.4Withprobabilityl–ρthematchcontinuesandthefirmenjoystheexpectedvalueofthejob.Withprobabilityρ thematchbreaksupnextperiod.InthiscasethefirmenjoysthevalueofavacancybutatthesametimeincursthefixedfiringcostgivenbyF.5Thefuturevalueofthejobisdiscountedbythediscountfactorβt+1whichisgivenbyβλt + 1 / λt.

Thevalueofavacancy,Vt,isasfollows:

€

Vt =−κ +Etβt,t+1[qt (1− ρ)Jt+1 + (1− qt )Vt+1]   (13)

4 Theproductionfunctionf(h)isassumedtobeadecreasingreturnstoscaleproductionfunction.5 Firingcostsarenotmodelledintheformofseverancepayment.

lABoR mARkEt institutions AnD WAgE AnD inflAtion DynAmic

398

whereκistheflowcostofpostingavacancy.Withprobability

€

qt (1− ρ) avacancywillbefillednextperiodandwillactuallybeproducing.Assumingfreeentryofvacancieswilldrivethevalueofavacancytozeroinordertoeliminateanyarbitrageopportunity.Thiswillgivethefollowingequilibriumcondition:

€

κ

qt= Etβt,t+1[(1− ρ)Jt+1]  

(14)

Themeaningofthisequilibriumconditionisthattheexpectedcostofavacancyisequaltotheexpectedbenefitreceivedfromfillingthatvacancy.

NowletWt andUt bethevalueofemploymentandunemploymentrespectivelyfromtheperspectiveofaworker.Valueofemploymentforaworkerisgivenby:

€

Wt = wtht −g(ht )

λt+Etβt,t+1[(1− ρ)Wt+1 + ρUt+1]  

(15)

Theterm

€

g(ht )λt

isthedisutilityfromsupplyinghoursofworkanditisexpressedintermsof current consumption inorder topreserveconsistencybetween the terms.Thevalueofunemploymentisgivenbythefollowingvaluefunction:

€

Ut= b+E

tβt,t+1[st (1− ρ)Wt+1 + (1− s

t(1− ρ))U

t+1]   (16)

whereb isthevalueofunemploymentbenefitsreceivedbytheworkerwhichisfinancedbyalump-sumgovernmenttax.

2.2.3WageBargaining

Iassumethatwagesaredeterminedbysurplussplittingassumption.However,thepresenceoffiringcostscreatesdifferencebetweenthesurplusforanewemploymentrelationshipandthesurplusforanexistingemploymentrelationshipfromtheperspectiveofafirm.Whenaworkerandafirmmeetforthefirsttimeforawagebargainthefirmdoesnotneedtopayafiringcostifthematchisnotsuccessful.ontheotherhand,foranexistingemploymentrelationshipifthematchbreaksupthefirmincursafixedfiringcost.Sowhencalculatingthesurplusforafirmthathasanexistingmatchoneneedstotakeintoaccountthefiringcoststhatareavoidedifthematchcontinues.Foranewmatchtheoutcomeofthesurplussplittingassumptionmaximizestheproduct:

€

(Wt−U

t)η(J

t

n −Vt)1−η  (17)

wherethefirsttermisthesurplusfortheworkerandthesecondtermisthesurplusforthefirm.Theparameterη reflectsthebargainingpoweroftheworker.Firmsandworkersmaximizethejointsurplusofthematch.Thewagethatmaximizesthejointsurplusgivesthefollowingfirstordercondition:

€

ηJt

n= (1−η)(W

t−U

t)   (18)

fAtih mAcit

399

AsImentionedabovethesurplusforafirmfromanexistingemploymentrelationshipwilldifferfromtheoneforanewmatchduetothefiringcostthatthefirmincurswhenthematchbreaks up. For an existing employment relationship the outcome of the surplus splittingassumptionmaximizesthefollowingproduct:

€

(Wt−U

t)η(J

t− (V

t−F))1−η   (19)

ThesecondtermreflectsthefactthatwhenthematchendsupthefirmendsupwiththevalueofvacancyandpaysthefixedfiringcostsF.Thewagethatmaximizesthejointsurplusgivesthefollowingoptimalitycondition:

€

η(Jt+F) = (1−η)(W

t−U

t)   (20)

usingthesebargainingequationsandthejobcreationconditiongivesusthewageforanewmatchandthewageforanexistingemploymentrelationship:

€

wt =ηxtat f (ht )

ht+κθtht

+(1− βρ)F

ht]+ (1−η)[

g(ht )

htλt+b

ht]   (21)

€

wtn=η

xtat f (ht )

ht+κθtht

−βρF

ht]+ (1−η)[

g(ht )

htλt+b

ht]  (22)

where

€

wtn referstothewageforaworkerthathasmadeanewmatchwiththefirm.Thewage

equationsshowthatworkersthatalreadyhaveajobbenefitfromahigherfiringcostwhereasworkersthatdonothaveajobareharmed.ThisisalsoconsistentwiththeempiricalevidencereportedbyoEcdmentioningthatworkersthatalreadyhaveajobarefavoredbyhigherfiringcostswhereasthehigherfiringcostsmakeitmoredifficultforoutsiderstofindajobandreducestheirwages.

ForthedeterminationofhoursworkedIassumethatworkersandfirmsjointlydeterminehours.Trigari(2006)definesthisbargainingprocedureasefficientbargaining.Theoptimalityconditionforthedeterminationofhoursworkedisgivenby:

€

xtat fh (ht ) =g'(ht )

λt  (23)

where

€

at fh (ht ) referstomarginalproductofhoursworkedand

€

g' (ht )λ

denotesmarginalrateofsubstitution.

2.3 Retail Firms

There is a continuum of retails firms on the unit interval indexed by j operating in amonopolisticallycompetitivemarket.retailfirmstransformtheintermediategoodswithatechnologyandresellthemtothehouseholdsasafinalconsumptiongood.definingyjt astheoutputproducedbyretailfirmj,finalgoods,denotedbyyt, aregivenbythefollowingcombinationofindividualretailgoods:

lABoR mARkEt institutions AnD WAgE AnD inflAtion DynAmic

400

€

yt = 0

1

∫ y jtε−1

ε dj

ε

ε−1

  (24)

whereεistheelasticityofsubstitutionacrossthedifferentiatedretailgoods.usingthisconditioneachretailfirmhasthefollowingdemandfunctiongivenby:

€

y jt = (p jt

pt)−εyt   (25)

wherepjt isthepricechargedbyretailfirmj andpt istheaggregatepriceindexgivenby:

€

pt = 0

1

∫ p jt1−εdj[ ]

1

1−ε  (26)

retailfirmsaresubjecttocalvo(1983)typenominalpricerigidity.Eachperiodaretailfirmisallowedtoadjustitspricewithprobability1–ϕ andthisprobabilityisindependentofhistoryofpriceadjustments.Theretailfirmsthathavethechancetosettheirpricewillchoosetheirpriceinordertomaximizetheirexpectedfuturediscountedprofitssubjecttothedemandfortheirgoodandtotheconditionthatthepricethattheysetatdatet prevailsatdatet + k withprobabilityϕt + k.Thefirstorderconditionfortheretailfirm’sprofitmaximizationproblemisgivenby:

€

p jt = ςEtk=0

∞

∑ψ t,t +kmct +kn (27)

where

€

ς = εε −1istheflexiblepricemarkupandtheterm

€

mctn = ptmct isthenominalmarginal

costatdatet. Therelevantweights

€

ψ t,t +k arewrittenas:

€

ψ t,t+k =ϕ kβt,t+kRjt,t+k

Etk=0

∞

∑ϕ kβt,t+kRjt,t+k

  (28)

whereRjt, t + k istherevenueattimet + k giventhatthelastpriceadjustmentisdoneinperiodt

2.4 Monetary Authority

Iassumethattheshort-terminterestrateisthepolicyinstrumentofthemonetarypolicyandthemoneysupplyisadjustedaccordingly.IsetaTaylortyperuleforthenominalinterestratergivenby:

€

rt = β−(1−τ )

(rt−1)τEt (π t+1)

απ(1−τ )

(yt − y)αy(1−τ )

eεtr

  (29)

wheretheparameterτmeasuresthedegreeofinterestratesmoothing.Thenominalinterestraterespondstoinflationandthedeviationofoutputfromitssteadystatevalue.Theresponsecoefficientsare

€

απ and

€

αy respectively.Thelastterminthepolicyfunction,

€

εtr ,isthei.i.d

monetarypolicyshock.

fAtih mAcit

401

Iclosethemodelbywritinganaggregateresourceconstraint.Thatisgivenby:

€

yt = ct +κvt + ρFnt−1 (30)

Thisshowsthatoutputgoestoconsumptionorvacancypostingcostorfiringcost.

III.cALIBrATIoN

InordertoinvestigatethebusinesscyclepropertiesofthetheoreticalmodelIassignnumericalvaluestothestructuralparameters.Isetthediscountfactor,β,to0.99whichimpliesa4%annualinterestrate.Thevalueofthehabitpersistenceparameterissetas0.6.Isetthelaborsupplyelasticityas1/3whichisconsistentwiththemicroeconomicstudiesthatestimatethelaborsupplyelasticitycloseto0andnothigherthan0.5.

FollowingtheliteratureIchooseacobb-douglasformforthematchingfunction.Thematchingfunctionisgivenby

€

m = Mv1−µuµ .FollowingKrauseandLubik(2006)Isetthelevelparameter,M, as0.6.InaccordancewithTrigari(2006)Isettheelasticityparameter,μ, equalto0.5.6Theparameterfortheseparationrateissetas0.1.Therearesomeempiricalestimatesintheliteraturefortheu.S.separationrate.Hall(1995)reportsthisratetobebetween8to10percent.davis,Haltiwanger,andSchuh(1996)estimatestheu.S.separationrateas8%.ThevaluethatIsetisconsistentwiththeempiricalestimates.Forthevacancypostingcost,followingKrauseandLubik(2006)Icalibrateitas0.16.Thatimpliesthatatthesteadystateabout4%oftheoutputgoestovacancypostingcost.FollowingKrauseandLubik(2006)thevaluefromunemploymentgivenbyz issetas0.4andtherelativebargainingpoweroftheworkerissetas0.5.ThisvalueisthesameastheelasticityparameterofthematchingfunctionandsosatisfiestheHosios(1990)condition.

Inowcalibratethestructuralparametersfortheretailsector.Isettheprobabilitythatafirmisnotallowedtochangeitspriceinagivenperiodequalto0.67whichimpliespricesonaveragearefixedbythreequarters.Isettheflexiblepricemarkupas10%whichimplies�

Lastly,Icalibratetheparametersfortheexogenousshocksandthemonetarypolicyrule.IassumethatthelogarithmoftheaggregateproductivityshockfollowsanAr(1)processwithacoefficient0.923whichisusedbycanzoneri,cumby,anddiba(2007).Againfollowingcanzoneri,cumby,anddiba(2007)theinterestratesmoothingparameterissetas0.824andthecoefficientsoninflationandoutputaresetequalto2.02and0.184respectively.

IV.rESuLTS

InthissectionIpresentthemodelresultsintermsofinvestigatinghowthelevelofthebenefitreplacementrateandfiringcostsaffectthewageandinflationdynamics.

6 Fortheestimationofthematchingfunction,PetrongoloandPissarides(2001)reportthisparametertobebetween0.5and0.7intheirsurveyoftheliterature.

lABoR mARkEt institutions AnD WAgE AnD inflAtion DynAmic

402

4.1 Volatility and Persistence

Ifirstpresentthemodelresultsforthevolatilityofinflationandwagesfordifferentlevelsoffiringcost.Table 1showsthestandarddeviationsofthesevariablesforthreedifferentfiringcost levels.Thefirstcolumnisfor thecasewheretherearenofiringcosts.InthesecondcolumntheF parameterissettogenerateafiringcostthatisequaltoonewageatthesteadystateandinthethirdcolumnitissettogenerateafiringcostequaltotwowagesatthesteadystate.Thestandarddeviationsaremeasuredrelativetothatofoutput.Asthetableshowsasfiringcostsgoupthevolatilityofinflationandwagesgodown.Thisisconsistentwiththeempiricalresults thatarepresentedabove.Itwouldalsobenice tocompare theordersofmagnitudeonFigures 1and2toTable 1bygivingasenseoftwocountriesinthedata.InthisregardIchooseGermanyandNetherlands.IntermsoftheemploymentprotectionlegislationindexthevalueforNetherlandsishigherthanthatofGermany.InthedataonecanseethatthevolatilityofinflationforNetherlandsisabout30%lowerthanthatofGermany.IntermsofthevolatilityofwagesthenumberforNetherlandsisagainlowerthanthevolatilityofwagesinGermany.Animportantpointtonotehereisthatthevolatilityofwagesishigherthanwhatmightbeobservedinactualdata.Thereasonforthisisthatthemodelabstractsfrommanyrealrigiditiesincludingstaggeredwagecontractingoron-the-jobsearchmechanism.

Next,Iwilllookatcoefficientsofautocorrelationforinflationandrealwagesfordifferentlevelsoffiringcostsandthebenefitreplacementrateinordertoassesshowtheselabormarketinstitutionsaffectthepersistenceinthesevariables.Table 2showstheresultsfordifferentlevelsoffiringcosts.Ilookatcoefficientsofautocorrelationuptofivelagsfortwodifferentfiringcostlevels.Theresultsshowthatwhenfiringcostsgoupbothinflationandrealwagesbecomemorepersistent.Intermsofeconomicinterpretationthisresultisnotsurprising.Iffiringcostsarehigherthaninresponsetoexogenousshocksfirmshaveverylittleroomtomoveintermsofadjustingtheirpricesandinturntheirwages.Thismakesbothvariablesmorepersistentandlessvolatile.Atthispointitwillbeillustrativetogiveanexamplefromacoupleofcountriesinthedatathatmaysupportthemodelresultsintermsofautocorrelation.TobeconsistentwiththemodeltheautocorrelationsarecalculatedfortheH-Pfiltereddata.IpresenttheresultsuptofivelagsandforfourdifferentcountriesnamelyNetherlands,Germany,France,anddenmark.Table 3showstheresultsforrealwagesandTable 4showstheresultsforinflation.Itisseenfromtheresultsthatcountriesthataresubjecttomorerigidemploymentprotectionlegislationtendtohavemorepersistentinflationandwages.Animportantthingtonotehereisthattheobservedautocorrelationsaremuchhigherthantheonesobtainedfromthemodel.Thereasonforthisisthat,asImentionedbefore,themodelabstractsfrommanyrealrigiditiesthatmayhelpinobtainingtheobservedpatterninwages.Incorporatingstaggeredwagecontractingoron-the-jobsearchmechanismmaybearemedyinthisregard.

BesidesthefiringcostsIalsoinvestigatethepersistenceininflationandwagesfordifferentlevelsofthebenefitreplacementrate.Ilookfortwodiffferentlevelsofthebenefitreplacementrate.In thefirstoneIset theb parameterwhichgivesmeat thesteadystate thevalueofunemploymentbenefitstothewageis0.56.Inthesecondonetheparameterissetalowervaluethatatthesteadystatethisratiois0.19.BothofthesearewithintherangeofbenefitreplacementrateobservedinoEcdcountriesreportedbyNickell(2006).Table 5showstheresultsforinflationandwages.Itisseenthatespeciallyforinflationthepersistencebecomes

fAtih mAcit

403

moreobviouswhenthebenefitreplacementrategoesup.Whenthebenefitreplacementrateishigherworkersarenotwillingtotakelargewagecutsinresponsetoexogenousshocks.Thismakesbothwagesandinflationmorepersistentinresponsetotheseshocks.

4.2 Impulse Responses

Anotherimportantissueaboutthelabormarketinstitutionsiswhetherdifferencesintheseaffecthowmacrovariablesrespondtoexogenousshocks.InthissectionIinvestigatehowtheleveloffiringcostsandthebenefitreplacementrateaffectthepatternthatinflationandwages show in response toexogenous shocks inparticular toproductivityandmonetarypolicyshock.Figure 4showstheresponseofinflationtoapositiveproductivityshockfortwodifferentfiringcostlevels.onecanseethatwhenfiringcostsarelower,onimpactinflationshowsalargerresponse.Inthiscaseonimpactinflationgoesdownby0.55%whereasinthecasewherefiringcostsarehigherinflationgoesdownbyabout0.45%.However,itisseenthatwhenfiringcostsarelowerinflationadjustsmorequicklyinresponsetotheproductivityshock.Thisisactuallyconsistentwiththepreviousresultsthatarepresented.Thatiswhenfiringcostsarehigher,inresponsetotheexogenousshocksinflationbecomesmorepersistent.Figure 5showsthesameresultsforamonetarypolicyshockthatraisestheinterestrates.onecanseethesamepatternintheresponseofinflationthatisseeninproductivityshock.oncetheeconomyishitbyamonetarypolicyshockonimpactinflationgoesdownbyabout0.4%whenthefiringcostsareloweranditgoesdownby0.25%whenfiringcostsarehigher.However,again,inflationdoesnotshowapersistentresponsewhenfiringcostsareloweranditadjustsmorequicklycomparedtothecasewherefiringcostsarehigher.

Thepatternthatisseenininflationcanalsobeseenintheresponseofwages.Figure 6shows the response ofwages in response to a positive productivity shock.one can seethatonimpactlowerfiringcostscreatealargerresponse.However,thispatterndoesnotpersistandadjustveryquickly.Lowerfiringcostsgeneratemoreflexibilityforthefirmsinadjustingtheirwagesandcreateslargerandquickeradjustmentsinwagesinresponsetotheproductivityshock.

Besidestheleveloffiringcostthelevelofbenefitreplacementrateisalsoeffectivefortheresponseofinflationandwagestoproductivityandmonetarypolicyshocks.Figure 7showstheresponseofinflationtoapositiveproductivityshockfortwodifferentbenefitreplacementratesthathavebeenusedinprevioussection.Thesamepatternisobservedthathasbeenseenforfiringcosts.Whentherearelessgenerousunemploymentbenefitsonimpact,inflationshowsa larger response to thepositiveproductivityshockbut itadjusts faster than is thecasewhentherearemoregenerousunemploymentbenefits.Figure 8showsthesameresultsnowforamonetarypolicyshockandexactlythesamepatternisobserved.Lowerbenefitreplacementratesmakeinflationlesspersistentandthiscanbeobservedintheresponseofinflationtoaproductivityandmonetarypolicyshock.Figure 9showstheresponseofwagestoamonetarypolicyshockagainfortwodifferentbenefitreplacementratelevels.Whenthebenefitreplacementrateislower,wagesshowaslightlylargerresponseatthebeginningbutthenadjustveryquickly.Itisimportanttonotethattheleveloffiringcostsseemstoberelativelymoreinfluentialonimpulseresponsescomparedtotheeffectofthebenefitreplacementrate.

lABoR mARkEt institutions AnD WAgE AnD inflAtion DynAmic

404

V.coNcLuSIoN

InthispaperIdevelopadynamicstochasticgeneralequilibriummodelinanotherwisestandardNKmodel.Thispaperdiffersfromthepreviousliteraturethatusesthesamemodellingstrategywiththerespecttotheincorporationoffiringcosts.Thatiswhenanexistingemploymentrelationshipbreaksupthefirmincursafixedfiringcost.InthiscontextIinvestigatehowtheleveloffiringcostandunemploymentbenefitaffectthebusinesscycledynamics,particularly,wageandinflationdynamics.

Ibuildadynamicstochasticgeneralequilibriummodelforaclosedeconomyandinvestigatehowinflationandwagesreactfordifferentlevelsoffiringcostsandbenefitreplacementrate.Ifindthathigherlevelsoffiringcostgeneratelessvolatileandmorepersistentpatterninwagesandinflationinresponsetoproductivityandmonetarypolicyshocks.Thesamepatternholdsaswellforbenefitreplacementrates.Ialsofindthatwhenfiringcostsarelower,inresponsetotheproductivityandmonetarypolicyshocks,inflationandwagesshowlargerresponsesonimpactbuttheyadjustquickly.

Animportantextensionofthisworkmaybetoinvestigatetheroleofotherimportantlabormarketinstitutionsforbusinesscycledynamics.Amongthesedurationofbenefits,uniondensityandcoverage,andbargainingcoordinationandcentralizationcanbelisted.Theroleoftheseandotherlabormarketinstitutionscanbeempiricallyinvestigatedintermsoftheireffectonbusinesscycledynamics.developingtheoreticalmodelsthatmayprovidesupportfortheempiricalevidencemayalsobeanimportantfutureworkinthisdirection.

rEFErENcES

Andolfatto,d.(1996).BusinesscyclesandLaborMarketSearch,American Economic Review.861:112-132.

Arseneau,d.M. andK.c. Sanjay (2007). optimal Fiscal andMonetary Policywith costlyWageBargaining,Journal of Monetary Economics.55:1401-1414.

calvo,G.(1983).StaggeredPricesinautilityMaximizingFramework,Journal of Monetary Economics.12:383-398.

campolmi,A.andE.Faia(2010).LaborMarketInstitutionsandInflationVolatilityintheEuroArea,Journal of Economic Dynamics and Control.July2010.

canzoneri,B.M.,r,E.cumby,andB.diba(2007).ThecostofNominalrigidityinNNSModels,Journal of Money, Credit and Banking.39:1563–1586.

Kai, c and T. Linzert (2005). The role of realWage rigidity and Labor Market Frictions forunemploymentandInflationdynamics,EcBWorkingPaperNo.556.

Steven,d.,J.Haltivanger,andS.Schuh(1996).Job Creation and Destruction.cambridge,MA:TheMITPress.

Faia,E.(2006).optimalMonetaryPolicyruleswithLaborMarketFrictions,EcBWorkingPaper698.Fuhrer,J.(2000).HabitFormationinconsumptionandItsImplicationforMonetary-PolicyModels,

American Economic Review. 90:367-390.Gertler,M.,L.Sala,andA.Trigari(2008).AnEstimatedMonetarydSGEModelwithunemployment

andStaggeredNominalWageBargaining,Journal of Money, Credit and Banking.40:1713-1764.Macit,F.(2010).MonetaryPolicyandProductivityShockswithEndogenousHoursandon-the-Job

Search,Journal of Economic and Social Research.forthcoming.Merz,M. (1995).Search in theLaborMarketand therealBusinesscycle ,Journal of Monetary

Economics. 36:269-300.oEcd2008.Economicoutlookdatabase.

fAtih mAcit

405

Nickell,William(2006).ThecEP-oEcdInstitutionsdataSet,cEPdiscussionPaperNo759.Petrongolo,B.andc.Pissarides(2001).LookingintotheBlackBox:ASurveyoftheMatchingFunction,

Journal of Economic Literature.39:390-431.Thomas,c.(2006).Firingcosts,labormarketsearchandthebusinesscycle,LondonSchoolofEconomics.Thomas,c.(2008).Searchandmatchingfrictionsandoptimalmonetarypolicy,Journal of Monetary

Economics.55:936-956.Trigari,A.(2006).TheroleofSearchFrictionsandBargainingforInflationdynamics,IGIErWorking

PaperNo.304.Veracierto,M.(2004).FiringcostsandBusinesscycleFluctuations,WorkingPaper,Federalreserve

Bankofchicago.

TABLES

Table 1:StandarddeviationofVariablesfordifferentFiringcostLevels

F=0 F=w F=2wInflation 1.11 0.55 0.45realWage 2.30 1.58 1.21

Table 2:coefficientsofAutocorrelationforInflationandrealWages

1 2 3 4 5InflationwithF=w 0.4817 0.2193 0.0626 -0.0370 -0.0997InflationwithF=2w 0.5868 0.3241 0.1418 0.0125 -0.0777realWagewithF=w 0.3979 0.1564 0.0292 -0.0482 -0.0963realWagewithF=2w 0.5135 0.2794 0.1298 0.0215 -0.0581

Table 3:coefficientsofAutocorrelationforrealWages

1 2 3 4 5 EPLNetherlands 0.7714 0.6170 0.4918 0.4297 0.2142 3.07Germany 0.5973 0.3748 0.1530 -0.0068 -0.2758 2.66France 0.7257 0.5771 0.4418 0.3208 0.1143 2.37denmark 0.4756 0.2082 0.0876 0.1188 -0.2053 1.49

Table 4:coefficientsofAutocorrelationforInflation

1 2 3 4 5 EPLNetherlands 0.7943 0.5020 0.1955 -0.0977 -0.2634 3.07Germany 0.7621 0.4930 0.2660 0.0308 -0.0667 2.66France 0.6482 0.4460 0.3452 -0.0276 -0.1141 2.37denmark 0.6305 0.2939 -0.0294 -0.2077 -0.3028 1.49

lABoR mARkEt institutions AnD WAgE AnD inflAtion DynAmic

406

Table 5:coefficientsofAutocorrelationforInflationandrealWages

1 2 3 4 5

Inflationwith

bw = 0.56 0.3788 0.1169 -0.0169 -0.0878 -0.1225

Inflationwith

bw = 0.19 0.3226 0.0617 -0.0727 -0.1237 -0.1350

realWagewith

bw = 0.56 0.3589 0.1158 -0.0085 -0.0785 -0.1158

realWagewith

bw = 0.19 0.3658 0.1045 -0.0480 -0.1165 -0.1404

FIGurES

Figure 1:TherelationshipBetweentheVolatilityofInflationandEmploymentProtection

Inflation with

0.56=wb

0.3788 0.1169 -0.0169 -0.0878 -0.1225

Inflation with

0.19=wb

0.3226 0.0617 -0.0727 -0.1237 -0.1350

Real Wage with

0.56=wb

0.3589 0.1158 -0.0085 -0.0785 -0.1158

Real Wage with

0.19=wb

0.3658 0.1045 -0.0480 -0.1165 -0.1404

0 0.5 1 1.5 2 2.5 3 3.5 4 4.50.2

0.4

0.6

0.8

1

1.2

1.4

1.6

Employment Protection Legislation Index

The S

tandar

d Devi

ation o

f Infla

tion

data 1 linear

Figure 1: The relationship between the volatility of inflation and employment

StandarddeviationofInflation

EmploymentProtectionLegislationIndex

fAtih mAcit

407

Figure 2:TherelationshipBetweentheVolatilityrealWagesandEmploymentProtection

Figure 3:TherelationshipBetweentheVolatilityMarginalcostandEmploymentProtection

0 0.5 1 1.5 2 2.5 3 3.50.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

2.2

2.4

Employment Protection Index

Standa

rd Devia

tion of W

ages

data points linear fit

Figure 2: The relationship between the volatility real wages and

0.5 1 1.5 2 2.5 3 3.5

0.8

1

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

Employment Protection Legislation Index

Standa

rd Devi

ation o

f Marg

inal C

ost

data points linear

Figure 3: The relationship between the volatility marginal cost and employment protection

-0.006

-0.005

-0.004

-0.003

-0.002

-0.001

01 3 5 7 9 11 13 15 17 19 21 23

Lower firing costHigher firing cost

Figure 4: Impulse response of inflation for a positive productivity shock

StandarddeviationofWages

StandarddeviationofM

arginalc

ost

EmploymentProtectionIndex

EmploymentProtectionLegislationIndex

lABoR mARkEt institutions AnD WAgE AnD inflAtion DynAmic

408

Figure 5:ImpulseresponseofInflationforaMonetaryPolicyShock

Figure 4:ImpulseresponseofInflationforaPositiveProductivityShock

0.5 1 1.5 2 2.5 3 3.5

0.8

1

1.2

1.4

1.6

1.8

2

2.2

2.4

2.6

Employment Protection Legislation Index

Standa

rd Devi

ation o

f Marg

inal C

ost

data points linear

Figure 3: The relationship between the volatility marginal cost and employment protection

-0.006

-0.005

-0.004

-0.003

-0.002

-0.001

01 3 5 7 9 11 13 15 17 19 21 23

Lower firing costHigher firing cost

Figure 4: Impulse response of inflation for a positive productivity shock

-0.004

-0.0035

-0.003

-0.0025

-0.002

-0.0015

-0.001

-0.0005

0

0.0005

1 3 5 7 9 11 13 15 17 19 21 23

Lower firing costHigher firing cost

Figure 5: Impulse response of inflation for a monetary policy shock

-0.008

-0.006

-0.004

-0.002

0

0.002

0.004

0.006

1 3 5 7 9 11 13 15 17 19 21 23Lower firing costHigher firing cost

Figure 6: Impulse response of wages for a positive productivity shock

fAtih mAcit

409

Figure 6:ImpulseresponseofWagesforaPositiveProductivityShock

Figure 7:ImpulseresponseofInflationtoaPositiveProductivityShock

-0.004

-0.0035

-0.003

-0.0025

-0.002

-0.0015

-0.001

-0.0005

0

0.0005

1 3 5 7 9 11 13 15 17 19 21 23

Lower firing costHigher firing cost

Figure 5: Impulse response of inflation for a monetary policy shock

-0.008

-0.006

-0.004

-0.002

0

0.002

0.004

0.006

1 3 5 7 9 11 13 15 17 19 21 23Lower firing costHigher firing cost

Figure 6: Impulse response of wages for a positive productivity shock

-0.009

-0.008

-0.007

-0.006

-0.005

-0.004

-0.003

-0.002

-0.001

0

1 3 5 7 9 11 13 15 17 19 21 23

Higer BRR

Lower BRR

Figure 7: Impulse response of inflation to a positive productivity shock

-0.007

-0.006

-0.005

-0.004

-0.003

-0.002

-0.001

0

0.001

1 3 5 7 9 11 13 15 17 19 21 23

Higher BRR

Lower BRR

Figure 8: Impulse response of inflation to a monetary policy shock

lABoR mARkEt institutions AnD WAgE AnD inflAtion DynAmic

410

Figure 8:ImpulseresponseofInflationtoaMonetaryPolicyShock

Figure 9:ImpulseresponseofWagestoaMonetaryPolicyShock

-0.009

-0.008

-0.007

-0.006

-0.005

-0.004

-0.003

-0.002

-0.001

0

1 3 5 7 9 11 13 15 17 19 21 23

Higer BRR

Lower BRR

Figure 7: Impulse response of inflation to a positive productivity shock

-0.007

-0.006

-0.005

-0.004

-0.003

-0.002

-0.001

0

0.001

1 3 5 7 9 11 13 15 17 19 21 23

Higher BRR

Lower BRR

Figure 8: Impulse response of inflation to a monetary policy shock

-0.016

-0.014

-0.012

-0.01

-0.008

-0.006

-0.004

-0.002

0

0.002

1 3 5 7 9 11 13 15 17 19 21 23

Higher BRR

Lower BRR

Figure 9: Impulse response of wages to a monetary policy shock