通货膨胀、价格分散和福利：消费者搜索的作用（英）

FinanceandEconomicsDiscussionSeries FederalReserveBoard,Washington,D.C.ISSN1936-2854(Print) ISSN2767-3898(Online) Inflation,PriceDispersion,andWelfare:TheRoleofConsumerSearch FranciscaSara-Zaror2024-047 Pleasecitethispaperas: Sara-Zaror,Francisca(2024).“Inflation,PriceDispersion,andWelfare:TheRoleofCon-sumerSearch,”FinanceandEconomicsDiscussionSeries2024-047.Washington:BoardofGovernorsoftheFederalReserveSystem,https://doi.org/10.17016/FEDS.2024.047. NOTE:StaffworkingpapersintheFinanceandEconomicsDiscussionSeries(FEDS)arepreliminarymaterialscirculatedtostimulatediscussionandcriticalcomment.TheanalysisandconclusionssetfortharethoseoftheauthorsanddonotindicateconcurrencebyothermembersoftheresearchstaffortheBoardofGovernors.ReferencesinpublicationstotheFinanceandEconomicsDiscussionSeries(otherthanacknowledgement)shouldbeclearedwiththeauthor(s)toprotectthetentativecharacterofthesepapers. Inflation,PriceDispersion,andWelfare:TheRoleofConsumerSearch FranciscaSara-Zaror∗†‡May21,2024 Abstract Instandardmacroeconomicmodels,thecostsofinflationaretightlylinkedtothepricedispersionofidenticalgoods.Therefore,understandinghowpricedispersionempiricallyrelatestoinflationiscrucialforwelfareanalysis.Inthispaper,Istudytherelationshipbetweensteady-stateinflationandpricedisper-sionforacrosssectionofU.S.retailproductsusingscannerdata.Bycomparingpricesofitemswiththesamebarcode,mymeasureofrelativepricedispersioncontrolsforproductheterogeneity,overcominganimportantchallengeinthelit-erature.Idocumentanewfact:pricedispersionofidenticalgoodsincreasessteeplyaroundzeroinflationandbecomesflatterasinflationincreases,display-ingaΥ-shapedpattern.Currentsticky-pricemodelsareinconsistentwiththisfinding.Idevelopamenu-costmodelwithidiosyncraticproductivityshocksandsequentialconsumersearchthatreproducesthenewfactandexhibitsrealisticprice-settingbehavior.Inthemodel,inflation-inducedpricedispersionincreasesshoppers’incentivestosearchforlowpricesandthuscompetitionamongretailers.Thepositivewelfare-maximizinginflationrateoptimallytradesofftheefficiencygainsfromlowermarkupsandtheresourcesspentonsearch. JELCodes:E31,E50,L11,L16. ∗FederalReserveBoard.E-mail:francisca.sara-zaror@frb.gov †Iamgreatlyindebtedtomyco-chairsFernandoAlvarezandMikeGolosov,andmydissertationcommitteemembersGregKaplanandRobShimer,forinvaluableadvice,support,andencouragementateverystageofthisproject.IamgratefultoAgustinHurtado,JoshMorris-Levenson,HyejinPark,andMartaPratoforinsightfulfeedback;tomydiscussantsKieranLarkin,GaetanoGaballo,andJaneRyngaert,aswellasparticipantsoftheASSA2023,MidwestMacroMeetings2023,SED2023,Inflation:DriversandDynamicsConference2023,XXVWorkshopinInternationalEconomicsandFinance,andmultipleseminarsforhelpfulcomments. ‡Researcher(s)ownanalysescalculated(orderived)basedinpartondatafromNielsenConsumer LLCandmarketingdatabasesprovidedthroughtheNielsenIQDatasetsattheKiltsCenterforMar-ketingDataCenteratTheUniversityofChicagoBoothSchoolofBusiness.TheconclusionsdrawnfromtheNielsenIQdataarethoseoftheresearcher(s)anddonotreflecttheviewsofNielsenIQ.NielsenIQisnotresponsiblefor,hadnorolein,andwasnotinvolvedinanalyzingandpreparingtheresultsreportedherein. 1Introduction Asalientfeatureofmicro-leveldataisthatthepricesofidenticalgoodsvaryacrosssellers.Incurrentmacroeconomicmodels,inflationisacrucialdeterminantofpricedis-persion,andthecostsofinflationmainlyarisefrominflation-inducedpricedispersion.Inthispaper,Iaddressthreequestions:Whatistheempiricalrelationshipbetweeninflationandpricedispersion?Whatdoesthisrelationshipimplyforcurrentmonetarytheories?Whatdowelearnforthewelfareanalysisofinflation? Studyinghowpricedispersionempiricallyrelatestoinflationimposesseveralchal-lenges.Thefirstistomeasurepricedispersionaccurately.Forthis,weneedtoobservethepricesdifferentsellerschargeforidenticalgoods.Typically,granulardatasetswhichhavebeenessentialtoestablishfactsonmicro-pricerigiditydonotsatisfythisrequirement.Forinstance,quotesinthemicro-dataunderlyingtheCPIaregroupedintoproductcategorieswithlimitedinformationaboutthespecificproductbeingmea-suredortheabilitytolinkidenticalproductsacrosssellers.Thus,thecomparisonofpricesacrosssellers,evenwithinnarrowproductcategories,wouldnottakeproductheterogeneityintoaccount(Nakamuraetal.,2018). IovercomethischallengeusinghighlydetailedscannerdataforretailproductsintheU.S.Thedatasetcontainspricesandquantitiesofproductssoldinover35,000storesacrossthecountrybetween2006and2020.Theobservationsareidentifiedbytheproductbarcode,theweek,andtheretailerwherethetransactionwascarriedout.Morethan3millionbarcodesareavailable,eachofthemsoldby50storesonaverage.Eachretailerbelongstooneofoverathousandgeographicallydispersedcounties.Withthesedata,Icancomparepricesofproductswithidenticalbarcodessoldacrossdifferentretailersonthesamedateandcounty.Hence,mymeasureofrelativepricedispersioncontrolsforseveralsourcesofheterogeneitywhichaprioriareunrelatedtoinflation. Ashortcomingofthisdatasetisthatitisonlyavailableforarelativelylowandstableaggregateinflationperiod.Therefore,t