Household Savings in Russia during the Transition.docx
《Household Savings in Russia during the Transition.docx》由会员分享,可在线阅读,更多相关《Household Savings in Russia during the Transition.docx(38页珍藏版)》请在冰豆网上搜索。
HouseholdSavingsinRussiaduringtheTransition
HouseholdSavingsinRussia
duringtheTransition
MarkC.FoleyWilliamPyle
DepartmentofEconomicsDepartmentofEconomics
DavidsonCollegeMiddleburyCollege
Davidson,NC28035USAMiddlebury,VT05753USA
phone:
(704)894-2248phone:
(802)443-3240
mafoley@davidson.eduwpyle@middlebury.edu
Abstract
WeexploitpaneldatafromthesecondphaseoftheRussianLongitudinalMonitoringSurvey(RLMS)toinvestigatethehouseholdcharacteristicsthatexplainsavingduringaperiodofextremedislocation.Amongourmorenoteworthyfindings,weconfirmtherobustnessoftheU-shapedage-savingsrelationshiptomultivariatespecifications.Wealsofindevidenceofshort-termconsumptionsmoothingbehaviorinthatsavingsratesarelowerinhouseholdssufferingfromunemploymentorpaymentarrears.Conditioningonincomelevel,wealsoshowhowthecompositionofhouseholdincomeinfluencesthepropensitytosave.Andfinally,weturnupstrongsupportforaninverserelationshipbetweenthehousehold’sstockofdurablesanditssavingrate.
JEL:
D91,P20.
Note:
TheauthorsthankMichaelAlexeev,RickEricson,IvanKhotulev,andseminarparticipantsattheHarrimanInstituteatColumbiaUniversityforcommentsonearlierversions.
HouseholdSavingsinRussiaduringtheTransition
Introduction
Inthispaper,weexploitpaneldatafromthesecondphaseoftheRussianLongitudinalMonitoringSurvey(RLMS)toinvestigatethehouseholdcharacteristicsthatexplainsavingduringaperiodofextremedislocation.Ourinvestigationstartsfromthepre-suppositionthatthesavingsbehaviorthathasemergedinRussiamaynotconformtothatfoundinmorematureandstablemarketenvironments.Theexperienceofpriceliberalization,justoveradecadeago,offeredaninitial“lesson”ofsortsinthefutilityofsavingsasRussianssawtherealvalueoftheirrubleholdings,stashed“underthemattress”orheldinthestatesavingsbank,rapidlydecline.Thesubsequentperiodofhyperinflation,thewidespreadfailuresoflargefinancialinstitutionsandtheunexpecteddevaluationoftherublein1998alsoprovidedsimilarlydiscouraging“lessons”astothevalueoftraditionalformsofsaving.Consider,furthermore,thatRussianshavebeenforcedtolearnothersortsof“lessons”intheircountry’snascentlaborandconsumergoodsmarkets.Forthefirsttimeingenerations,theyhavehadtodealwithuncertaintiesassociatedwithpaymentarrearsandunemployment.Butasincomestreamshavebecomelesspredictable,marketizationhasrenderedthedistributionandavailabilityofconsumergoodsmuchmoreso.
Whatsortofsavingpatternshavebeenproducedinsuchapotentiallydisorientingperiodoftransition?
Theconfluenceandshearmagnitudeofthesechangesleadsus,insomesense,toexpecttheunexpected.Findinginter-householdvariationthatappearstocontradictexperienceselsewhereshouldnotsurpriseus.Gregoryetal.(1999),forinstance,reportaU-shapedrelationshipbetweentheageofthehouseholdheadandsavingsratesinoneroundoftheRLMS.Thisfindingisatoddswithboththeage-savingsrelationshipfoundinothercountriesandtheinvertedUpredictedbythestandardlife-cyclehypothesis.
Webuilduponthisandotherwork,investigatingtherelationshipbetweenhouseholdcharacteristicsandsavingsinapanelsettingfrom1994to2001.Amongourmorenoteworthyfindings,weconfirmtherobustnessoftheU-shapedsavings-agerelationshiptomultivariatespecifications.Althoughthismightbeinterpretedasevidenceagainstconsumptionsmoothingoverthelongrun,wedofindevidenceofshort-termsmoothingbehaviorinthatsavingsratesarelowerforhouseholdssufferingfromunemploymentorpaymentarrears.Wealsoshowhowthecompositionofhouseholdincomeinfluencesthepropensitytosave;specifically,thosehouseholdsthatbenefitdisproportionatelyfromnon-standard(informal)sourcesofsupporttendtosavemore.Andfinally,weturnupstrongevidenceforaninverserelationshipbetweenthesizeofahousehold’sstockofdurablegoodsanditssavingrate.
ThesefindingsbearupontwoimportantaspectsofRussia’seconomictransition.Sincehouseholdsavingscanbeacriticalsourceofsupportfornewinvestmentprojects,therelationshipshighlightedherecouldsuggesthowcertaindevelopmentsorpolicieswouldinfluencetheaggregatesavingsrateandthusthepaceofeconomicrestructuringanddevelopment.Onanotherlevel,ourconclusionsmaysuggestthatthesavingspatternacrosshouseholdsis,ingoodpart,afunctionoftheinstitutionalunder-developmentoftheRussianeconomy.Forinstance,ifthebehaviorsuggeststhatRussianhouseholdshaveadaptedtoananemicmarketforconsumercreditorthegeneralabsenceofsocialinsuranceprograms,wemaybebetterplacedtopredictthesavings-relatedconsequencesofthedevelopmentinthoseinstitutions.
Ourpaperisstructuredasfollows.Inthenextsection,wediscusshowweusetheRLMStoestimatehouseholdsavingsratesinRussia.Then,afterexploringbivariaterelationshipsbetweensavingsratesandvarioushouseholdcharacteristics,weestimatethedeterminantsofhouseholdsavingsinapanelregression.Weconcludewithadiscussionoftheresultsandtheirimplications.
MeasuringHouseholdSavingsinRussia
SincetheRLMSprovidesadetailedaccountingofhouseholdbudgetaryflowsandonlyminimalinformationonwealth,bydefault,wecomputesavingsasthedifferencebetweenmonthlyincomeandconsumptionflows.Thismeasure,asopposedtoafirstdifferenceofassets,directlyreflectsdecisionsabouthowmuchhouseholdsconsumeandsaveandindicatestheresourcesthathouseholdshavesetasidethatcouldbeusedtofinanceinvestment.
Weproceedwithanaccountingofthesub-componentsthatareincludedinthesetwoflows.Incomeismeasuredasthesumofmultiplecashandnon-cashpaymentsreceivedbyalladultsinahouseholdduringthemonthprecedingthesurvey.Wages,whichconstitutethelargestshareofhouseholdincome,includebothcashand(theestimatedvalueof)in-kindpaymentsfromprimaryandanysecondaryjobs.AsshowninFigure1,byRoundX,employmentincomeamountstoalmosthalfoftotalhouseholdincome.Wagearrears,althoughanimportantelementoftheRussianstory,presentdifficultiesofinterpretation.Eventhoughtheycouldbeconstruedasfutureincome,theRLMSdoesnotallowustodeterminetherateatwhichahouseholdmightdiscountaruble’sIOUsincethatwouldbeafunctionofitsparticularbeliefsaboutifandwhenthedebtwouldbesettled.Forthesereasons,wehaveexcludedthemfromourcalculations.
Publictransfers,thesecondlargestcomponentofhouseholdincome(consistentlyaroundthirtypercent),areprimarilycomposedofpensionsbutalsoincludeunemploymentandchildbenefitsaswellashousingandutilitysubsidies.Therentandinterestincomegenerated,respectively,byahouseholds’realandfinancialassets,haveconsistentlybeennegligibleinaggregateterms;asshowninFigure1,innosinglerounddoesthis“investmentincome”accountformorethanhalfapercentagepointofaveragehousehold’sincome.
Non-standardsourcesofincomefigureprominentlyinthetransitionalcontext.Alargenumberofhouseholds,forinstance,benefitsignificantlyfrommoniesreceivedfromfamily,friendsandnon-publicphilanthropicorganizations(KuhnandStillman,2002).Onaverage,theseprivatetransfershaveconsistentlyamountedtojustundertenpercentofhouseholdincome.Aslightlylargershareofincomeoriginatesfromself-producedfoodproducts,whichincludethesaleoflivestockandwildfoods.Muchofthis“homeproduction”amountstoautoconsommation(i.e.,foodproductsproducedandconsumedbythesamehousehold),butanon-trivialportionissoldforcash.WiththeexceptionofRoundVIII,whichwasadministeredsoonafterthe1998crisis,therelativeimportanceofhomeproductionseemstobeindecline.
InFigure2,weseparateouthouseholdexpendituresintoseveralsub-components.Autoconsommationgetsrecordedas(homeproduced)consumptioninadditiontoincome(Deaton,1997).Non-home-producedfoodproducts,wefind,haveconsistentlyconstitutedabouthalfofallhouseholdexpenditures.Acatchallcategory(“other”)thatcapturesexpendituresonservicesandutilitieshasincreasedinrelativesizesinceRoundV.
ThefinalcategoryreflectedinFigure2is“durables.”Sincetheyprovideastreamofbenefitsoverasustainedperiodoftime,itisunderstandabletodiscountthecontributionofexpendituresonrefrigerators,cars,dachas,etc.tocurrentconsumption.Moreover,sincetheymayserveasanadequatestoreofvalue,theymightberegardedasanimportantformofsaving,particularlyinaneconomy,likeRussia’s,thathasexperiencedhyperinflation,unexpectedcurrencydevaluationsandmassivefailuresoffinancialinstitutionsallwithinthepastdecade.Russians,thatis,mightunderstandablyconsiderthemselvestohavefewotherreliablealternativesforsaving.Indeed,sociologicalresearchhasturnedupevidencethatsuggeststhathouseholdsself-interprettheirpurchasesofdurablesinjustthismanner(Shevchenko,2002).Therefore,intheanalysisthatfollowswerecordtwomeasuresofsavings,oneinclusiveandoneexclusiveofdurables.
Havingnowlaidoutthetwostreamswhosedifferenceconstitutessavings,itisstraightforwardtonotethatthesavingsrateissimplycomputedassavingsdividedbyincome.Intheestimatesandanalysisthatfollow,wepurgethedataofacutedis-savers–i.e.,householdswitheithernegativenetincomeorthatreportspendingmorethantwicethevalueoftheirin