劳动与社会保障 外文翻译.docx

劳动与社会保障 外文翻译.docx

《劳动与社会保障 外文翻译.docx》由会员分享，可在线阅读，更多相关《劳动与社会保障 外文翻译.docx（9页珍藏版）》请在冰豆网上搜索。

劳动与社会保障外文翻译

劳动与社会保障外文翻译

毕业论文外文资料翻译

题目西部农村新型农村养老保险参保

意愿的影响因素研究

——以青海省乐都县八里桥村为例

学院政治与公共管理学院

专业劳动与社会保障

班级社保0803班

学生赵继平

学号220080706103

指导教师刘艳丽

二〇一二年三月二十五日

ThissectionofarticlesselectedfromATheoryoftheConsumptionFunction,P26-30.ThewriterisFriedmanMilton.

Theassumptionsthatthefirsttwocorrelations—betweenthepermanentandtransitorycomponentsofincomeandofconsumption—arezeroseemverymildandhighlyplausible.Indeed,bythemselves,theyhavelittlesubstantivecontentandcanalmostbeassimplycompletingortranslatingthedefinitionsoftransitoryandpermanentcomponents;thequalitativenotionthatthetransitorycomponentisintendedtoembodyisofanaccidentalandtransientadditiontoorsubtractionfromincome,whichisalmostequivalenttosayinganadditionorsubtractionthatisnotcorrelatedwiththerestofincome.Themergingoferrorsofmeasurementwithtransitorycomponentscontributesfurthertotheplausibilitythatthesecorrelationsarezero.

Foragroupofindividuals,itisplausibletosupposethattheabsolutesizeofthetransitorycomponentvarieswiththesizeofthepermanentcomponent:

thatagivenrandomeventproducesthesamepercentageratherthanthesameabsoluteincreaseordecreaseintheincomesofunitswithdifferentpermanentcomponents.Thismaymakemoreconvenientanalternativedefinitionoftransitorycom-ponentthatissuggestedbelow;itisnot,however,inconsistentwithzerocorrelation.Zerocorrelationimpliesonlythatthe'averagetransitorycomponent—thealgebraicaverageinwhichpositiveandnegative.componentsoffsetoneanother—isthesameforallvaluesofthepermanentcomponent.Forexample,supposethatthetransitorycomponentisequallylikelytobeplusorminus10percentofthepermanentcomponent.Theaveragetransitorycomponentisthenzeroforallvaluesofthepermanentcomponent,althoughtheaverageabsolutevalue,whichdisregardsthesignofthecomponents,isdirectlyproportionaltothepermanentcomponent.

Theplausibilityoftakingourdefinitionoftransitorycomponentstoimplyazerocorrelationforagroupofconsumerunitsdependssomewhatonthecriteriadeterminingmembershipinthegroup.Theclearestexampleisaclassificationof-unitsbythesizeoftheirmeasuredincome.Foreachsuchgroup,thecorrelationbetweenpermanentandtransitorycomponentsisnecessarilynegative,sincewithacommonmeasuredincomethepermanentcomponentcanberelativelyhighonlyifthetransitorycomponentisrelativelylow,andconversely.

Theassumptionthatthethirdcorrelationin（3.3）—betweenthetransitorycomponentsofincomeandconsumption—iszeroisamuchstrongerassumption.Itisprimarilythisassumptionthatintroducesimportantsubstantivecontentintothehypothesisandmakesitsusceptibleofcontradictionbyawiderangeofphenomenacapableofbeingobserved.Theultimatetestofitsacceptabilityisofcoursewhethersuchphenomenaareinfactobserved,andmostofwhatfollowsisdevotedtothisquestion.Itishardlyworthproceedingtosuchmorerefinedtests,however,unlesstheassumptioncanpass—oratleastnotfailmiserably—themuchcrudertestofconsistencywithcasualobservationofone'sselfandone'sneighbors,sosomecommentsontheintuitiveplausibilityoftheassumptionarenotoutoforder.

Thecommonnotionthatsavings,oratleastcertaincomponentsofsavings,area"residual"speaksstronglyfortheplausibilityoftheassumption.Forthisnotionimpliesthatconsumptionisdeterminedbyratherlong-termconsiderations,sothatanytransitorychangesinincomeleadprimarilytoadditionstoassetsortotheuseofpreviouslyaccumulatedbalancesratherthantocorrespondingchangesinconsumption.

Yetfromanotherpointofview,theassumptionseemshighlyimplausible.Willnotamanwhoreceivesanunexpectedwindfalluseatleastsomepartofitin"riotousliving,"i.e.inconsumptionexpenditures?

Wouldhebelikelytoaddthewholeofittohiswealth?

Theanswertothesequestionsdependsgreatlyonbow"consumption"isdefined.Theoffhandaffirmativeanswerreflectsinlargemeasure,Ibelieve,animplicitdefinitionofconsumptionintermsofpurchases,includingdurablegoods,ratherthanintermsofthevalueofservices.Ifthelatterdefinitionisadopted,asseemshighlydesirableinapplyingthehypothesistoempiricaldata—thoughunfortunatelyIhavebeenabletodosotoonlyalimitedextent—muchthatoneclassifiesoffhandasconsumptionisreclassifiedassavings.Isnotthewindfalllikelytobeusedforthepurchaseofdurablegoods?

Or,toputitdifferently,isnotthetimingofthereplacementofdurablegoodsandofadditionstothestockofsuchgoodslikelytosomeextenttobeadjustedsoastocoincidewithwindfalls?

Twootherconsiderationarguefortheplausibilityoftheassumptionthattransitorycomponentsofincomeandconsumptionareuncorrelated.First,theaboveidentificationofawindfallwithtransitoryincomeisnotprecise.Suppose,forexample,inheritancesareincluded.inaparticularconceptofmeasuredincome.Consideraconsumerunitwhosereceiptsremainunchangedoverasuccessionoftimeperiodsexceptthatitreceivesaninheritanceinthefinalperiod.Iftheinheritancewasexpectedtooccursometimeorother,itwillalreadyhavebeenallowedforinpermanentincome;thetransitorycomponentofincomeisonlytheexcessoftheinheritanceoverthiselementofpermanentincome.Therereasonwhythereceiptoftheinheritanceshouldmakeconsumptioninthefinalperioddifferentfromthatofprecedingperiods,exceptthroughinabilitytoborrowinadvanceonthestrengthoftheinheritance.Butthisimpliesthatthereceiptoftheinheritancechangesw（theratioofwealthtoincome）in（2.6）;itisthereforealreadytakenintoaccountinthehypothesis.Thereisnoessentialdifferenceiftheinheritanceisunexpected.Theeffectoftheinheritanceisthentoincreasethepermanentincomeoftheunit,andthiswilljustifyahigherconsumptioninthefinalperiod;againthetransitorycomponentisonlytheexcessofthewindfalloverthiselementofpermanentincome,anditisnolongerintuitivelyobviousthatitshouldleadtoanincreaseincurrentconsumption.7Thesecondconsiderationisthatjustasthereareinstancesinwhichonewouldexpectatransitoryincreaseinincometoproduceatransitoryincreaseinconsumption,soalsothereareinstancesinwhich,onewouldexpectthereverse.Thesimplestexampleiswhenatransitoryincreaseinincomereducesopportunitiesfor-consumptionaswhenitisobtainedbyworkinglongerhoursorgoingtoabackwardcountry.Suchnegativeandpositivecorrelationswilltendtooffsetoneanother.

Theprecedingremarksabstractfromerrorsofmeasurement.Yet,asnoted,inanystatisticalanalysiserrorsofmeasurementwillingeneralbeindissolublymergedwiththecorrectlymeasuredtransitorycomponents.Theeffectonthecorrelationbetweenstatisticallyrecordedtransitorycomponentsofincomeandconsumptiondependscriticallyonhowthestatisticaldataareobtained.Ifincomeandconsumptionaremeasuredindependently,the.errorsofestimatemightbeexpectedtobeindependentaswellandthereforetocontributetowardasmallorzeroobservedcorrelationbetweentransitorycomponentsofincomeandconsumption.Ontheotherhand,ifconsumptionisestimated,asitfrequentlyis,bymeasuringpendentlysavingsandincomeandsubtractingtheformerfromthelatter,thenmeasuredconsumptionandmeasuredincomehavecommonerrorsofmeasurement.Thistendstowardapositiveobservedcorrelationbetweentransitorycomponentsofincomeandconsumption.

Thepurposeoftheseremarksisnottodemonstratethatazerocorrelationistheonlyplausibleassumption—neitherevidencelikethatalludedtonoranyothercanjustifysuchaconclusion.Itspurposeisrathertoshowthatcommonobservationdoesnotrenderitabsurdtosupposethatahypothesisembodyingazerocorrelationcanyieldafairlycloseapproximationtoobservedconsumerbehavior.Theassumptionthatthecorrelationbetweentransitorycomponentsofincomeandconsumptioniszerocould,ofcourse,bereplacedbythelessrestrictiveassumptionthatitisapositivenumberbetweenzeroandunity,butthiswouldgreatlyweakenthehypothesisandreduceitspotentialusefulnessforpredictingbehavior.Itseemshighlyundesirabletodosountilandunlessasignificantcontradictionarisesbetweenthestrongerhypothesisandempiricalevidenceonconsumerbehavior.

Aparticularlysimplespecialcaseofthehypothesisarisesif,inadditionto（3.3）,itisassumedthatthemeantransitorycomponentsofconsumptionandincomearezero,or

Where

standsforthemeanofthevariabledesignatedbyitssubscript.Thisassumptioniseminentlyreasonableiftheprobabilitydistributioninquestionissufficientlycomprehensive.Ingeneral,however,weshallwanttouse.conditionalprobabilitydistributions,forexample,thedistributionoftransitory'componentsinaparticularyear,orformembersofaparticulargroup.Insuchcases,itwillgenerallybeundesirabletoassumethat（3.4）holds,justasforthesingleconsumerunitviewedexpostitisundesirabletoassumethatthetransitorycomponentsthemselvesarenecessarilyzero.

Itmaybedesirableornecessarytoimposeadditionalconditionsonthe'probabilitydistributionstofacilitatetheestimationoftheparametersofthesystemfromobserveddata.Ishall,however,largelyneglecttheproblemofstatisticalestimation,andsoweneednotgointosuchconditions.

Amoreimportantqualificationisthat,for