Human capital theory 清华大学王一江.docx
《Human capital theory 清华大学王一江.docx》由会员分享,可在线阅读,更多相关《Human capital theory 清华大学王一江.docx(12页珍藏版)》请在冰豆网上搜索。
Humancapitaltheory清华大学王一江
Humancapitaltheory
i.Knowledge,abilityandskillsobtainedthroughaperson’s“investment”inacquiringthemarecalledhumancapital(HC).
∙HumancapitaltheoryinmoderneconomicsispioneeredbyT.ShultzandG.Becker,bothbecameNobelPrizelaureatespartlybecauseoftheircontributionstoHC.
∙HCtheoryisthecornerstoneofmoderneconomics.Thetheoryisusedtoexplainmanylabormarketphenomena.
∙ThecostsofinvestmentinHCareintheformsofa)directmonetaryexpenses;b)forgoneearningsor“opportunitycost”;andc)psychiclossesbecauselearningisoftendifficultandtedious.
∙Likeanyinvestment,investmentinHCisforafuturereturn,intheformofhigherfutureearnings.
∙
InclassicHCtheory,anindividual’sHCinvestmentdecisionisbasedonacost-benefitanalysislikeinanyinvestmentdecision.
Grossbenefits
EarningA
A
Lost
income
pay-ageofworker
ments
B
ii.Determiningtheoptimallevelofinvestment
∙Presentvalue=B1/(1+r)+B2/(1+r)2+…+BT/(1+r)T
∙Thispresentvalueiscomparedwiththecurrentmarginalcosttodetermineifanadditionalunitofexpenditureisworthwhile.
∙Internalrateofreturnmethod.Usethesameformulaasabove.Askthequestion“howlargecouldthediscountratebeandstillrendertheinvestmentprofitable?
”
∙Exampleofvaluinghumanassets:
E&S,p.315.
iii.Implicationsofthetheory
∙ThosewhodislikeschoolswillinvestlessinHC.
∙Thosewhohavefeweropportunitieswillinvestless.
MC’
MC
MB
∙Youngerpeoplearemorelikelytoattendcollege.
∙Present-orientedpeoplearelesslikelytogotocollegethanforward-lookingpeople(e.g.,therearepeopleforthemcurrentincomeismoreimportant).
∙Collegeattendancewilldecreaseifthecostsofcollegerise.
∙CollegeattendancewilldecreaseifincomedifferencebetweencollegeandHSgraduatesnarrows.
∙Thosewhobeartheprimaryresponsibilitiesforchildbearingandraising(historicallywomen)arelesslikelytoseekasmucheducation.
∙But,asthenumberofchildrendecreases,womenwillseekmoreeducation.
Generalvs.Firm-SpecificHumanCapital
1.Introduction.
i.DefinitionsGvs.SHC
●Conceptualdefinitions(generalHC,firmSHC,andIndustrySHC.)
●Aretheypracticallyseparable?
ii.Implications
●Firmsarenotwillingtoinvestinworkers’GHC.
●LaborturnoverisadverselyaffectedbyinvestmentinSHC.Mincer&Jovanovic(1981)reportdatathattheannualseparationratedropsby90%fromthefirstyeartothe6thyearwiththefirm.
●GHChasnoeffectonlaborturnover.
2.SHCinvestmentandmatching/turnover(Parsons.1986,inthespiritofMortensen1978;HashimotoandYu,1979;andHarshimoto,1981.)
i.Assumptions
●Theworkerandthefirmmustundertakeaninvestment(inSHC,or“organizationalcapital”)iftheworkeristobeanefficientemployee.Investmentcostisc.
●Aftertheinvestment,theworker’sproductivityinthefirmandthatinthelabormarketaresubjecttorandomshocks.Theformermaybeduetoareversalofdemandforthefirm’sproduct.Thelattermaybeduetosomestructureintheeconomy.(Comment:
twoshocksarenotnecessary.Itistherelativevaluesoftheworker’sproductivityinandoutsideofthefirmthatmatters.)
●Theworker’srealizedproductivityisthus
Vi=μi+θi,i=0,1(w/investment)
Vi=0,(w/oinvestment)
μ0:
worker’sexpectedproductivityatthefirm.
μ1:
worker’sexpectedproductivityatanotheremployment
θ0:
Arandomvariable,shocktoproductivityattheemployment
θ1:
Arandomvariable,shocktoproductivityatotheremployment
E(θ)=0
●Thefirmmaximizesprofitandtheworkerincome.
ii.Results.
●μ0–μ1hastobegreaterthanthecosttomakeinvestmentintheworkerprofitableonaverage,μ0–μ1>c.
●EfficientseparationoccurswhenV0θ1
m
θ0
(-m,0)(0,0)
●Separationislessifθ1&θ0arepositivelycorrelated.Thereisnoseparationiftheyareperfectlyandpositivelycorrelated.
●Ifθ1&θ0areindependent,increasedvarianceineitherofthemwillincreasetheprobabilityofseparation.
●Sinceseparationisefficient,distinctionbetweenlayoffs&quitsisnotnecessaryandmeaningful.
iii.Informationrequirementinsharingthegain
●Definingfactorpayments
w=V1+α(V0–V1)
=V1+αm+α(θ0–θ1)
andπ=V0–w
=(1-α)(V0–V1)
=(1–α)m+(1–α)(θ0–θ1)
where0<α<1isapredeterminedsharingrule.
●Ifinvestmentcostμ0–μ1=m>c,butαm●Underthissharingrule,separationisefficient,i.e.,itoccurswhentheworker’soutsideproductivityisgreaterthanatthecurrentfirm.
●However,toimplementthesharingrule(asmightbesignedinacontract),anunrelated3rdpartymustbeabletoobservetherandomelementsθ0andθ1,aswellasthepermanentfactorsμ0&μ1(theirdifferencem).Withoutrelevantinformation,suchacontractisnotfeasible.
iv.Comments
●Parsonsoverstatedtheinformationalrequirement.If,forexample,thetwoinvolvedpartiesknowμ0&μ1atthebeginning,andlearnaboutθ0andθ1aftertheinvestment,theycanbargainingtodeterminethesharesofthenetgainwithoutinterventionbya3rdparty.Whenthe2partieslearnaboutθ0andθ1iscritical.
●Bargainingdoesn’tworkonlyifthereistwo-sidedasymmetricinformation,e.g.,thefirmknowsθ0buttheworkerdoesn’t.Theworkerknowsθ1butthefirmdoesn’t.
●Afixed-wageandup-or-outcontractisneededtoprovideincentiveswhentheinformationalasymmetryis2-sidedandsoismoralhazard.
Up-Or-OutContractasanIncentivetoEncourageSHCInvestment
1.Thequestion.
i.Inmanyorganizations,ifpromotiondoesn’toccurwithinsomesettime,individualsarenotretainedevenwhenitappearsproductivetodoso.Examplesincludetenure,partnership,andpromotionruleinthemilitary.
ii.Itisnotclearwhylessproductivejuniorworkersaretoldtoleaveratherthanbeingofferedthejobatreducedwages.
iii.Two-sideduncertaintycanexplainthispuzzle.Withtwo-sideduncertainty,bothfirmandworkeraresubjecttoamoralhazardproblem,inwhichworkerhasatendencytounderinvestinhishumancapitalwhilefirmthetendencytounderreportworker’sproductivity.Themodel
2.ThemodelofKahnandHuberman(1988,JOLE).
i.Assumptions.
●Aworker’soutputatthefirmiseitherb2orb1,b2>b1.
●Theopportunitycostoftheworker’slaborisr●Theworkerhastoreceiveatleastutilitylevelv.
●Thefirmproducesnothingiftheworkerisletgo.
●Theworkercanmakeaninvestmenttoincreasetheprobabilityofb2fromf(0)tof
(1).f(0)(1).Theinvestmentisfirmspecific.
●Investmenthascostianditisefficientfortheinvestmenttobeundertaken,i.e.,theinvestmenthasthenetsocialgainZwith
Z≡[f
(1)–f(0)](b2–b1)–i>0
(1)
●Productivityislearnedbeforetheworkisactuallydone,sothatinprinciplethedecisiontoworkornotcanbemadecontingentonthedraw.
●Bothfirmandworkerarerisk-neutral.
ii.Thecaseofperfectinformation
Underperfectinformation,thetwosideswillsignanefficientcontractwiththefollowingcontents.
●Theworkerwillundertaketheinvestment.
●Hewillworkforthefirmnomatterwhatthedrawofproductivitymaybe.
●Heiscompensatedy=v+i(=r+i?
)
iii.Whenthefirmhasprivateinformationregardingtheworker’sproductivity.
●Normally,thiswillresultinsomedistortion.
●However,inthisparticularcase,itdoesn’talterthefullinformationcontractbecausetheworker’soutsideproductivityisneverashigh.Aslongasicanbecontractedandtheinvestmentismade,thesamefixedwageshouldbeofferedandtheworkeralwaysretained.
iv.Wheninvestmentisprivateinformation
●Thisisastraightforwardmoralhazardproblem.
●Thecontracttiescompensationtooutput.Buttheworkerisstillalwaysretained.
●Lety2andy1representthepayofferedinthetworealizations,withy2>y1.Theincentivecompatibilityconditionforundertakingtheinvestmentis
f
(1)y2+[1–f
(1)]y1–i
(2)
>f(0)y2+[1–f(0)]y1.(IC)
●Fortheworkertoacceptthecontract,itisnecessarythat
f
(1)y2+[1–f
(1)]y1–i>v.(IR)(3)
●Together,y2andy1arefoundwhenthetwoconditionsaretreatedasequalitieswilldeterminelevelsofy2andy1thatwillmaximizethefirm’sprofits.Notethat
y2>v+i>v>y1(4)
●Itremainstobedeterminedthatthecontractentailsahigherprofitthanifitdoesn’tinducetheworkertoinvest.Thisrequiresthat
f
(1)(b2-y2)+[1–f
(1)](b1-y1)–i
>f(0)b2+[1–f(0)]b1-v.(5)
●(5)canbereducedto
(1).Thismeansthat,providedtheinvestmentissociallyproductiveandagentsareriskneutral,moralhazardproblemalonecreatesnoinefficiency.Thereasonforthisisthat,aslongastheinvestmentismade,gainorlossbyeitherpartyisonlyatransferbetweenthetwoparties,notanetsocialloss.
●
(2)guaranteesthattheinvestmentwillbemade.
v.Two-sidedinformationalproblemandup-or-outcontract.
Theproblem:
●Theworkerhasprivateinformationregardinginvestment.
●Thefirmhasprivateinformationregardingtheproductivitydraw,i.e.,whetherit’sb1orb2.
●Becauseiisprivateinformation,theworkerwillnotinvestifheispaidafixedwage&guaranteedemployment(asunder