EXPECTATION
ideathattheonlyriskherunsisthatofagenuinechangeinthenewsoverthenearfuture，astothelikelihoodofwhichhecanattempttoformhisownjudgment，andwhichisunlikelytobeverylarge。For，assumingthattheconventionholdsgood，itisonlythesechangeswhichcanaffectthevalueofhisinvestment，andheneednotlosehiSsleepmerelybecausehehasnotanynotionwhathisinvestmentwillbeworthtenyearshence。Thusinvestmentbecomesreasonably'safe'fortheindividualinvestorovershortperiods，andhenceoverasuccessionofshortperiodshowevermany，ifhecanfairlyrelyontherebeingnobreakdownintheconventionandonhisthereforehavinganopportunitytorevisehisjudgmentandchangehisinvestment，beforetherehasbeentimeformuchtohappen。
Investmentswhichare'fixed'forthecommunityarethusmade'liquid'
fortheindividual。
Ithasbeen，Iamsure，onthebasisofsomesuchprocedureasthisthatourleadinginvestmentmarketshavebeendeveloped。Butitisnotsurprisingthataconvention，inanabsoluteviewofthingssoarbitrary，shouldhaveitsweakpoints。Itisitsprecariousnesswhichcreatesnosmallpartofourcontemporaryproblemofsecuringsufficientinvestment。
V
Someofthefactorswhichaccentuatethisprecariousnessmaybebrieflymentioned。
1Asaresultofthegradualincreaseintheproportionoftheequityinthecommunity'saggregatecapitalinvestmentwhichisownedbypersonswhodonotmanageandhavenospecialknowledgeofthecircumstances，eitheractualorprospective，ofthebusinessinquestion，theelementofrealknowledgeinthevaluationofinvestmentsbywhosewhoownthemorcontemplatepurchasingthemhasseriouslydeclined。
2Day-to-dayfluctuationsintheprofitsofexisting[Page154]THE
GENERALTHEORYOFEMPLOYMENT
investments，whichareobviouslyofanephemeralandnon-significantcharacter，tendtohaveanaltogetherexcessive，andevenanabsurd，influenceonthemarket。Itissaid，forexample，thatthesharesofAmericancompanieswhichmanufactureicetendtosellatahigherpriceinsummerwhentheirprofitsareseasonallyhighthaninwinterwhennoonewantsice。Therecurrenceofabank-holidaymayraisethemarketvaluationoftheBritishrailwaysystembyseveralmillionpounds。
3Aconventionalvaluationwhichisestablishedastheoutcomeofthemasspsychologyofalargenumberofignorantindividualsisliabletochangeviolentlyastheresultofasuddenfluctuationofopinionduetofactorswhichdonotreallymakemuchdifferencetotheprospectiveyield；sincetherewillbenostrongrootsofconvictiontoholditsteady。
Inabnormaltimesinparticular，whenthehypothesisofanindefinitecontinuanceoftheexistingstateofaffairsislessplausiblethanusualeventhoughtherearenoexpressgroundstoanticipateadefinitechange，themarketwillbesubjecttowavesofoptimisticandpessimisticsentiment，whichareunreasoningandyetinasenselegitimatewherenosolidbasisexistsforareasonablecalculation。
4Butthereisonefeatureinparticularwhichdeservesourattention。Itmighthavebeensupposedthatcompetitionbetweenexpertprofessionals，possessingjudgmentandknowledgebeyondthatoftheaverageprivateinvestor，wouldcorrectthevagariesoftheignorantindividuallefttohimself。Ithappens，however，thattheenergiesandskilloftheprofessionalinvestorandspeculatoraremainlyoccupiedotherwise。Formostofthesepersonsare，infact，largelyconcerned，notwithmakingsuperiorlong-termforecastsoftheprobableyieldofaninvestmentoveritswholelife，butwithforeseeingchangesintheconventionalbasisofvaluationashorttimeaheadofthegeneralpublic。Theyareconcerned，notwithwhataninvestmentis[Page155]LONG-TERM
EXPECTATION
reallyworthtoamanwhobuysit'forkeeps'，butwithwhatthemarketwillvalueitat，undertheinfluenceofmasspsychology，threemonthsorayearhence。Moreover，thisbehaviourisnottheoutcomeofawrong-headedpropensity。Itisaninevitableresultofaninvestmentmarketorganisedalongthelinesdescribed。Foritisnotsensibletopay25foraninvestmentofwhichyoubelievetheprospectiveyieldtojustifyavalueof30，ifyoualsobelievethatthemarketwillvalueitat20threemonthshence。
Thustheprofessionalinvestorisforcedtoconcernhimselfwiththeanticipationofimpendingchanges，inthenewsorintheatmosphere，ofthekindbywhichexperienceshowsthatthemasspsychologyofthemarketismostinfluenced。Thisistheinevitableresultofinvestmentmarketsorganisedwithaviewtoso-called'liquidity'。Ofthemaximsoforthodoxfinancenone，surely，ismoreanti-socialthanthefetishofliquidity，thedoctrinethatitisapositivevirtueonthepartofinvestmentinstitutionstoconcentratetheirresourcesupontheholdingof'liquid'securities。
Itforgetsthatthereisnosuchthingasliquidityofinvestmentforthecommunityasawhole。Thesocialobjectofskilledinvestmentshouldbetodefeatthedarkforcesoftimeandignorancewhichenvelopourfuture。
Theactual，privateobjectofthemostskilledinvestmentto-dayis'tobeatthegun'，astheAmericanssowellexpressit，tooutwitthecrowd，andtopassthebad，ordepreciating，half-crowntotheotherfellow。
Thisbattleofwitstoanticipatethebasisofconventionalvaluationafewmonthshence，ratherthantheprospectiveyieldofaninvestmentoveralongtermofyears，doesnotevenrequiregullsamongstthepublictofeedthemawsoftheprofessional；¾；itcanbeplayedbyprofessionalsamongstthemselves。Norisitnecessarythatanyoneshouldkeephissimplefaithintheconventionalbasisofvaluationhavinganygenuinelong-termvalidity。Foritis，sotospeak，agameofSnap，[Page156]THE
GENERALTHEORYOFEMPLOYMENT
ofOldMaid，ofMusicalChairs¾；apastimeinwhichheisvictorwhosaysSnapneithertoosoonnortoolate，whopassedtheOldMaidtohisneighbourbeforethegameisover，whosecuresachairforhimselfwhenthemusicstops。Thesegamescanbeplayedwithzestandenjoyment，thoughalltheplayersknowthatitistheOldMaidwhichiscirculating，orthatwhenthemusicstopssomeoftheplayerswillfindthemselvesunseated。
Or，tochangethemetaphorslightly，professionalinvestmentmaybelikenedtothosenewspapercompetitionsinwhichthecompetitorshavetopickoutthesixprettiestfacesfromahundredphotographs，theprizebeingawardedtothecompetitorwhosechoicemostnearlycorrespondstotheaveragepreferencesofthecompetitorsasawhole；sothateachcompetitorhastopick，notthosefaceswhichhehimselffindsprettiest，butthosewhichhethinkslikeliesttocatchthefancyoftheothercompetitors，allofwhomarelookingattheproblemfromthesamepointofview。Itisnotacaseofchoosingthosewhich，tothebestofone'sjudgment，arereallytheprettiest，noreventhosewhichaverageopiniongenuinelythinkstheprettiest。Wehavereachedthethirddegreewherewedevoteourintelligencestoanticipatingwhataverageopinionexpectstheaverageopiniontobe。
Andtherearesome，Ibelieve，whopractisethefourth，fifthandhigherdegrees。
Ifthereaderinterjectsthattheremustsurelybelargeprofitstobegainedfromtheotherplayersinthelongrunbyaskilledindividualwho，unperturbedbytheprevailingpastime，continuestopurchaseinvestmentsonthebestgenuinelong-termexpectationshecanframe，hemustbeanswered，firstofall，thatthereare，indeed，suchserious-mindedindividualsandthatitmakesavastdifferencetoaninvestmentmarketwhetherornottheypredominateintheirinfluenceoverthegame-players。Butwemustalsoaddthatthereare[Page157]LONG-TERM
EXPECTATION
severalfactorswhichjeopardisethepredominanceofsuchindividualsinmoderninvestmentmarkets。Investmentbasedongenuinelong-termexpectationissodifficultto-dayastobescarcelypracticable。Hewhoattemptsitmustsurelyleadmuchmorelaboriousdaysandrungreaterrisksthanhewhotriestoguessbetterthanthecrowdhowthccrowdwillbehave；and，givenequalintelligence，hemaymakemoredisastrousmistakes。Thereisnoclearevidencefromexperiencethattheinvestmentpolicywhichissociallyadvantageouscoincideswiththatwhichismostprofitable。Itneedsmoreintelligencetodefeattheforcesoftimeandourignoranceofthefuturethantobeatthegun。Moreover，lifeisnotlongenough；¾；humannaturedesiresquickresults，thereisapeculiarzestinmakingmoneyquickly，andremotergainsarediscountedbytheaveragemanataveryhighrate。Thegameofprofessionalinvestmentisintolerablyboringandover-exactingtoanyonewhoisentirelyexemptfromthegamblinginstinct；
whilsthewhohasitmustpaytothispropensitytheappropriatetoll。
Furthermore，aninvestorwhoproposestoignorenear-termmarketfluctuationsneedsgreaterresourcesforsafetyandmustnotoperateonsolargeascale，ifatall，withborrowedmoney¾；afurtherreasonforthehigherreturnfromthepastimetoagivenstockofintelligenceandresources。Finallyitisthelong-terminvestor，hewhomostpromotesthepublicinterest，whowillinpracticecomeinformostcriticism，whereverinvestmentfundsaremanagedbycommitteesorboardsorbanks。[90]Foritisintheessenceofhisbehaviourthatheshouldbeeccentric，unconventionalandrashintheeyesofaverageopinion。Ifheissuccessful，thatwillonlyconfirmthegeneralbeliefinhisrashness；andif[Page158]THE
GENERALTHEORYOFEMPLOYMENT
intheshortrunheisunsuccessful，whichisverylikely，hewillnotreceivemuchmercy。Worldlywisdomteachesthatitisbetterforreputationtofailconventionallythantosucceedunconventionally。
5Sofarwehavehadchieflyinmindthestateofconfidenceofthespeculatororspeculativeinvestorhimselfandmayhaveseemedtobetacitlyassumingthat，ifhehimselfissatisfiedwiththeprospects，hehasunlimitedcommandovermoneyatthemarketrateofinterest。Thisis，ofcourse，notthecase。Thuswemustalsotakeaccountoftheotherfacetofthestateofconfidence，namely，theconfidenceofthelendinginstitutionstowardsthosewhoseektoborrowfromthem，sometimesdescribedasthestateofcredit。Acollapseinthepriceofequities，whichhashaddisastrousreactionsonthemarginalefficiencyofcapital，mayhavebeenduetotheweakeningeitherofspeculativeconfidenceorofthestateofcredit。Butwhereastheweakeningofeitherisenoughtocauseacollapse，recoveryrequirestherevivalofboth。Forwhilsttheweakeningofcreditissufficienttobringaboutacollapse，itsstrengthening，thoughanecessaryconditionofrecovery，isnotasufficientcondition。
VI
Theseconsiderationsshouldnotliebeyondthepurviewoftheeconomist。
Buttheymustberelegatedtotheirrightperspective。IfImaybeallowedtoappropriatethetermspeculationfortheactivityofforecastingthepsychologyofthemarket，andthetermenterprisefortheactivityofforecastingtheprospectiveyieldofassetsovertheirwholelife，itisbynomeansalwaysthecasethatspeculationpredominatesoverenterprise。
Astheorganisationofinvestmentmarketsimproves，theriskofthepredominanceofspeculationdoes，however，increase。Inoneofthegreatestinvestmentmarketsintheworld，namely，NewYork，the[Page159]LONG-TERM
EXPECTATION
influenceofspeculationintheabovesenseisenormous。Evenoutsidethefieldoffinance，Americansareapttobeundulyinterestedindiscoveringwhataverageopinionbelievesaverageopiniontobe；andthisnationalweaknessfindsitsnemesisinthestockmarket。Itisrare，oneistold，foranAmericantoinvest，asmanyEnglishmenstilldo，'forincome'；andhewillnotreadilypurchaseaninvestmentexceptinthehopeofcapitalappreciation。Thisisonlyanotherwayofsayingthat，whenhepurchasesaninvestment，theAmericanisattachinghishopes，notsomuchtoitsprospectiveyield，astoafavourablechangeintheconventionalbasisofvaluation，i。e。thatheis，intheabovesense，aspeculator。Speculatorsmaydonoharmasbubblesonasteadystreamofenterprise。Butthepositionisseriouswhenenterprisebecomesthebubbleonawhirlpoolofspeculation。
Whenthecapitaldevelopmentofacountrybecomesaby-productoftheactivitiesofacasino，thejobislikelytobeill-done。ThemeasureofsuccessattainedbyWallStreet，regardedasaninstitutionofwhichthepropersocialpurposeistodirectnewinvestmentintothemostprofitablechannelsintermsoffutureyield，cannotbeclaimedasoneoftheoutstandingtriumphsoflaissez-fairecapitalism¾；whichisnotsurprising，ifIamrightinthinkingthatthebestbrainsofWallStreethavebeeninfactdirectedtowardsadifferentobject。
Thesetendenciesareascarcelyavoidableoutcomeofourhavingsuccessfullyorganised'liquid'investmentmarkets。Itisusuallyagreedthatcasinosshould，inthepublicinterest，beinaccessibleandexpensive。Andperhapsthesameistrueofstockexchanges。ThatthesinsoftheLondonStockExchangearelessthanthoseofWallStreetmaybedue，notsomuchtodifferencesinnationalcharacter，astothefactthattotheaverageEnglishmanThrogmortonStreetis，comparedwithWallStreettotheaverageAmerican，inaccessibleandveryexpensive。Thejobber's'turn'，thehigh[Page160]THE
GENERALTHEORYOFEMPLOYMENT
brokeragechargesandtheheavytransfertaxpayabletotheExchequer，whichattenddealingsontheLondonStockExchange，sufficientlydiminishtheliquidityofthemarketalthoughthepracticeoffortnightlyaccountsoperatestheotherwaytoruleoutalargeproportionofthetrinsactioncharacteristicofWallStreet。[91]Theintroductionofasubstantialgovernmenttransfertaxonalltransactionsmightprovethemostserviceablereformavailable，withaviewtomitigatingthepredominanceofspeculationoverenterpriseintheUnitedStates。
Thespectacleofmoderninvestmentmarketshassometimesmovedmetowardstheconclusionthattomakethepurchaseofaninvestmentpermanentandindissoluble，likemarriage，exceptbyreasonofdeathorothergravecause，mightbeausefulremedyforourcontemporaryevils。Forthiswouldforcetheinvestortodirecthismindtothelong-termprospectsandtothoseonly。Butalittleconsiderationofthisexpedientbringsusupagainstadilemma，andshowsushowtheliquidityofinvestmentmarketsoftenfacilitates，thoughitsometimesimpedes，thecourseofnewinvestment。Forthefactthateachindividualinvestorflattershimselfthathiscommitmentis'liquid'
thoughthiscannotbetrueforallinvestorscollectivelycalmshisnervesandmakeshimmuchmorewillingtorunarisk。Ifindividualpurchasesofinvestmentswererenderedilliquid，thismightseriouslyimpedenewinvestment，solongasalternativewaysinwhichtoholdhissavingsareavailabletotheindividual。Thisisthedilemma。Solongasitisopentotheindividualtoemployhiswealthinhoardingorlendingmoney，thealternativeofpurchasingactualcapitalassetscannotberenderedsufficientlyattractiveespeciallytothemanwhodoes[Page161]LONG-TERM
EXPECTATION
notmanagethecapitalassetsandknowsverylittleaboutthem，exceptbyorganisingmarketswhereintheseassetscanbeeasilyrealisedformoney。
Theonlyradicalcureforthecrisesofconfidencewhichafflicttheeconomiclifeofthemodernworldwouldbetoallowtheindividualnochoicebetweenconsuminghisincomeandorderingtheproductionofthespecificcapital-assetwhich，eventhoughitbeonprecariousevidence，impresseshimasthemostpromisinginvestmentavailabletohim。Itmightbethat，attimeswhenhewasmorethanusuallyassailedbydoubtsconcerningthefuture，hewouldturninhisperplexitytowardsmoreconsumptionandlessnewinvestment。Butthatwouldavoidthedisastrous，cumulativeandfar-reachingrepercussionsofitsbeingopentohim，whenthusassailedbydoubts，tospendhisincomeneitherontheonenorontheother。
Thosewhohaveemphasisedthesocialdangersofthehoardingofmoneyhave，ofcourse，hadsomethingsimilartotheaboveinmind。Buttheyhaveoverlookedthepossibilitythatthephenomenoncanoccurwithoutanychange，oratleastanycommensuratechange，inthehoardingofmoney。
VII
Evenapartfromtheinstabilityduetospeculation，thereistheinstabilityduetothecharacteristicofhumannaturethatalargeproportionofourpositiveactivitiesdependonspontaneousoptimismratherthanonamathematicalexpectation，whethermoralorhedonisticoreconomic。Most，probably，ofourdecisionstodosomethingpositive，thefullconsequencesofwhichwillbedrawnoutovermanydaystocome，canonlybetakenasaresultofanimalspirits¾；ofaspontaneousurgetoactionratherthaninaction，andnotastheoutcomeofaweightedaverageofquantitativebenefitsmultipliedbyquantitativeprobabilities。Enterprise[Page162]THE
GENERALTHEORYOFEMPLOYMENT
onlypretendstoitselftobemainlyactuatedbythestatementsinitsownprospectus，howevercandidandsincere。OnlyalittlemorethananexpeditiontotheSouthPole，isitbasedonanexactcalculationofbenefitstocome。Thusiftheanimalspiritsaredimmedandthespontaneousoptimismfalters，leavingustodependonnothingbutamathematicalexpectation，enterprisewillfadeanddie；¾；thoughfearsoflossmayhaveabasisnomorereasonablethanhopesofprofithadbefore。
Itissafetosaythatenterprisewhichdependsonhopesstretchingintothefuturebenefitsthecommunityasawhole。Butindividualinitiativewillonlybeadequatewhenreasonablecalculationissupplementedandsupportedbyanimalspirits，sothatthethoughtofultimatelosswhichoftenovertakespioneers，asexperienceundoubtedlytellsusandthem，isputasideasahealthymanputsasidetheexpectationofdeath。
Thismeans，unfortunately，notonlythatslumpsanddepressionsareexaggeratedindegree，butthateconomicprosperityisexcessivelydependentonapoliticalandsocialatmospherewhichiscongenialtotheaveragebusinessman。IfthefearofaLabourGovernmentoraNewDealdepressesenterprise，thisneednotbetheresulteitherofareasonablecalculationorofaplotwithpoliticalintent；¾；itisthemereconsequenceofupsettingthedelicatebalanceofspontaneousoptimism。Inestimatingtheprospectsofinvestment，wemusthaveregard，therefore，tothenervesandhysteriaandeventhedigestionsandreactionstotheweatherofthoseuponwhosespontaneousactivityitlargelydepends。
Weshouldnotconcludefromthisthateverythingdependsonwavesofirrationalpsychology。Onthecontrary，thestateoflong-termexpectationisoftensteady，and，evenwhenitisnot，theotherfactorsexerttheircompensatingeffects。Wearemerelyremindingourselvesthathumandecisionsaffectingthefuture，whetherpersonalorpoliticaloreconomic，cannot[Page163]LONG-TERM
EXPECTATION
dependonstrictmathematicalexpectation，sincethebasisformakingsuchcalculationsdoesnotexist；andthatitisourinnateurgetoactivitywhichmakesthewheelsgoround，ourrationalselveschoosingbetweenthealternativesasbestweareable，calculatingwherewecan，butoftenfallingbackforourmotiveonwhimorsentimentorchance。
VIII
Thereare，moreover，certainimportantfactorswhichsomewhatmitigateinpracticetheeffectsofourignoranceofthefuture。Owingtotheoperationofcompoundinterestcombinedwiththelikelihoodofobsolescencewiththepassageoftime，therearemanyindividualinvestmentsofwhichtheprospectiveyieldislegitimatelydominatedbythereturnsofthecomparativelynearfuture。Inthecaseofthemostimportantclassofverylong-terminvestments，namelybuildings，theriskcanbefrequentlytransferredfromtheinvestortotheoccupier，oratleastsharedbetweenthem，bymeansoflong-termcontracts，theriskbeingoutweighedinthemindoftheoccupierbytheadvantagesofcontinuityandsecurityoftenure。Inthecaseofanotherimportantclassoflong-terminvestments，namelypublicutilities，asubstantialproportionoftheprospectiveyieldispracticallyguaranteedbymonopolyprivilegescoupledwiththerighttochargesuchratesaswillprovideacertainstipulatedmargin。Finallythereisagrowingclassofinvestmentsentereduponby，orattheriskof；publicauthorities，whicharefranklyinfluencedinmakingtheinvestmentbyageneralpresumptionoftherebeingprospectivesocialadvantagesfromtheinvestment，whateveritscommercialyieldmayprovetobewithinawiderange，andwithoutseekingtobesatisfiedthatthemathematicalexpectationoftheyieldisatleastequaltothecurrentrateofinterest，¾；thoughtheratewhichthepublic[Page164]THE
GENERALTHEORYOFEMPLOYMENT
authorityhastopaymaystillplayadecisivepartindeterminingthescaleofinvestmentoperationswhichitcanafford。
Thusaftergivingfullweighttotheimportanceoftheinfluenceofshort-periodchangesinthestateoflong-termexpectationasdistinctfromchangesintherateofinterest，wearestillentitledtoreturntothelatterasexercising，atanyrate，innormalcircumstances，agreat，thoughnotadecisive，influenceontherateofinvestment。Onlyexperience，however，canshowhowfarmanagementoftherateofinterestiscapableofcontinuouslystimulatingtheappropriatevolumeofinvestment。
FormyownpartIamnowsomewhatscepticalofthesuccessofamerelymonetarypolicydirectedtowardsinfluencingtherateofinterest。IexpecttoseetheState，whichisinapositiontocalculatethemarginalefficiencyofcapital-goodsonlongviewsandonthebasisofthegeneralsocialadvantage，takinganevergreaterresponsibilityfordirectlyorganisinginvestment；
sinceitseemslikelythatthefluctuationsinthemarketestimationofthemarginalefficiencyofdifferenttypesofcapital，calculatedontheprinciplesIhavedescribedabove，willbetoogreattobeoffsetbyanypracticablechangesintherateofinterest。
TheGeneralTheoryofInterest[Page165]
Chapter13
THEGENERALTHEORYOFTHERATEOFINTEREST
I
Wehaveshowninchapter11that，whilstthereareforcescausingtherateofinvestmenttoriseorfallsoastokeepthemarginalefficiencyofcapitalequaltotherateofinterest，yetthemarginalefficiencyofcapitalis，initself；adifferentthingfromtherulingrateofinterest。
Thescheduleofthemarginalefficiencyofcapitalmaybesaidtogovernthetermsonwhichloanablefundsaredemandedforthepurposeofnewinvestment；
whilsttherateofinterestgovernsthetermsonwhichfundsarebeingcurrentlysupplied。Tocompleteourtheory，therefore，weneedtoknowwhatdeterminestherateofinterest。
Inchapter14anditsAppendixweshallconsidertheanswerstothisquestionwhichhavebeengivenhitherto。Broadlyspeaking，weshallfindthattheymaketherateofinteresttodependontheinteractionofthescheduleofthemarginalefficiencyofcapitalwiththepsychologicalpropensitytosave。Butthenotionthattherateofinterestisthebalancingfactorwhichbringsthedemandforsavingintheshapeofnewinvestmentforthcomingataivenrateofinterestintoequalitywiththesupplyofsavingwhichresultsatthatrateofinterestfromthecommunity'spsychologicalpropensitytosave，breaksdownassoonasweperceivethatitisimpossibletodeducetherateofinterestmerelyfromaknowledgeofthesetwofactors。
[Page166]THE
GENERALTHEORYOFEMPLOYMENT
What，then，isourownanswertothisquestion？
II
Thepsychologicaltime-preferencesofanindividualrequiretwodistinctsetsofdecisionstocarrythemoutcompletely。Thefirstisconcernedwiththataspectoftime-preferencewhichIhavecalledthepropensitytoconsume，which，operatingundertheinfluenceofthevariousmotivessetforthinBookIII，determinesforeachindividualhowmuchofhisincomehewillconsumeandhowmuchhewillreserveinsomeformofcommandoverfutureconsumption。
Butthisdecisionhavingbeenmade，thereisafurtherdecisionwhichawaitshim，namely，inwhatformhewillholdthecommandoverfutureconsumptionwhichhehasreserved，whetheroutofhiscurrentincomeorfromprevioussavings。Doeshewanttoholditintheformofimmediate，liquidcommandi。e。inmoneyoritsequivalent？Orishepreparedtopartwithimmediatecommandforaspecifiedorindefiniteperiod，leavingittofuturemarketconditionstodetermineonwhattermshecan，ifnecessary，convertdeferredcommandoverspecificgoodsintoimmediatecommandovergoodsingeneral？Inotherwords，whatisthedegreeofhisliquidity-preference¾；whereanindividual'sliquidity-preferenceisgivenbyascheduleoftheamountsofhisresources，valuedintermsofmoneyorofwage-units，whichhewillwishtoretainintheformofmoneyindifferentsetsofcircumstances？
Weshallfindthatthemistakeintheacceptedtheoriesoftherateofinterestliesintheirattemptingtoderivetherateofinterestfromthefirstofthesetwoconstituentsofpsychologicaltime-preferencetotheneglectofthesecond；anditisthisneglectwhichwemustendeavourtorepair。
Itshouldbeobviousthattherateofinterestcannot[Page167]THE
GENERALTHEORYOFINTEREST
beareturntosavingorwaitingassuch。Forifamanhoardshissavingsincash，heearnsnointerest，thoughhesavesjustasmuchasbefore。
Onthecontrary，themeredefinitionoftherateofinteresttellsusinsomanywordsthattherateofinterestistherewardforpartingwithliquidityforaspecifiedperiod。Fortherateofinterestis，initself；
nothingmorethantheinverseproportionbetweenasumofmoneyandwhatcanbeobtainedforpartingwithcontroloverthemoneyinexchangeforadebt[92]forastatedperiodoftime。[93]
Thustherateofinterestatanytime，beingtherewardforpartingwithliquidity，isameasureoftheunwillingnessofthosewhopossessmoneytopartwiththeirliquidcontroloverit。Therateofinterestisnotthe'price'whichbringsintoequilibriumthedemandforresourcestoinvestwiththereadinesstoabstainfrompresentconsumption。Itisthe'price'whichequilibratesthedesiretoholdwealthintheformofcashwiththeavailablequantityofcash；¾；whichimpliesthatiftherateofinterestwerelower，i。e。iftherewardforpartingwithcashwerediminished，theaggregateamountofcashwhichthepublicwouldwishtoholdwouldexceedtheavailablesupply，andthatiftherateofinterestwereraised，therewouldbeasurplusofcashwhichnoonewouldbewillingtohold。Ifthisexplanationiscorrect，thequantityofmoneyisthe[Page168]THE
GENERALTHEORYOFEMPLOYMENT
otherfactor，which，inconjunctionwithliquidity-preference，determinestheactualrateofinterestingivencircumstances。Liquidity-preferenceisapotentialityorfunctionaltendency，whichfixesthequantityofmoneywhichthepublicwillholdwhentherateofinterestisgiven；sothatifristherateofinterest，MthequantityofmoneyandLthefunctionofliquidity-
preference，wehaveM=Lr。Thisiswhere，andhow，thequantityofmoneyentersintotheeconomicscheme。
Atthispoint，however，letusturnbackandconsiderwhysuchathingasliquidity-preferenceexists。Inthisconnectionwecanusefullyemploytheancientdistinctionbetweentheuseofmoneyforthetransactionofcurrentbusinessanditsuseasastoreofwealth。Asregardsthefirstofthesetwouses，itisobviousthatuptoapointitisworthwhiletosacrificeacertainamountofinterestfortheconvenienceofliquidity。
But，giventhattherateofinterestisnevernegative，whyshouldanyoneprefertoholdhiswealthinaformwhichyieldslittleornointeresttoholdingitinaformwhichyieldsinterestassuming，ofcourse，atthisstage，thattheriskofdefaultisthesameinrespectofabankbalanceasofabond？Afullexplanationiscomplexandmustwaitforchapter15。Thereis，however，anecessaryconditionfailingwhichtheexistenceofaliquidity-preferenceformoneyasameansofholdingwealthcouldnotexist。
Thisnecessaryconditionistheexistenceofuncertaintyastothefutureoftherateofinterest，i。e。astothecomplexofratesofinterestforvaryingmaturitieswhichwillruleatfuturedates。Foriftheratesofinterestrulingatallfuturetimescouldbeforeseenwithcertainty，allfutureratesofinterestcouldbeinferredfromthepresentratesofinterestfordebtsofdifferentmaturities，whichwouldbeadjustedtotheknowledgeofthefuturerates。Forexample，if1dristhevaluelnthepresentyear1of£；1deferredryearsanditis[Page169]THE
GENERALTHEORYOFINTEREST
knownthatndrwillbethevalueintheyearnof£；1deferredryearsfromthatdate，wehave1dnrndr=¾；¾；¾；¾；；
1dnwhenceitfollowsthattherateatwhichanydebtcanbeturnedintocashnyearshenceisgivenbytwooutofthecomplexofcurrentratesofinterest。Ifthecurrentrateofinterestispositivefordebtsofeverymaturity，itmustalwaysbemoreadvantageoustopurchaseadebtthantoholdcashasastoreofwealth。
If，onthecontrary，thefuturerateofinterestisuncertainwecannotsafelyinferthatndrwillprovetobeequalto1dnr/1dnwhenthetimecomes。Thusifaneedforliquidcashmayconceivablyarisebeforetheexpiryofnyears，thereisariskofalossbeingincurredinpurchasingalong-termdebtandsubsequentlyturningitintocash，ascomparedwithholdingcash。Theactuarialprofitormathematicalexpectationofgaincalculatedinaccordancewiththeexistingprobabilities¾；ifitcanbesocalculated，whichisdoubtful¾；mustbesufficienttocompensatefortheriskofdisappointment。
Thereis，moreover，afurthergroundforliquidity-preferencewhichresultsfromtheexistenceofuncertaintyastothefutureoftherateofinterest，providedthatthereisanorganlsedmarketfordealingindebts。Fordifferentpeoplewillestimatetheprospectsdifferentlyandanyonewhodiffersfromthepredominantopinionasexpressedinmarketquotationsmayhaveagoodreasonforkeepingliquidresourcesinordertoprofit，ifheisright，fromitsturningoutinduecoursethatthe1dr'swereinamistakenrelationshiptooneanother。[94]
Thisiscloselyanalogoustowhatwehavealready[Page170]THE
GENERALTHEORYOFEMPLOYMENT
discussedatsomelengthinconnectionwiththemarginalefficiencyofcapital。Justaswefoundthatthemarginalefficiencyofcapitalisfixed，notbythe'best'opinion，butbythemarketvaluationasdeterminedbymasspsychology，soalsoexpectationsastothefutureoftherateofinterestasfixedbymasspsychologyhavetheirreactionsonliquidity-preference；¾；butwiththisadditionthattheindividual，whobelievesthatfutureratesofinterestwillbeabovetheratesassumedbythemarket，hasareasonforkeepingactualliquidcash，[95]whilsttheindividualwhodiffersfromthemarketintheotherdirectionwillhaveamotiveforborrowingmoneyforshortperiodsinordertopurchasedebtsoflongerterm。Themarketpricewillbefixedatthepointatwhichthesalesofthe'bears'
andthepurchasesofthe'bulls'arebalanced。
Thethreedivisionsofliquidity-preferencewhichwehavedistinguishedabovemaybedefinedasdependingonithetransactions-motive，i。e。
theneedofcashforthecurrenttransactionofpersonalandbusinessexchanges；
iitheprecautionary-motive，i。e。thedesireforsecurityastothefuturecashequivalentofacertainproportionoftotalresources；andiiithespeculative-motive，i。e。theobjectofsecuringprofitfromknowingbetterthanthemarketwhatthefuturewillbringforth。Aswhenwewerediscussingthemarginalefficiencyofcapital，thequestionofthedesirabilityofhavingahighlyorganisedmarketfordealingwithdebtspresentsuswithadilemma。For，intheabsenceofanorganisedmarket，liquidity-preferenceduetotheprecautionary-motivewouldbegreatlyincreased；whereastheexistenceofanorganisedmarketgivesan[Page171]THE
GENERALTHEORYOFINTEREST
opportunityforwidefluctuationsinliquidity-preferenceduetothespeculative-motive。
Itmayillustratetheargumenttopointoutthat，iftheliquidity-preferencesduetothetransactions-motiveandtheprecautionary-motiveareassumedtoabsorbaquantityofcashwhichisnotverysensitivetochangesintherateofinterestassuchandapartfromitsreactionsonthelevelofincome，sothatthetotalquantityofmoney，lessthisquantity，isavailableforsatisfyingliquidity-preferencesduetothespeculative-motive，therateofinterestandthepriceofbondshavetobefixedatthelevelatwhichthedesireonthepartofcertainindividualstoholdcashbecauseatthatleveltheyfeel'bearish'ofthefutureofbondsisexactlyequaltotheamountofcashavailableforthespeculative-motive。Thuseachincreaseinthequantityofmoneymustraisethepriceofbondssufficientlytoexceedtheexpectationsofsome'bull'andsoinfluencehimtosellhisbondforcashandjointhe'bear'brigade。If，however，thereisanegligibledemandforcashfromthespeculative-motiveexceptforashorttransitionalinterval，anincreaseinthequantityofmoneywillhavetolowertherateofinterestalmostforthwith，inwhateverdegreeisnecessarytoraiseemploymentandthewage-unitsufficientlytocausetheadditionalcashtobeabsorbedbythetransactions-motiveandtheprecautionary-motive。
Asarule，wecansupposethatthescheduleofliquidity-preferencerelatingthequantityofmoneytotherateofinterestisgivenbyasmoothcurvewhichshowstherateofinterestfallingasthequantityofmoneyisincreased。Forthereareseveraldifferentcausesallleadingtowardsthisresult。
Inthefirstplace，astherateofinterestfalls，itislikely，cet。
par。，thatmoremoneywillbeabsorbedbyliquidity-preferencesduetothetransactions-motive。Forifthefallintherateofinterestincreasesthenationalincome，theamountofmoneywhichitisconvenientto[Page172]THE
GENERALTHEORYOFEMPLOYMENT
keepfortransactionswillbeincreasedmoreorlessproportionatelytotheincreaseinincome；whilst，atthesametime，thecostoftheconvenienceofplentyofreadycashintermsoflossofinterestwillbediminished。
Unlesswemeasureliquidity-preferenceintermsofwage-unitsratherthanofmoneywhichisconvenientinsomecontexts，similarresultsfollowiftheincreasedemploymentensuingonafallintherateofinterestleadstoanincreaseofwages，i。e。toanincreaseinthemoneyvalueofthewage-unit。Inthesecondplace，everyfallintherateofinterestmay，aswehavejustseen，increasethequantityofcashwhichcertainindividualswillwishtoholdbecausetheirviewsastothefutureoftherateofinterestdifferfromthemarketviews。
Nevertheless，circumstancescandevelopinwhichevenalargeincreaseinthequantityofmoneymayexertacomparativelysmallinfluenceontherateofinterest。Foralargeincreaseinthequantityofmoneymaycausesomuchuncertaintyaboutthefuturethatliquidity-preferencesduetotheprecautionary-motivemaybestrengthened；whilstopinionaboutthefutureoftherateofinterestmaybesounanimousthatasmallchangeinpresentratesmaycauseamassmovementintocash。Itisinterestingthatthestabilityofthesystemanditssensitivenesstochangesinthequantityofmoneyshouldbesodependentontheexistenceofavarietyofopinionaboutwhatisuncertain。Bestofallthatweshouldknowthefuture。Butifnot，then，ifwearetocontroltheactivityoftheeconomicsystembychangingthequantityofmoney，itisimportantthatopinionsshoulddifferThusthismethodofcontrolismoreprecariousintheUnitedStates，whereeveryonetendstoholdthesameopinionatthesametime，thaninEnglandwheredifferencesofopinionaremoreusual。
[Page173]THE
GENERALTHEORYOFINTEREST
III
Wehavenowintroducedmoneyintoourcausalnexusforthefirsttime，andweareabletocatchafirstglimpseofthewayinwhichchangesinthequantityofmoneyworktheirwayintotheeconomicsystem。If，however，wearetemptedtoassertthatmoneyisthedrinkwhichstimulatesthesystemtoactivity，wemustremindourselvesthattheremaybeseveralslipsbetweenthecupandthelip。Forwhilstanincreaseinthequantityofmoneymaybeexpected，cet。par。，toreducetherateofinterest，thiswillnothappeniftheliquidity-preferencesofthepublicareincreasingmorethanthequantityofmoney；andwhilstadeclineintherateofinterestmaybeexpected，cet。par。，toincreasethevolumeofinvestment，thiswillnothappenifthescheduleofthemarginalefficiencyofcapitalisfallingmorerapidlythantherateofintere~t；andwhilstanincreaseinthevolumeofinvestmentmaybeexpected，cet。par。，toincreaseemployment，thismaynothappenifthepropensitytoconsumeisfallingoff。Finally，ifemploymentincreases，priceswillriseinadegreepartlygovernedbytheshapesofthephysicalsupplyfunctions，andpartlybytheliabilityofthewage-unittoriseintermsofmoney。Andwhenoutputhasincreasedandpriceshaverisen，theeffectofthisonliquidity-preferencewillbetoincreasethequantityofmoneynecessarytomaintainagivenrateofinterest。
IV
Whilstliquidity-preferenceduetothespeculative-motivecorrespondstowhatinmyTreatiseonMoneyIcalled'thestateofbearishness'，itisbynomeansthesamething。For'bearishness'istheredefinedasthefunctionalrelationship，notbetweentherateofinterestorpriceofdebtsandthequantityofmoney，butbetweenthepriceofassetsanddebts，takentogether，[Page174]THE
GENERALTHEORYOFEMPLOYMENT
andthequantityofmoney。Thistreatment，however，involvedaconfusionbetweenresultsduetoachangeintherateofinterestandthoseduetoachangeinthescheduleofthemarginalefficiencyofcapital，whichI
hopeIhavehereavoided。
V
Theconceptofhoardingmayberegardedasafirstapproximationtotheconceptofliquidity-preference。Indeedifweweretosubstitute'propensitytohoard'for'hoarding'，itwouldcometosubstantiallythesamething。Butifwemeanby'hoarding'anactualincreaseincash-holding，itisanincompleteidea¾；andseriouslymisleadingifitcausesustothinkof'hoarding'and'not-hoarding'assimplealternatives。Forthedecisiontohoardisnottakenabsolutelyorwithoutregardtotheadvantagesofferedforpartingwithliquidity；¾；itresultsfromabalancingofadvantages，andwehave，therefore，toknowwhatliesintheotherscale。Moreoveritisimpossiblefortheactualamountofhoardingtochangeasaresultofdecisionsonthepartofthepublic，solongaswemeanby'hoarding'theactualholdingofcash。Fortheamountofhoardingmustbeequaltothequantityofmoneyor¾；onsomedefinitions¾；tothequantityofmoneyminuswhatisrequiredtosatisfythetransactions-motive；andthequantityofmoneyisnotdeterminedbythepublic。Allthatthepropensityofthepublictowardshoardingcanachieveistodeterminetherateofinterestatwhichtheaggregatedesiretohoardbecomesequaltotheavailablecash。Thehabitofoverlookingtherelationoftherateofinteresttohoardingmaybeapartoftheexplanationwhyinteresthasbeenusuallyregardedastherewardofnot-spending，whereasinfactitistherewardofnot-hoarding。
TheClassicalTheoryofInterest[Page175]
Chapter14
THECLASSICALTHEORYOFTHERATEOFINTEREST
I
Whatistheclassicaltheoryoftherateofinterest？Itissomethinguponwhichwehaveallbeenbroughtupandwhichwehaveacceptedwithoutmuchreserveuntilrecently。YetIfinditdifficulttostateitpreciselyortodiscoveranexplicitaccountofitintheleadingtreatisesofthemodernclassicalschool。[96]
Itisfairlyclear，however，thatthistraditionhasregardedtherateofinterestasthefactorwhichbringsthedemandforinvestmentandthewillingnesstosaveintoequilibriumwithoneanother。Investmentrepresentsthedemandforinvestibleresourcesandsavingrepresentsthesupply，whilsttherateofinterestisthe'price'ofinvestibleresourcesatwhichthetwoareequated。Justasthepriceofacommodityisnecessarilyfixedatthatpointwherethedemandforitisequaltothesupply，sotherateofinterestnecessarilycomestorestundertheplayofmarketforcesatthepointwheretheamountofinvestmentatthatrateofinterestisequaltotheamountofsavingatthatrate。
TheaboveisnottobefoundinMarshall'sPrinciplesinsomanywords。Yethistheoryseemstobethis，anditiswhatImyselfwasbroughtuponandwhatItaughtformanyyearstoothers。Take，forexample，thefollowingpassagefromhisPrinciples:'Interest，beingthepricepaidfortheuseofcapitalinany[Page176]THE
GENERALTHEORYOFEMPLOYMENT
market，tendstowardsanequilibriumlevelsuchthattheaggregatedemandforcapitalinthatmarket，atthatrateofinterest，isequaltotheaggregatestockforthcomingatthatrate'。[97]OragaininProfessorCassel'sNatureandNecessityofInterestitisexplainedthatinvestmentconstitutesthe'demandforwaiting'andsavingthe'supplyofwaiting'，whilstinterestisa'price'whichserves，itisimplied，toequatethetwo，thoughhereagainIhavenotfoundactualwordstoquote。ChapterviofProfessorCarver'sDistributionofWealthclearlyenvisagesinterestasthefactorwhichbringsintoequilibriumthemarginaldisutilityofwaitingwiththemarginalproductivityofcapital。[98]SirAlfredFluxEconomicPrinciples，p。95writes:'Ifthereisjusticeinthecontentionsofourgeneraldiscussion，itmustbeadmittedthatanautomaticadjustmenttakesplacebetweensavingandtheopportunitiesforemployingcapitalprofitably……Savingwillnothaveexceededitspossibilitiesofusefulness……solongastherateofnetinterestisinexcessofzero。'ProfessorTaussigPrinciples，vol。ii。p。29drawsasupplycurveofsavingandademandcurverepresenting'thediminishingproductivenessoftheseveralinstalmentsofcapital'，havingpreviouslystatedp。20
that'therateofinterestsettlesatapointwherethemarginalproductivityofcapitalsufficestobringoutthemarginalinstalmentofsaving'。[
99]Walras，in[Page177]THE
CLASSICALTHEORYOFINTEREST
AppendixIIIIofhisÉ；lé；mentsd'é；conomiepure，wherehedealswith'l'é；changed'é；pargnescontrecapitauxneufs'，arguesexpresslythat，correspondingtoeachpossiblerateofinterest，thereisasumwhichindividualswillsaveandalsoasumwhichtheywillinvestinnewcapitalassets，thatthesetwoaggregatestendtoequalitywithoneanother，andthattherateofinterestisthevariablewhichbringsthemtoequality；sothattherateofinterestisfixedatthepointwheresaving，whichrepresentsthesupplyofnewcapital，isequaltothedemandforit。Thusheisstrictlyintheclassicaltradition。
Certainlytheordinaryman¾；banker，civilservantorpolitician¾；broughtuponthetraditionaltheory，andthetrainedeconomistalso，hascarriedawaywithhimtheideathatwheneveranindividualperformsanactofsavinghehasdonesomethingwhichautomaticallybringsdowntherateofinterest，thatthisautomaticallystimulatestheoutputofcapital，andthatthefallintherateofinterestisjustsomuchasisnecessarytostimulatetheoutputofcapitaltoanextentwhichisequaltotheincrementofsaving；
and，further，thatthisisaself-regulatoryprocessofadjustmentwhichtakesplacewithoutthenecessityforanyspecialinterventionorgrandmotherlycareonthepartofthemonetaryauthority。Similarly¾；andthisisanevenmoregeneralbelief，evento-day¾；eachadditionalactofinvestmentwillnecessarilyraisetherateofinterest，ifitisnotoffsetbyachangeinthereadinesstosave。
Nowtheanalysisofthepreviouschapterswillhavemadeitplainthatthisaccountofthemattermustbeerroneous。Intracingtoitssourcethereasonforthedifferenceofopinion，letus，however，beginwiththematterswhichareagreed。
Unliketheneo-classicalschool，whobelievethatsavingandinvestmentcanbeactuallyunequal，theclassicalschoolproperhasacceptedtheviewthattheyareequal。Marshall，forexample，surelybelieved，[Page178]THE
GENERALTHEORYOFEMPLOYMENT
althoughhedidnotexpresslysayso，thataggregatesavingandaggregateinvestmentarenecessarilyequal。Indeed，mostmembersoftheclassicalschoolcarriedthisbeliefmuchtoofar；sincetheyheldthateveryactofincreasedsavingbyanindividualnecessarilybringsintoexistenceacorrespondingactofincreasedinvestment。Noristhereanymaterialdifference，relevantinthiscontext，betweenmyscheduleofthemarginalefficiencyofcapitalorinvestmentdemand-scheduleandthedemandcurveforcapitalcontemplatedbysomeoftheclassicalwriterswhohavebeenquotedabove。Whenwecometothepropensitytoconsumeanditscorollarythepropensitytosave，wearenearertoadifferenceofopinion，owingtotheemphasiswhichtheyhaveplacedontheinfluenceoftherateofinterestonthepropensitytosave。Buttheywould，presumably，notwishtodenythatthelevelofincomealsohasanimportantinfluenceontheamountsaved；whilstI，formypart，wouldnotdenythattherateofinterestmayperhapshaveaninfluencethoughperhapsnotofthekindwhichtheysupposeontheamountsavedoutofagivenincome。AllthesepointsofagreementcanbesummedupinapropositionwhichtheclassicalschoolwouldacceptandIshouldnotdispute；namely，that，ifthelevelofincomeisassumedtobegiven，wecaninferthatthecurrentrateofinterestmustlieatthepointwherethedemandcurveforcapitalcorrespondingtodifferentratesofinterestcutsthecurveoftheamountssavedoutofthegivenincomecorrespondingtodifferentratesofinterest。
Butthisisthepointatwhichdefiniteerrorcreepsintotheclassicaltheory。Iftheclassicalschoolmerelyinferredfromtheabovepropositionthat，giventhedemandcurveforcapitalandtheinfluenceofchangesintherateofinterestonthereadinesstosaveoutofgivenincomes，thelevelofincomeandtherateofinterestmustbeuniquelycorrelated，therewouldbenothingtoquarrelwith。Moreover，thisproposition[Page179]THE
CLASSICALTHEORYOFINTEREST
wouldleadnaturallytoanotherpropositionwhichembodiesanimportanttruth；namely，that，iftherateofinterestisgivenaswellasthedemandcurveforcapitalandtheinfluenceoftherateofinterestonthereadinesstosaveoutofgivenlevelsofincome，thelevelofincomemustbethefactorwhichbringstheamountsavedtoequalitywiththeamountinvested。
But，infact，theclassicaltheorynotmerelyneglectstheinfluenceofchangesinthelevelofincome，butinvolvesformalerror。
Fortheclassicaltheory，ascanbeseenfromtheabovequotations，assumesthatitcanthenproceedtoconsidertheeffectontherateofinterestofe。g。ashiftinthedemandcurveforcapital，withoutabatingormodifyingitsassumptionastotheamountofthegivenincomeoutofwhichthesavingsaretobemade。Theindependentvariablesoftheclassicaltheoryoftherateofinterestarethedemandcurveforcapitalandtheinfluenceoftherateofinterestontheamountsavedoutofagivenincome；
andwhene。g。thedemandcurveforcapitalshifts，thenewrateofinterest，accordingtothistheory，isgivenbythepointofintersectionbetweenthenewdemandcurveforcapitalandthecurverelatingtherateofinteresttotheamountswhichwillbesavedoutofthegivenincome。Theclassicaltheoryoftherateofinterestseemstosupposethat，ifthedemandcurveforcapitalshiftsorifthecurverelatingtherateofinteresttotheamountssavedoutofagivenincomeshiftsorifboththesecurvesshift，thenewrateofinterestwillbegivenbythepointofintersectionofthenewpositionsofthetwocurves。Butthisisanonsensetheory。Fortheassumptionthatincomeisconstantisinconsistentwiththeassumptionthatthesetwocurvescanshiftindependentlyofoneanother。Ifeitherofthemshift，then，ingeneral，incomewillchange；withtheresultthatthewholeschematismbasedontheassumptionofagivenincomebreaksdown。
Thepositioncouldonlybesavedbysome[Page180]THE
GENERALTHEORYOFEMPLOYMENT
complicatedassumptionprovidingforanautomaticchangeinthewage-unitofanamountjustsufficientinitseffectonliquidity-preferencetoestablisharateofinterestwhichwouldjustoffsetthesupposedshift，soastoleaveoutputatthesamelevelasbefore。Infact，thereisnohinttobefoundintheabovewritersastothenecessityforanysuchassumption；
atthebestitwouldbeplausibleonlyinrelationtolong-periodequilibriumandcouldnotformthebasisofashort-periodtheory；andthereisnogroundforsupposingittoholdeveninthelong-period。Intruth，theclassicaltheoryhasnotbeenalivetotherelevanceofchangesinthelevelofincomeortothepossibilityofthelevelofincomebeingactuallyafunctionoftherateoftheinvestment。
Theabovecanbeillustratedbyadiagram[100]asfollows:
InthisdiagramtheamountofinvestmentorsavingIismeasuredvertically，andtherateofinterestrhorizontally。X1X1'
isthefirstpositionoftheinvestmentdemand-schedule，andX2X2'
isasecondpositionofthiscurve。ThecurveY1relatesthe[Page181]THE
CLASSICALTHEORYOFINTEREST
amountssavedoutofanincomeY1tovariouslevelsoftherateofinterest，thecurvesY2，Y3，etc。，beingthecorrespondingcurvesforlevelsofincomeY2，Y3，etc。LetussupposethatthecurveY1istheY-curveconsistentwithaninvestmentdemand-scheduleX1X1'
andarateofinterestr1。Nowiftheinvestmentdemand-scheduleshiftsfromX1X1'toX2X2'，incomewill，ingeneral，shiftalso。Buttheabovediagramdoesnotcontainenoughdatatotelluswhatitsnewvaluewillbe；and，therefore，notknowingwhichistheappropriateY-curve，wedonotknowatwhatpointthenewinvestmentdemand-schedulewillcutit。If，however，weintroducethestateofliquidity-preferenceandthequantityofmoneyandthesebetweenthemtellusthattherateofinterestisr2，thenthewholepositionbecomesdeterminate。FortheY-curvewhichintersectsX2X2'atthepointverticallyabover2，namely，thecurveY2，willbetheappropriatecurve。ThustheX-curveandtheY-curvestellusnothingabouttherateofinterest。Theyonlytelluswhatincomewillbe，iffromsomeothersourcewecansaywhattherateofinterestis。
Ifnothinghashappenedtothestateofliquidity-preferenceandthequantityofmoney，sothattherateofinterestisunchanged，thenthecurveY2'
whichintersectsthenewinvestmentdemand-scheduleverticallybelowthepointwherethecurveY1intersectedtheoldinvestmentdemand-schedulewillbetheappropriateY-curve，andY2'
willbethenewlevelofincome。
Thusthefunctionsusedbytheclassicaltheory，namely，theresponseofinvestmentandtheresponseoftheamountsavedoutofagivenincometochangeintherateofinterest，donotfurnishmaterialforatheoryoftherateofinterest；buttheycouldbeusedtotelluswhatthelevelofincomewillbe，givenfromsomeothersourcetherateofinterest；
and，alternatively，whattherateofinterestwillhavetobe，ifthelevelofincomeistobemaintainedatagivenfiguree。g。thelevelcorrespondingtofullemployment。
[Page182]THE
GENERALTHEORYOFEMPLOYMENT
Themistakeoriginatesfromregardinginterestastherewardforwaitingassuch，insteadofastherewardfornot-hoarding；justastheratesofreturnonloansorinvestmentsinvolvingdifferentdegreesofrisk，arequiteproperlyregardedasthereward，notofwaitingassuch，butofrunningtherisk。Thereis，intruth，nosharplinebetweentheseandtheso-called'pure'rateofinterest，allofthembeingtherewardforrunningtheriskofuncertaintyofonekindoranother。Onlylntheeventofmoneybeingusedsolelyfortransactionsandneverasastoreofvalue，wouldadifferenttheorybecomeappropriate。[101]
Thereare，however，twofamiliarpointswhichmight，perhaps，havewarnedtheclassicalschoolthatsomethingwaswrong。Inthefirstplace，ithasbeenagreed，atanyratesincethepublicationofProfessorCassel'sNatureandNecessityofInterest，thatitisnotcertainthatthesumsavedoutofagivenincomenecessarilyincreaseswhentherateofinterestisincreased；whereasnoonedoubtsthattheinvestmentdemand-schedulefallswitharisingrateofinterest。ButiftheY-curvesandtheX-curvesbothfallastherateofinterestrises，thereisnoguaranteethatagivenY-curvewillintersectagivenX-curveanywhereatall。ThissuggeststhatitcannotbetheY-curveandtheX-curvealonewhichdeterminetherateofinterest。
Inthesecondplace，ithasbeenusualtosupposethatanincreaseinthequantityofmoneyhasatendencytoreducetherateofinterest，atanyrateinthefirstinstanceandintheshortperiod。Yetnoreasonhasbeengivenwhyachangeinthequantityofmoneyshouldaffecteithertheinvestmentdemand-scheduleorthereadinesstosaveoutofagivenincome。
ThustheclassicalschoolhavehadquiteadifferenttheoryoftherateofinterestinvolumeIdealingwiththetheoryofvaluefromwhattheyhavehadinvolumeIIdealingwiththetheoryofmoney。Theyhaveseemedun-
[Page183]THE
CLASSICALTHEORYOFINTEREST
disturbedbytheconflictandhavemadenoattempt，sofarasIknow，tobuildabridgebetweenthetwotheories。Theclassicalschoolproper，thatistosay；sinceitistheattempttobuildabridgeonthepartoftheneo-classicalschoolwhichhasledtotheworstmuddlesofall。Forthelatterhaveinferredthattheremustbetwosourcesofsupplytomeettheinvestmentdemand-schedule；namely，savingsproper，whicharethesavingsdealtwithbytheclassicalschool，plusthesummadeavailablebyanyincreaseinthequantityofmoneythisbeingbalancedbysomespeciesoflevyonthepublic，called'forcedsaving'orthelike。
Thisleadsontotheideathatthereisa'natural'or'neutral'[102
]orequilibrium'rateofinterest，namely，thatrateofinterestwhichequatesinvestmenttoclassicalsavingsproperwithoutanyadditionfrom'forcedsavings'；andfinallytowhat，assumingtheyareontherighttrackatthestart，isthemostobvioussolutionofall，namely，that，ifthequantityofmoneycouldonlybekeptconstantinallcircumstances，noneofthesecomplicationswouldarise，sincetheevilssupposedtoresultfromthesupposedexcessofinvestmentoversavingsproperwouldceasetobepossible。Butatthispointweareindeepwater。'Thewildduckhasdiveddowntothebottom¾；asdeepasshecanget¾；andbittenfastholdoftheweedandtangleandalltherubbishthatisdownthere，anditwouldneedanextraordinarilycleverdogtodiveafterandfishherupagain。'
Thusthetraditionalanalysisisfaultybecauseithasfailedtoisolatecorrectlytheindependentvariablesofthesystem。Savingandinvestmentarethedeterminatesofthesystem，notthedeterminants。Theyarethetwinresultsofthesystem'sdeterminants，namely，thepropensitytoconsume，thescheduleofthemarginalefficiencyofcapitalandtherateofinterest。
These[Page184]THE
GENERALTHEORYOFEMPLOYMENT
determinantsare，indeed，themselvescomplexandeachiscapableofbeingaffectedbyprospectivechangesintheothers。Buttheyremainindependentinthesensethattheirvaluescannotbeinferredfromoneanother。Thetraditionalanalysishasbeenawarethatsavingdependsonincomebutithasoverlookedthefactthatincomedependsoninvestment，insuchfashionthat，wheninvestmentchanges，incomemustnecessarilychangeinjustthatdegreewhichisnecessarytomakethechangeinsavingequaltothechangeininvestment。
Norarethosetheoriesmoresuccessfulwhichattempttomaketherateofinterestdependon'themarginalefficiencyofcapital'。Itistruethatinequilibriumtherateofinterestwillbeequaltothemarginalefficiencyofcapital，sinceitwillbeprofitabletoincreaseordecrease
thecurrentscaleofinvestmentuntilthepointofequalityhasbeenreached。
Buttomakethisintoatheoryoftherateofinterestortoderivetherateofinterestfromitinvolvesacircularargument，asMarshalldiscoveredafterhehadgothalf-wayintogivinganaccountoftherateofinterestalongtheselines。[103]Forthe'marginalefficiencyofcapital'partlydependsonthescaleofcurrentinvestment，andwemustalreadyknowtherateofinterestbeforewecancalculatewhatthisscalewillbe。Thesignificantconclusionisthattheoutputofnewinvestmentwillbepushedtothepointatwhichthemarginalefficiencyofcapitalbecomesequaltotherateofinterest；andwhatthescheduleofthemarginalefficiencyofcapitaltellsus，is，notwhattherateofinterestis，butthepointtowhichtheoutputofnewinvestmentwillbepushed，giventherateofinterest。
Thereaderwillreadilyappreciatethattheproblemhereunderdiscussionisamatterofthemostfundamentaltheoreticalsignificanceandofoverwhelmingpracticalimportance。Fortheeconomicprinciple，[Page185]THE
CLASSICALTHEORYOFINTEREST
onwhichthepracticaladviceofeconomistshasbeenalmostinvariablybased，hasassumed，ineffect，that，cet。par。，adecreaseinspendingwilltendtolowertherateofinterestandanincreaseininvestmenttoraiseit。Butifwhatthesetwoquantitiesdetermineis，nottherateofinterest，buttheaggregatevolumeofemployment，thenouroutlookonthemechanismoftheeconomicsystemwillbeprofoundlychanged。AdecreasedreadinesstospendwillbelookedoninquiteadifferentlightIf，insteadofbeingregardedasafactorwhichwill，cet。par。，increaseinvestment，itisseenasafactorwhichwill，cet。par。，diminishemployment。
AppendixtoChapter14[Page186]
AppendixtoChapter14
APPENDIXONTHERATEOFINTERESTINMARSHALL'S
PRINCIPLESOFECONOMICS，RICARDO'SPRINCIPLES
OFPOLITICALECONOMY，ANDELSEWHERE
I
ThereisnoconsecutivediscussionoftherateofinterestintheworksofMarshall，EdgeworthorProfessorPigou，¾；nothingmorethanafewobiterdicta。Apartfromthepassagealreadyquotedabovep。139theonlyimportantcluestoMarshall'spositionontherateofinterestaretobefoundinhisPrinciplesofEconomics6thedn。，BookVI。p。534andp。593，thegistofwhichisgivenbythefollowingquotations:
'Interest，beingthepricepaidfortheuseofcapitalinanymarket，tendstowardsanequilibriumlevelsuchthattheaggregatedemandforcapitalinthatmarket，atthatrateofinterest，isequaltotheaggregatestock[104]forthcomingthereatthatrate。Ifthemarket，whichweareconsidering，isasmallone¾；sayasingletown，orasingletradeinaprogressivecountry¾；anincreaseddemandforcapitalinitwillbepromptlymetbyanincreasedsupplydrawnfromsurroundingdistrictsortrades。Butifweareconsideringthewholeworld，oreventhewholeofalargecountry，asonemarketforcapital，wecannotregardtheaggregatesupplyofitasalteredquicklyandtoaconsiderableextentbyachangeintherateofinterest。Forthegeneralfundofcapitalistheproductoflabourandwaiting；andtheextra[Page187]APPENDIX
TOCHAPTER14
work，[105]andtheextrawaiting，towhichariseintherateofinterestwouldactasanincentive，wouldnotquicklyamounttomuch，ascomparedwiththeworkandwaiting，ofwhichthetotalexistingstockofcapitalistheresult。Anextensiveincreaseinthedemandforcapitalingeneralwillthereforebemetforatimenotsomuchbyanincreaseofsupply，asbyariseintherateofinterest；[106]whichwillcausecapitaltowithdrawitselfpartiallyfromthoseusesinwhichitsmarginalutilityislowest。Itisonlyslowlyandgraduallythattheriseintherateofinterestwillincreasethetotalstockofcapital'p。534。
'Itcannotberepeatedtoooftenthatthephrase"therateofinterest"
isapplicabletooldinvestmentsofcapitalonlyinaverylimitedsense。
[107]Forinstance，wemayperhapsestimatethatatradecapitalofsomeseventhousandmillionsisinvestedinthedifferenttradesofthiscountryatabout3percentnetinterest。Butsuchamethodofspeaking，thoughconvenientandjustifiableformanypurposes，isnotaccurate。Whatought[Page188]THE
GENERALTHEORYOFEMPLOYMENT
tobesaidisthat，takingtherateofnetinterestontheinvestmentsofnewcapitalineachofthosetrades[i。e。onmarginalinvestments]tobeabout3percent；thentheaggregatenetincomerenderedbythewholeofthetrade-capitalinvestedinthevarioustradesissuchthat，ifcapitalisedat33years'purchasethatis，onthebasisofinterestat3percent，itwouldamounttosomeseventhousandmillionpounds。Forthevalueofthecapitalalreadyinvestedinimprovinglandorerectingabuilding，inmakingarailwayoramachine，istheaggregatediscountedvalueofitsestimatedfuturenetincomes[orquasi-rents]；andifitsprospectiveincome-yieldingpowershoulddiminish，itsvaluewouldfallaccordinglyandwouldbethecapitalisedvalueofthatsmallerincomeafterallowingfordepreciation'p。593。
InhisEconomicsofWelfare3rdedn。，p。163，ProfessorPigouwrites:'Thenatureoftheserviceof"waiting"hasbeenmuchmisunderstood。
Sometimesithasbeensupposedtoconsistintheprovisionofmoney，sometimesintheprovisionoftime，and，onbothsuppositions，ithasbeenarguedthatnocontributionwhateverismadebyittothedividend。Neithersuppositioniscorrect。"Waiting"simplymeanspostponingconsumptionwhichapersonhaspowertoenjoyimmediately，thusallowingresources，whichmighthavebeendestroyed，toassumetheformofproductioninstruments[108]……Theunitof"waiting"is，therefore，theuseofagivenquantityofresources[109]¾；forexample，labourormachinery¾；foragiventime……Inmoregeneraltermswemaysaythattheunitofwaitingisayear-value-unit，or，inthesimpler，iflessaccurate，languageofDrCassel，ayear-pound……Acautionmaybeaddedagainstthecommonviewthattheamountofcapitalaccumulatedinanyyearisnecessarilyequaltotheamountof"savings"madeinit。Thisisnotso，evenwhensavingsareinterpretedtomeannetsavings，thuseliminatingthesavingsofonemanthatarelenttoincreasetheconsumptionofanother，andwhentemporaryaccumulationsofunusedclaimsuponservicesintheformofbank-moneyareignored；formanysavingswhicharemeanttobecomecapital[Page189]APPENDIX
TOCHAPTER14
infactfailoftheirpurposethroughmisdirectionintowastefuluses。'
[110]
ProfessorPigou'sonlysignificantreferencetowhatdeterminestherateofinterestis，Ithink，tobefoundinhisIndustrialFluctuations1stedn。，pp。251-3，wherehecontrovertstheviewthattherateofinterest，beingdeterminedbythegeneralconditionsofdemandandsupplyofrealcapital，liesoutsidethecentraloranyotherbank'scontrol。Againstthisviewhearguesthat:'Whenbankerscreatemorecreditforbusinessmen，theymake，intheirinterest，subjecttotheexplanationsgiveninchapterxiii。ofparti。，[111]aforcedlevyofrealthingsfromthepublic，thusincreasingthestreamofrealcapitalavailableforthem，andcausingafallintherealrateofinterestonlongandshortloansalike。Itistrue，inshort，thatthebankers'rateformoneyisboundbyamechanicaltietotherealrateofinterestonlongloans；butitisnottruethatthisrealrateisdeterminedbyconditionswhollyoutsidebankers'control。'
Myrunningcommentsontheabovehavebeenmadeinthefootnotes。TheperplexitywhichIfindinMarshall'saccountofthematterisfundamentallydue，Ithink，totheincursionoftheconcept'interest'，whichbelongstoamonetaryeconomy，intoatreatisewhichtakesnoaccountofmoney。
'Interest'hasreallynobusinesstoturnupatallinMarshall'sPrinciplesofEconomics，¾；itbelongstoanotherbranchofthesubject。
[Page190]THE
GENERALTHEORYOFEMPLOYMENT
ProfessorPigou，conformablywithhisothertacitassumptions，leadsusinhisEconomicsofWelfaretoinferthattheunitofwaitingisthesameastheunitofcurrentinvestmentandthattherewardofwaitingisquasi-rent，andpracticallynevermentionsinterest，whichisasitshouldbe。Neverthelessthesewritersarenotdealingwithanon-monetaryeconomyifthereissuchathing。Theyquiteclearlypresumethatmoneyisusedandthatthereisabankingsystem。Moreover，therateofinterestscarcelyplaysalargerpartinProfessorPigou'sIndustrialFluctuationswhichismainlyastudyoffluctuationsinthemarginalefficiencyofcapitalorinhisTheoryofUnemploymentwhichismainlyastudyofwhatdetermineschangesinthevolumeofemployment，assumingthatthereisnoinvoluntaryunemploymentthaninhisEconomicsofWelfare。