Classical, Loanable-Fund, and Keynesian Interest Theories

Author(s): Alvin H. Hansen

Source: The Quarterly Journal of Economics, Vol. 65, No. 3 (Aug., 1951), pp. 429-432
Published by: Oxford University Press
Stable URL: http://www.jstor.org/stable/1882223 .
Accessed: 18/06/2014 22:41
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to The Quarterly
Journal of Economics.

http://www.jstor.org

This content downloaded from 195.78.109.96 on Wed, 18 Jun 2014 22:41:42 PM

All use subject to JSTOR Terms and Conditions

CLASSICAL, LOANABLE-FUND,
AND KEYNESIAN INTEREST THEORIES
By

ALVIN

H. HANSEN

Keynes attacked the classical theoryof intereston the ground

that it is indeterminate. Accordingto classical theorythe rate is
determinedby the intersectionof the investmentdemand-schedule
and the saving-schedule- schedulesdisclosingthe relationofinvestment and saving to the rate of interest. No solution,however,is
possible because the position of the saving-schedulewill vary with
the level of real income. As income rises,the schedule will shiftto
the right. Thus we cannot know what the rate of interestwill be
unlesswe already know the incomelevel. And we cannot know the
income level withoutalready knowingthe rate of interest,since a
lowerinterestrate will mean a largervolume of investment,and so,
a higherlevel of real income.The classicalanalysis,
via themultiplier,
thereforeoffersno solution.
Now exactlythe same criticismapplies to the Keynesiantheory.
Accordingto the Keynesian theorythe rate of interestis determined
by the intersectionofthe supply-scheduleofmoney(perhapsinterest
inelastic, if rigorouslyfixed by the monetaryauthority) and the
schedule). This
demand-scheduleformoney(the liquidity-preference
schedanalysisalso is indeterminatebecause the liquidity-preference
ule will shiftup or down withchangesin the incomelevel. Here we
schedule including
are concernedwith the total liquidity-preference
both the "transactions"demand and the "asset" demand formoney.
If we separate the total demand schedule for money into its two
componentparts,we could perhapsargue that the "pure" liquiditypreferenceschedule is independentof the level of income.' But
this does not help matters,since we cannot know,given the total
moneysupply,how much moneywill be available to holdas an asset
unlesswe firstknowthe level ofincome. Thus the Keynesiantheory,
like the classical,is indeterminate.In the Keynesiancase the money
supplyand demand-schedulescannot give the rate of interestunless
we already know the incomelevel; in the classical case the demand
and supply schedulesforsaving offerno solutionuntil the incomeis
known. Keynes' criticismof the classical theoryapplies equally to
his own theory.
1. In fact since expectations are influencedby the level of income this is
not a permissible assumption. The liquidity preferencecase is thereforeeven
weaker than here indicated.
429

This content downloaded from 195.78.109.96 on Wed, 18 Jun 2014 22:41:42 PM

All use subject to JSTOR Terms and Conditions

430

QUARTERLY JOURNAL OF ECONOMICS

Preciselythe same is trueofthe loanable-fundtheory.According

to the loanable-fundanalysis, the rate of interestis determinedby
the intersectionof the demand-scheduleforloanable fundswith the
supply-schedule. Now the supply-scheduleof loanable funds is
compoundedof saving (in the Robertsoniansense) plus net additions
to loanable funds from new money and the dishoardingof idle
balances. But since the "savings" portion of the schedule varies
withthelevel of"disposable" income,2it followsthatthetotal supplyschedule ofloanable fundsalso varieswithincome. Thus this theory
is also indeterminate.
In the loanable-fundtheory,the relevant supply-scheduleis
conceivedofin termsof loanable funds(i.e., "voluntary"saving plus
new money). In the neo-classical theory of Pigou, however, the
is conceivedin termsofsavingout ofcurrent
relevantsupply-schedule
income. "Saving is definedas the excess of total income received
over income received for services in providingfor consumption."'
Again, in the same vein, "aggregate money saving" is definedas
the "excess of money income over expenditureson consumption
goods."4 Here income, consumption,and saving, all apply to the
same period. Money savings are that part of currentincomewhich
is not consumed. Now currentincome is derived from current
expenditures. Whetheror not currentincomeis fedin part fromthe
injectionofnew moneyor fromthe activationofidle balances,makes
whateverfromthe standpointof the Pigouvian or neono difference
classical definition.5Income is income whetherit springsfromthe
spending of funds borrowed frombanks or fromthe spending of
"prior" income; and saving fromsuch incomeis saving even though
bank creditplayed a role in the processof incomecreation.6Accordingly,in Pigouvian or neo-classicaltheory,"saving" is in effectthe
same thingas "loanable funds." In Robertsonianlanguage,however,
"loanable funds" consist of voluntary saving (i.e., saving out of
"disposable" income) plus borrowedbank fundsand activated idle
balances. In Pigouvian language,saving out of currentincomemay
well exceed "voluntary"(or Robertsonian)savingin so faras current
incomeis increasedby bank loans or the injectionof idle balances.
2. "Disposable income" is here used in the Robertsonian sense, i.e.,
"yesterday's" income.
3. See A. C. Pigou, Employmentand Equilibrium, p. 30.

4. Ibid,p. 31.

5. "It is important to be clear about the implications of these definitions

when people or governments borrow from the banks. Everybody agrees that
money so borrowedonly becomes income when it is paid out, forservicesrendered,
to factors of production" (ibid, p. 30).
6. Ibid, p. 30.

This content downloaded from 195.78.109.96 on Wed, 18 Jun 2014 22:41:42 PM

All use subject to JSTOR Terms and Conditions

INTEREST

THEORIES

431

Thus the Pigouvian supply-scheduleof savingsamountsto the same

thing as the Robertsonianor Swedish supply-scheduleof loanable
funds. It is thereforenot necessaryto distinguishfurtherbetween
them,and hereafterI shall referonly to the neo-classical7theoryon
the one side, and the Keynesian on the other.
The neo-classicalformulationand the Keynesian formulation,
taken together,do supplyus with an adequate theoryof the rate of
interest. From the neo-classical formulationwe get a family of
saving-schedulesat various income levels. These togetherwith the
schedule8give us the Hicksian "IS curve." In
investment-demand
otherwords,the neoclassical formulationtells us what the various
scheduleand
levels of incomewill be (given the investment-demand
rates of interest.
familyof saving-schedules)at different
From the Keynesian formulationwe get a familyof liquidity
preferenceschedulesat various income levels. These togetherwith
the supply of money fixedby the monetaryauthority,give us the
Hicksian "L curve" (whichI preferto call the "LM curve").9 The
LM curve tells us what the various rates of interestwill be (given
curves)
the quantityof moneyand the familyof liquidity-preference
at different
levels ofincome. But the liquidityschedulealone cannot
tell us what the rate of interestwill be.
The "IS curve" and the "LM curve" are functionsrelatingthe
two variables: (1) incomeand (2) the rate interest. Income and the
rate of interestare thereforedeterminedtogetherat the point of
intersectionof these two curves or schedules. At this point income
and the rate of intereststand in a relationto each othersuch that:
(1) investmentand savingare in equilibrium(i.e., actual savingequals
desiredsaving) and (2) the demand formoneyis in equilibriumwith
the supply of money (i.e., the desiredamount of moneyis equal to
the actual supplyof money).
Thus a determinatetheoryofinterestis based on: (1) the investment demand function,(2) the saving-function(or converselythe
7. The classical theory may be said to coincide with the neo-classical or
Pigouvian theory in the special case in which no new money is-being created by
the banking system and in which idle balances are not being dishoarded.
8. Perhaps a family of investment-demand schedules, one for each level
of income. Everyone will agree that a change in the level of income affectsthe
volume of investment,but not everyone will agree that the level of income is a
determinantof net investment.

9. See my MonetaryTheoryand Fiscal Policy,Chapter5. The "LM"

curve representsa situation in which L = M in an equilibrium sense, L meaning

the demand for money, and M the supply of money. Similarly the "IS" curve
indicates a condition in which I = S in an equilibrium sense (i.e., the multiplier
process has fullyworked itselfout).

This content downloaded from 195.78.109.96 on Wed, 18 Jun 2014 22:41:42 PM

All use subject to JSTOR Terms and Conditions

QUARTERLY JOURNAL OF ECONOMICS

432

consumptionfunction),(3) the liquidity preferencefunction,and

(4) the quantityof money. The Keynesian analysis,looked at as a
whole, involved all of these. But Keynes never broughtthem all
togetherin a comprehensiveway to formulatean integratedinterest
theory. He failed to point out specificallythat liquiditypreference
plus the quantityof moneycan give us not the rate of interest,but
only an "LM curve." It was leftforHicks' to supply us with the
tools needed fora comprehensiveanalysis.
ALVIN
HARVARD

UNIVERSITY

1. Econometrica,Volume V, 1937, 147-59.

This content downloaded from 195.78.109.96 on Wed, 18 Jun 2014 22:41:42 PM

All use subject to JSTOR Terms and Conditions

H.

HANSEN.